This might be far out, but could it also hint that we are very unproductive while using social media and apps, and in general not creating any value (which to my knowledge are some of the biggest time-consumers when on mobile)?
I mean, when I'm on social media, I'm rarely cross-checking information, let alone doing something useful with the information I find. So I don't need to search. It's basically pure consumption.
Perhaps that's just the general thing about mobile - that people aren't productive while not near their computers. Personally, I never use my smartphone for anything then note-taking, and even that is a tedious task. That being said, I actually Google more often since I got an Android phone, due to the fact that the Google-search bar is on my front-page, and so I don't need to open my browser manually.
It's not inherent to mobile, but the form that smartphones have taken.
I remember owning a Treo way back and at that time, a phone was something I could easily compose text and be productive on. Ever since my first smartphone with a touch screen keyboard I've thought of the phone as something to read or play games on. Entering text has become enough of a chore that I stopped thinking of a phone as a productivity tool. It's good at keeping me entertained on the subway or being a music player, but that's it.
Swede here, and yes it works in swedish to, not just works its actually usable.
Havent tried to write a swedish sms with it but just saying "hitta närmaste butik" will give me the closest groverystores nearby and so on.
That's not true, I use Cortana in French all the time and it works pretty well. I always use "Hey Cortana, rappelle moi de faire tel truc dans 2 heures" (Hey Cortana, remind me of doing this thing in 2 hours), and it always understands me and creates a reminder.
Well, I'm Scottish and I don't have a particularly strong accent (at least by the standard of Scottish accents) and my accent appears to be beyond any kind of voice technology (Apple or Android).
It is better than you think. In my experience Italian, Spanish, and English voice recognition work almost flawlessly now.
I'm also learning Chinese using a mixture of tools on my smartphone, including speech recognition. I can't tell how good this is in general, but it is clear that the transcription works when I pronounce something correctly.
The worst performance I get is when I try to speak in my home (American English) dialect. "It can't see the forest for the trees" becomes "Kansas City forest with trees." So yes you may need to code switch.
A couple of years ago I found the voice search useless, but these days even e.g. searching for channel names on Youtube that doesn't even match dictionary words works near flawlessly despite the fact I speak English with a broad Scandinavian accept.
Same here. I have a pretty thick Spanish accent, but the Google voice search has improved so much that he understand almost everything I say. I still don't use it that often though, giving instructions to your phone still feels weird to me.
I'm not convinced nationality or accents have much to do with it, the overwhelming majority of American English speakers I know also can't use voice tools on their phones. My American English is pretty regular and it certainly never works for me. I suspect the common factor here is that the voice tools are junk, not that groups of people lie outside its area of function.
By contrast, I use VoiceAttack to tie voice commands to keyboard macros on my desktop, and it does what it's supposed to most of the time (but I bet it wouldn't if I had an accent.) Voice control definitely isn't an unsolvable problem, but it sure seems like one whenever I try to use it on my phone.
Moreover, voice functions are made so that they require you to be on-line, which cuts down their utility significantly. Also, I can type faster than it takes to say something and then wait for it to make a trip to Google's servers, get parsed and the result (that is just the search string!) to come back.
My first smartphone was an iPhone but I switched away. For minimal text entry like a search term I'll use the keyboard, it's easier than waiting for the voice input interface, hoping it works, etc.
For longer text, voice dictation also doesn't fix the problem. Sometimes it works but often you have to seek to arbitrary points in your sentence/paragraph and fix errors. I preferred a physical keyboard.
I agree, typing on a touch screen is horrible. The android "gesture typing" where you drag your thumb over the keys isn't too bad. I refuse to talk to machines, but that's just a personal thing I guess. I've never even tried the voice response features.
It feels awkward at first, but it really helps with a smartphone. You get used to it. I now use speech-to-text for the majority of the text input I perform on my mobile, including text messages, short emails, and search.
I also do everything I can to avoid the need for a lot of text input on mobile and save it for when I have a real keyboard. I agree with other users that the decline of physical keyboards on mobiles has had a terrible impact on their real-world productive use, but speech-to-text helps alleviate that a little bit at least, if you don't mind homonyms getting mixed in there occasionally.
Sure, but it really, really depends on the user. People with accents seem to have a hard time getting it to produce usable results. I only have anecdotes, but they seem to suggest that foreigners and first-generation English speakers have a hard time getting accuracy out of speech-to-text.
The one thing I'd really like to make voice control less awkward and faster is a push-to-talk button instead of the stupid activation phrases.
I want to just pull my phone out of my pocket, hold a button (or ideally hold a button on a smartwatch) and speak a command. Using the button release to finalise would also speed things up, because the phone doesn't have to wait for silence to know you're finished. It would also work better as a keyboard replacement. Select text box, hold button and speak to type.
It just seems so obvious to me that I don't know why no one's done it yet.
My only concern with voice commands is the tracking and search recording that inevitably takes place when using something like Google Now, Siri or Cortana.
If I don't use regular Google search due to privacy and tracking concerns, why would I use Google Now?
I have traditionally hated voice recognition since it's traditionally sucked. But Android voice recognition has gotten much, much better over the last few years. I certainly use it for search, and I'm even starting to use it for text entry at times. Coding would still be tedious "left paren function right paren", but outside of that, it's become quite usable.
Same phone, same story. The loss of a physical keyboard on my device removed the "input" use of it. Now it's really just voice and ready-only features. If there are websites I find or any other "work" that needs to be done I just forward it onto my computer and do it there.
I'll add too that I think the palm pre was the best form factor I've ever had: small, physical keyboard, yet gives me the full screen look and sleekness. And android still has yet to match it's universal search on type feature.
I've often wondered how many character clicks have been lost to typos and battling autocorrect on all phones.
While I like not having a physical keyboard, aside from text messages, I loathe typing anything of length on a smartphone.
I also find the UX of the iPhone's autocorrect to be really backward and frustrating; when I type a word, and I get a suggestion, I shouldn't have to lift my fingers from the keyboard to kill the suggestion. If I keep typing -- then keep what I am typing.
Ha! I had an Android phone for years and never used the search bar.
And yet, it's the most prominent feature in view whenever you see a "glamour shot" of a new Android device. It's a Google OS, so of course it's got to have a prominent search bar! But I never once used it—I suppose reinforcing the findings of this article. The ugliness (both visually and metaphorically) of a big Google search bar on the home screen of my phone is one reason I dropped Android.
It's not the only reason but it's one reason I don't use their launcher. Thankfully, it's dead simple to use a different one and doesn't require any weird hacks or jailbreaking. Nova Launcher is the closest thing I've found to an "ideal" Android launcher. Superficially looks the same by default but you can enable or disable the search bar, the persistent Google Now on the leftmost page (seriously, swiping up has always worked just fine for opening Now), and fixing their odd insistence on having so few rows/columns of icons and widgets when screens keep getting larger and higher-resolution.
I understand that you no longer use an Android device but hopefully this is helpful to anyone else who happens to read the comments and has similar wants. There are other launchers but I haven't kept up with them in a while since finding one that works best for me.
(in fairness, I had to throw a bunch of icons onto the Google Launcher desktop because it's not my default and I wanted to show how space is used. Also I put neutral-ish wallpapers on so as not to distract but they're not my usual ones.)
Yeah, exactly this. I find the stock iPhone-like launcher pretty useless. I ended up using Apex Launcher, because it lets me have much smaller icons and a denser grid.
Those are fresh shots I just did, no editing or setting things up. That's how it looks on my phone every day.
There are some display bugs, particularly with the system flashlight widget, but I got used to them and I can't live with anything that has bigger / less dense icons. I also find a simple black background to look cooler than a wallpaper (having a screen that displays black well helps).
Disclaimer: I work for an Alphabet company (but not Google itself any more).
I was a die-hard iPhone supporter since I got my 3G, but switched to Android when I started at Google back in June. One of the things I like best about Android is the search bar on the home screen - I use it constantly. Half the time what I'm looking for isn't a web page per se (directions, apps, email, etc), and having search integrated into the home screen is extremely convenient.
The one thing I really do miss is being able to hold down the home button to activate Siri. Saying "OK, Google" is only successful for me about 50% of the time (and I have a very standard American accent) - when it doesn't work I feel like Scotty in Star Trek IV: The One with the Whales: https://www.youtube.com/watch?v=LkqiDu1BQXY
I search more on my Android compared to my (former) iPhone but as my Galaxy S4 ages it has becoming increasingly slow to open the search app and slow to open the browser. Sometimes I avoid searching entirely just because the latency is too much work and involves jumping between the search app and chrome/firefox.
I'm curious but do you ever just search in the browser (skipping the separate search app)? I only ask because it never really occurs to me to use a second app to search when both Chrome and (my default) Firefox let you type in the URL bar to search directly. Same results, less steps and app switching.
I do search in the browser and this is generally quicker although part of the reason I search more often is the search box widget that I have on every home screen. Lack of support for true widgets is really an unfortunate gap in the iPhone experience for me but I may switch back for the next two years anyway as I'm not really a fanboy in either direction.
There are studies that suggest that teams that take time out of the day to joke and laugh with each other are still more productive than the teams that are heads down all the time.
(I'm grateful my natural inclination has proved out to be a good use of time, but now I don't even feel guilty about it, which is nice.)
If we were robots or Vulcans, perhaps socialization is a net loss, but for humans in general it is not. If you are trying to optimize your productivity, removing social contact is not necessarily moving in the right direction; the productivity optimum is not "zero social contact".
Of course "social networks" may be low-grade or negative socialization. Perhaps the fact it is "low-calorie" social interaction causes people to spend too much time on it when they could do better by other mechanisms. I don't know, really. I'm just pointing out that assuming social == not producing value is an oversimplification, if you're a human being.
I don't consider social networking to be about social contact, because the physical contact is missing and most of the time we aren't even talking with our "friends" on these social networks. A "like" or a retweet is not a dialog.
Phone calls, SMS messages, IRC chats or even emails are way more healthy than whatever happens on social networks because they are bidirectional. And I'm not even speaking about the magic that happens when shaking somebody's hand or when receiving a smile.
Yesterday I realized that it has been a long time since an old friend reached out by email to ask how I was doing. I still do, but I also don't use the most popular social media channels. I wonder if the simulation of contact provided by broadcast-style social media is enough to satiate the need to reach out and make direct contact. Perhaps I should delete my page from facebook. Despite the fact it hasn't been touched in four years it might give people the impression that they know how I am. It is easier to find me by actively searching for my name on the web. Then, even if one does this I've heard that facebook will sometimes make a stub page from media they have available to stand in. It feels like a nightmarish trap in which we train machines to supplant ourselves.
My close group of a friends always has a Facebook Messenger thread going. We went to college together but have since moved to different places all around the Midwest. We make time to see each other in person a few times a year, but the main way we keep in contact is a Facebook chat that's constantly rolling.
Well yeah, that's a reasonable use-case, but how often does it really happen? Plus you're probably talking about a handful of friends, but what about the other 150 people or more in one's list?
I disagree. For those of us who use social networking more for posting updates and discussing things with each other, it has lots of social-contact value. The two benefits it provides that aren't simultaneously given by phone calls, SMS, IRC or watercooler meetups:
- broadcast but pull (not push) communication; everyone (from the group that can see a particular post) can join or leave a discussion at any time, and reading it does not require commitment to active participation
- asynchronous communication; I can drop a comment and read replies at my own schedule, and there is no expectation for immediate response
And no, you can't really disconnect it from the other side of social networking, i.e. posting cat pictures and news articles - they provide a social object, around which a discussion can form[0].
Complaining about social networks as "inferior" form of social interaction seems like complaining about potatoes as inferior food because they weren't around in Europe 500 years ago. Yes, living only on potatoes isn't the healthiest thing, but they are a decent food and even if you make it the core of your diet, you won't get sick.
Oh, I would never say that potatoes are inferior, as that would be the same reductionist science mentality that has been making us obese and sick. In fact a social network like Facebook is nothing like the potato, but more like the Chicken McNuggets or the sweetened corn flakes or the soda made of high fructose corn syrup, or the ultra-pasteurized and low-fat milk, or the beef from corn-fed cows that need antibiotics because of liver abscesses that happen for being fed corn and animal tissue instead of grass.
Just as with fast food that depends on subsidized corn, ignorance and fossil fuels, social networks like Facebook represents the industrialization of our social interactions: fast, cheap, shallow and designed for mass consumption. And make no mistake, just like how the food demand is inelastic because our stomach doesn't grow bigger, so is our attention span as there are only 24 hours in a day, so the same strategies are employed by both sides, with Facebook being the new TV (how ironic, almost as ironic as organic food being industrialized).
And going back to the potato, we Europeans may have started eating it only 500 years ago, but the potato has been domesticated 7,000 to 10,000 years ago in South America. We simply discovered what other people have been eating for thousands of years. No such thing happened with social networks. Or with present day processed foods for that matter.
And you know, industrialized food and the modern lifestyle has been making us obese and gave us diabetes. We might also discover that this modern lifestyle is also bringing with it other gifts, like autism and I wouldn't be surprised at all if some conditions from the autistic spectrum are also caused by humans having shallower social interactions.
> (I'm grateful my natural inclination has proved out to be a good use of time, but now I don't even feel guilty about it, which is nice.)
Are you sure you are not just experiencing confirmation bias?
A lot of very high-octane code has been written by devs who eschew social contact (any distractions, really) for the duration of the push, so I suppose it just depends.
I have a hard time believing that any "study" can really tease out optimum productivity for all cases.
It could be something where a short term optimization does not apply to a long term problem.
Take running. Trying to get from point A to B as fast as possible requires different strategies depending upon how far apart they are. A very short distance is best done by going all out 100%, but if you do that for a marathon you'll be beaten by someone who paces themself.
Your analogy about running is very good, althought I think there's also another aspect - a lot depends on personal idiosyncrasies and relation between a person and a task. For instance, if I'm free to work on my personal projects, I prefer not to have any social contact with anyone except pizza delivery guy for extended periods of time. On the other hand, I need Hacker News and Facebook Messenger to get through the day at $dayjob.
"Are you sure you are not just experiencing confirmation bias?"
"I did a thing" -> "Study comes out which says the thing is a good thing" does not conform to any useful definition of "confirmation bias" I can think of. (I can stretch it, but the result is not a useful definition.)
I would really like to see those studies, and a link would be great. Our company is on Slack and while we all work hard, we take some time during the day to spontaneously crack jokes and send interesting articles around.
I can't speak for everyone in the group, but I feel a little recharged after a flurry of funny gifs or esoteric programming jokes.
I tried to Google them up before I posted, but I couldn't come up with specific-enough Google terms in the sea of "social". It's been on HN recently; HN being what it is, keep your eye out, it'll roll around again in the next few months.
this is interesting, it does explain why now on tap is such an important feature for Marshmallow. lets you cross check without switching out of the app
Or it could indicate that Charles Arthur's analysis of the data provided by Google is flawed and so using it to speculate on productivity is a bit of a fool's errand :-)
It is. I've actually disagreed with him in the past in some obscure correspondence I sent him (so minor he won't recall it!), but he's not always wrong :-)
I've responded to your comment there. Obviously, the data is open to lots of interpretations. I tried to use models that seem reasonable and reflect what we know about internet behaviour - particularly the Pareto distribution of most activities.
Communicating, managing projects, taking notes, reviewing/signing documents, sending payments, drafting designs, outlining documents, finding/ordering kit -- are these not productive?
Yeah, you're probably not doing hardcore document creation, or real research, away from a keyboard, comfortable chair, and a nice big screen, but that's not the shape of all productivity.
For some users, sure. But that'd be about the user, not the platform. And more specifically, that'd probably be mostly about the time and place of use, not even the user.
If those same people had a 27" screen during that same slice of time and space, they'd probably be doing the same thing.
I think apps have just done a good job keeping you in their little walled garden - they don't want you browsing the web, checking other pages, etc. They want your eyeballs on their screen so they can sell ads.
I think it's easy in theory (and probably hard in practice) to solve this problem. Google just needs to reduce the friction of web searches. I think they've gone a long way to achieving this goal on Android 6.0. For the uninitiated, they added that weird hold to search which automatically picks stuff out from the screen to search and often presents results in a way where it's really easy to switch back to what you were originally looking at.
Imagine if all you had to do was think about something to initiate a google search. Okay, that would probably get pretty dystopian but it would let you do many more searches even while your screen has been abducted by an app.
I think all google needs to do is reduce the frictions to search.
There are also some productive things that are doable with mobile devices. For me that is learning flash cards with Anki and data acquisition for openstreetmap (house numbers, etc.).
The problems they have is usually the small screen that makes it hard to view much content at once and that text input is really burdensome. Some tasks require these and are thus not suitable for the mobile devices. They however also have advantages to be usable in many situations computers aren't (for using a computer while standing you need a special desk while you can use the mobile devices in almost every position and place) and touch input may be better for some tasks.
For stuff that would be possible both for computers and for mobile devices equally good there's the barrier that it's already been developed for computers and that everybody needing it for work already uses a computer, so you would need to put a unbearable mass of work into it to bring it to already is with computers today.
I don't think people get less productive using smartphones (but I'm not 100% sure that that's really what you are implying). I think the demographic of web users has/and is still shifting. Desktops have always served the more technologically inclined people better and left less technologically "literate" people behind. Thanks to smartphones a whole new demographic of less "technologically literate" people gain access to the "web" (although mostly through native apps). It should be no surprise that we see less web searches, relatively speaking. The only move Google can do to capture this demographic as well is ever deeper integration of the search engine into Android.
I do search on Mobil as one offs. Quick search. When I'm looking to buy or compare I do it on desktop where I can have multiple large windows open which I switch back and forth and do additional searches based on serendipity.
On mobile I feel "caged in" psychologically, but that may be influenced by having had the desktop experience as my first and also my primary experience in search.
Definitely. You soon reach the limits of web navigation on mobiles. (Javascript, screen size, forms, etc).
What could help, is if there was an easy and smooth way to pursure browsing started on the mobile on the desktop, and vice versa to continue reading a document started on the desktop on the mobile.
One very good thing that Youtube does, is to record in the history the point at which we're in a video, so that if you browse to the history on another browser on another computer, it will continue the video from the point you left it. This allows you to switch between desktop and mobile and vice-versa. Cats can be watched on mobile, but if browse to a conference with slides and you want to take notes, you definitely want a bigger screen and a desktop computer.
It's similar for the browsing (including the bookmarks, and the tabs). If they were shared smoothly, one switch more easily between the two devices.
> What could help, is if there was an easy and smooth way to pursure browsing started on the mobile on the desktop, and vice versa to continue reading a document started on the desktop on the mobile.
Apple has that; they call it "Handoff". It works reasonably well.
I'm not 100% sure it means people are doing less searching in general, though. I still search a lot with my phone, but rarely using Google. Instead I search Yelp for businesses, search Twitter for news, search Facebook for people. I think it's more about a move in behavior to being more app-centric.
A big one for me is also: search Wikipedia for [lots of things]. On desktop I probably default to the web search bar for this kind of search, with the feeling that if a Wikipedia article is relevant, it'll probably be the first or second web search result anyway. So I don't explicitly go to Wikipedia first. But on mobile I default to trying the Wikipedia app first, mostly because the browsing experience is so much faster, and the pages more readable, at least on my older phone.
I think the speed of the experience is core. Browsers on mobile eat up a big part of device's resources, and the time it takes to load it and perform the search is so big that it doesn't make sense to use it for anything else than general web search. If "there's an app for that", one will use an app for that.
Just as a counter data point, if there's an app for that I rarely use it. Because usually I prefer the mobile browser version to the app for many reasons, not least of which is the way to get to the web version is always the same.
There is a wikipedia app? This is one of those cases where a little nag on their website pointing out they have an app would have been nice. I had no idea.
Although, I use exactly your desktop pattern on my phone. Go anywhere there is a search bar (they're all over on Android), and search "wikipedia foobar" and go straight to the first wikipedia hit. Works better than wikipedia's own search usually.
And their mobile site is actually pretty nice. I like it at least. I'll check out the app though.
The mobile site is pretty nice too, and pretty similar to the app. If you have a recent phone and are usually on fast data connections the app isn't as compelling, though I still like it a bit more.
One interface feature I like with the app is within-article navigation: if you swipe inwards from the right edge, you pull out a table of contents, and can easily jump to article sections by tapping one. This is useful enough that I find myself unintentionally trying to do it in mobile Firefox, hoping I can jump around a blog post or documentation's headings that way (I wonder if mobile FF could implement this? do enough pages use HTML headings in a semantically meaningful way?).
This is more a stylesheet choice, but the app also more aggressively collapses things other than the main article text, like infoboxes and tables. On the mobile web view I find I have to scroll down too much to get past big infoboxes and such.
On slower phones and slower data connections it's also speedier. Especially on longer pages and pages with more non-text stuff, my phone has a noticeable pause to render the page after it loads, while in the app it's nearly instant. The data transfer via the app is also low enough that it's usable even on slow connections (I'm semi-frequently on T-mobile's free global roaming, which is rate-limited to 2g speeds, ~128 kbps). Since the interface is already preloaded, all it needs to do to load an article is grab the article text through the MediaWiki API, which is usually in the range of 2-20kb. It then lazy-loads images later (which you can disable if you really want to be lean on data).
1. 50% of people do one or more searches on mobile. That's awesome from Google's POV!
2. The number of people who are using mobile for their Internet has exploded in only a few years, many of those people still use their desktop for search.
3. Many, many people still don't use their phone for anything other than texting and phone calls.
But here is the kicker:
4. Charles Arthur is extrapolating a measure of central tendency in a rather foolish way. He's taking the mean average of the number of searches from mobiles per day for an entire month, then he's extrapolating that none of those mobile users do any searches per day for an entire month.
In other words, he's saying that half of all users don't do ANY searches at all, whereas all that he can say is that on some days, some users don't do any Google searches on their mobile device.
And if that's all his statistics are showing, then I'm not exactly blown away with this insight.
> Many, many people still don't use their phone for anything other than texting and phone calls.
This is me, well except the texting part. I only have a smart phone for the occasional time I need navigation. The vast majority of days I use it to make one phone call.
That navigation part makes the cost to upgrade every few years worth it though.
"In other words, he's saying that half of all users don't do ANY searches at all, whereas all that he can say is that on some days, some users don't do any Google searches on their mobile device."
From the article: "who does what searches isn’t fixed; so someone who did zero searches yesterday might do 10 tomorrow. But equally, the 10-searcher yesterday does none or one or four today. And so on."
So yes. But on any given day, half of people don't do a mobile search. And the numbers from Google suggest that the average across a 30-day period is less than 30. If you think there should be a different model for how many people do how many searches than a Pareto distribution, please offer it.
I probably need to re read your article, but where are you getting the assumption that 50% of people aren't doing searches on any given day?
How do you know the spread of data isn't that there are 20% of users who don't use their mobile device for the Internet for 10 days in a row, but then do a lot of searches in a short space of time?
Maybe I'm missing something here, but by taking an average over 30 days, you seem to assume that users do similar numbers of searches every day.
Is 'searches per day' an important metric? It seems obvious that someone on a desktop PC doing a navigational search for 'facebook' and immediately clicking on the first link only costs Google money. Google get zero benefit from it. Eliminating those searches would be a good thing (so long as people still use Google for searches that Google can wring some cash out of).
Fewer navigational searches would increase the proportion of monetisable searches; wouldn't that be a good thing for Google?
Yes, that was in an Incognito window and there were plenty of ads for other SERPs. It's Google so it would vary based on tons of factors (user agent, location, if I have a cookie, etc etc).
That doesn't matter if people don't click them. If someone is doing a navigational search they're overwhelmingly likely to click the first result. The amount of revenue navigational searches generate compared to all other searches must be negligible.
If I have understood this correctly, Google are losing out on people searching for Yahoo, Google, Facebook and other large web properties, which people either can't be bothered to type in full into the address bar, or are clueless to the fact that you can type an address without using Google.
In which case, do these lost searches actually reduce revenue in terms of valid as click throughs? I would argue that most people searching for Facebook for example are simply going to click on the first Facebook link the results.
If Facebook really are going to advertise (paid) in Google for the search term Facebook, then obviously Google miss out on a relatively pointless ad, but in most cases, I can't see the loss here. Why would Facebook advertise for the term "Facebook"?
Unless Google has a policy saying no one else can advertise for the term "Facebook", Facebook has to advertise for it to prevent others from getting their ads in front of people who are looking for Facebook.
Search the name of virtually any large company, and their website will come up as both the first real result and the banner ad at the top of the results.
I know _ad hominem_ is generally uncalled for, but he's not known for doing anything else. As the Guardian tech editor he had a reputation for showcasing (i.e. Boot Up, which got pretty toxic) and writing articles attacking Android. Most of it based on tedious market analysis.
No, but I tend to consider people who call hardworking Wikipedia editors by the pejorative "wiki-fiddlers" less than reliable, and the man regularly wrote half-truths and sometimes resorted to outright dishonesty.
That you would publish him... Well, that's your judgement but I consider it to be lacking.
He never said that the mode was 0.5... he said "The mode (most common number) will be below [1 search per day] too."
And it is. According to his numbers, the mode is precisely 0. The most common number of searches is zero... more people do 0 Google searches in a day than do 7 or 1 or any other specific number.
This is not a USEFUL statistic, but (according to this data) it is accurate.
Actually, it's not. He's taken the mean average over the course of 30 days to get the average number of searches per day for all users. That doesn't give you the average number of searches per user per day!
In order to get the mode, you'd need to know how many searches each person does each day. You can't get the mode from the figures he's been provided.
But in the article, he wrote:
"I tried modelling what search activity probably looks like on mobile: I used a mean = 0.925 (as per Singhal) and mode = 0.5. The mode must be below the mean because of the long tail of higher values; 0.5 is a guess, but moving it around doesn’t have a large effect. This gives a median of 0.94, close to the mean, which you’d also expect."
He made some guesstimates which he used as input parameters to form a Pareto Distribution.
As for the mode always being less than the mean, that's wrong too.
I do many searches a day, but I use the built in search in iOS... pull down from the home screen and get a search box. Hits apps, emails, everything in an integrated search. The backend of this is "SIRI" (I believe it's the same platform that siri voice search uses) and ultimately it may result in google searches. For now.
My career has spent a fair bit of time working on search engines... google's moat is that everyone uses it, it has the audience. Providing competent search is something a company the size of Apple could replicate with some effort.
I thin the threat to google is real... I think google knew this a decade ago, and that's why they bought Android Inc. and developed the android OS.
I'm aware that my own experience clouds my perception but it's still surprising to me. One of the main uses of a smartphone for me (even since my first Treo) is that I've got a "pocket Google". Messaging and social media and games are all great but for me, having a pocket reference the size of a massive global library remains one of the "killer apps".
I used to joke that smart phones are the equivalent of Penny's "computer book" from the Inspector Gadget cartoons I watched as a kid. I remember being in grade school watching that show and thinking how awesome it would be to have a powerful computer I could carry around with me all the time and use to find any info I wanted.
Serious question: have you ever made a purchase because of (indirectly or directly) an ad that appeared on a search page? (Maybe this should be an 'Ask HN')
I've gotten a job because an ad for a job search website that looked interesting, I've found out about new products and (after some of my own searching) purchased them, and i've definitely gone back to products i've left because of newly advertised features. (Microsoft Azure being one of them)
I don't remember if these were specifically on a search page (chances are they weren't) but they were ads on the web.
I would be careful of a self-answered question like this, I would suspect a lot of people will say 'no' when the actual answer is 'yes' just because they didn't even realize they were clicking an ad or because it's something they do so rarely they think they never do it when in fact they very rarely do it. But very rarely is in fact 'woohoo loadsa money!' for everyone involved.
I'm fairly certain that in the last 10 years or so I've never bought anything because of a web ad or a TV ad for that matter. You could argue perhaps there is a subconscious element that drove a later purchase, but even if true there is no directly measurable correlation for the ad networks.
One perfectly valid outcome for an ad is simple awareness. If I go to McDonalds, I know the McRib is on their menu because a billboard said it was. The billboard didn't make me buy one, but I know it's an option. When I go to the store, I see a dozen brands of soda. I know Pepsi and Coke are popular choices because of the number of ads I see for them. I've never seen a Mr Pibb ad, so that brand seems kind of weird to me. For that matter, I know McDonalds is a popular restaurant simply because they run advertisements.
The goal of many ads isn't to make the consumer say "wow I need to buy that now!". It's so when the consumer is presented with a handful of choices, they say "yeah, I know this brand." And is certainly is measurable, in brand awareness studies.
I'd say so, yes. Even seeing the name being repeated helps. I've never played Game of War nor do I know what the core gameplay is like, but I've seen enough ads to know the name and tell you it's a mobile medieval combat game that's free to play.
The hardest part of relying on brand awareness in ads is you have to repeat the ads over and over and over, and hope people start to recognize the name before they get sick of seeing the ad.
Thank you for your concern, but those answers would be fine. I'm thinking of buying some adword ads and this article called the process to mind. Rarely is as good as never for my purposes, and clicking an ad that they didn't know was an ad is also a non-vote.
I expect a flamewar about whether or not people actually click on ads. I'll be happy to participate, because I feel the need to vent out some of my issues with the advertising industry ;).
Oh, and to answer your question - personally, no, not directly through an ad.
That is why I use my mobile; when I'm walking in the mountains during breaks in my head i'm working on things which I can do without sitting/standing behind my laptop which usually means researching things I don't know or that I could not get to work.
That's how I feel as well. Having my phone on hand means I have an answer to almost every question. And I wonder about a lot of things. I really cannot count the number of times I'm driving talking to my wife and we get to a question neither can answer when "Ask Google" is my response. We use voice search on our phones a lot. A whole lot.
Hardly surprising. On Google desktop most queries are navigational these days, something that's unnecessary when using mobile apps (spotlight search). And doing research on mobiles is a chore.
I suspect this is a naive guestimate by the author. There are many ways to search that don't involve going to google.com in a browser, and with the combination of Google Now and Google Maps/Nav, people surface a lot of queries through organic behavior. It can't be overstated, too, that Google is making the majority of the money from in-app advertising, so that additionally mitigates the decline.
If I were an investor, I might be worried about some aspects of Google, but this isn't one of them.
I'm probably biased, though. I am constantly "ok googling" to learn stuff. Having young children will do that to you.
There is also another factor that may influence this statistic: most android phones in the general population are slow as hell. While in IT, especially in mobile development, phones are fast-ish. Developers keep adding features (OS as well as app) when they should be doing performance optimization.
I have a 1 year old, low/mid range phone I bought because I just didn't have the cash for a flagship. I have only the "most common" apps installed. Mobile web is unusable. Opening chrome takes 2 minutes, doing the actual search a minute more, rendering the site ... well, let's just say I have JS turned off. Unless I don't have access to the internet some other way and it's really crucial, I'm just not going to bother.
Those are all carrier locked, so the actual cost of the phone might be higher than the retail price. I don't think net10 is spending huge dollars making their phones cheap, but I'm sure they spend a bit.
Yes but unfortunately people look at up front costs and not end costs.
$200 to someone living in poverty is more than a month of food, while $20 up front + $40/ month for a baseline limited text and call plan seems far more manageable.
I have a 4 year old phone. Without all of the kernel optimizations, forced removal of Google App and Google Play Services, I wouldn't be able to do anything. My battery lasts 4 days of normal use. Before my changes it couldn't last 10 hours.
So much amateurish bloat. I'm baffled by those engineers.
How did you force remove the Google apps and Google Play Services? Did this require root?
I have a Samsung Galaxy Note 10.1 2014 edition that I bought in February, and it lasts nowhere as long as my wife's 2012 iPad3. Such a disappointment. I will sell it, last Android device I buy.
Of course, root, without it I couldn't do a thing.
I have Samsung Galaxy Note 10.1 but older edition, battery is still fine, the OS got slow, even on factory reset, so, I decided to root it and runs much faster - pen use was not that great before, now it's perfect.
Note that the act of rooting didn't have anything to do with the speed up, after installing a custom OS with GPU optimizations for drawing graphical elements it worked great.
CyanogenMod is basically a stock Android without any of the Google apps, services or frameworks.
You can always install a very minimal version of the GApps later if you really need them.
Look for a Nano or Pico version. I can't remember which is the smaller of the two, but it basically just comes with Play Store and the Framework to make it work.
At the same time, isn't it nice that you can do these things? Every so often I look for similar projects or hacks to revitalize my old iPad2 but they don't seem to exist. At that point I don't care about warranties or the silly insistence that I can't be trusted to grant superuser privileges on a device I own. But even jailbroken, I've yet to find a way to reliably "slim down" the iPad's set of running processes so it's painful to use even as just a web browsing device.
I used to work at a place that gave everyone 2 machines for development: Their regular desktop, and a 5 year old machine, with a throttled internet connection. The programs had to be run on both. I wish more companies would do this.
Especially in mobile. Sure, your site loads fast when your in a major downtown area, but try using your 30 different java-scripts when your in Nowheresville Alabama
I was recently basically forced to get a new phone because my Galaxy Note 3 had become virtually unusable from software updates over the past few years. Common tasks would freeze the phone for seconds at a time. All animations were laggy. It was as if it was running on 10-15 year old hardware, not 2 year old hardware.
Just to be clear, I've experienced similar degradation with Apple products as well (specifically an older iPad), but never as severe as with the Note 3. My old iPad is still definitely useable; it's just not as smooth as it once was. The phone on the other hand had just become complete junk.
Meanwhile my girlfriend's Nexus 5 phone has been on a roller coaster of battery life changes. Not too long ago, an update caused her battery life to drain to 0 in roughly half a day of light usage. Then it was fixed. Then it was ruined again in the update to Marshmallow. She's going to have to get a new phone as well because hers too has become nearly unusable.
I agree. I damaged the charging port on my N5 and while it usually works, it occasionally doesn't and I didn't feel like depending on it. Still, I kept it around when I bought a Moto X last year and still fire it up occasionally and use it as a backup phone. I'm always surprised when I fire it up and (once it finishes syncing all of my email and stuff) it still runs great. I actually miss how the display is bonded to the glass so that there's just a little less of that barely perceived disconnect between my finger and the screen. Didn't think much of it but when I go back to the N5 it's a pleasant surprise. I feel like the last couple of years have seen mobile devices really hit a point where even a $350-400 device can remain useful for several years. Anything I own that's older (iPad2, Galaxy S2) ran like a dog within a couple of years.
Nexus phones are notorious for the rollercoaster of battery life. Before Marshmallow, my N6 would be dying by 5pm with random, low-medium usage. With a Marshmallow build from XDA, I hit 20% at 9pm. It's insane. The battery stats looks roughly similar from either OS though.
This isn't a rebuttal, but there are often (better) alternatives to the firmware that ships with devices, especially if it's a carrier maintained version.
I've used Cyanogenmod on my smart devices for as long as I've been buying them. CM has had sandboxing for apps for a few years now, and I was running 5.1.1 on my S3 months before it was available for my SO's S5 through her carrier. CM13, announced recently, will be built around Marshmallow.
For older devices, you can either try the newest CM versions or go back to earlier stable major releases. And they're all bloatware and CarrierIQ free.
I stayed away from Cyanogenmod because... well, the rule of conservation of hackergy dictates that playing around with rooting my phone would take me away from hacking electronics and constantly reconfiguring Emacs. But at this point I think after I get back from China, if I won't buy a phone here, I'm going to flash my S4 with Cyanogen, praying it will help - the way the phone has been slowing down for the past two years is utterly ridiculous, and I blame it on overheating CPU or something because I just can't believe this shit could be software - neither the system nor the apps I use should ever need that much processing power.
Oh yes. My S4 has been slowing down over the last 2 years to the point I simply don't do things like browsing too much, because they involve spending 30 seconds waiting to switch between apps, etc. Many sites are unusable due to the amount of heavy JS processing they do. I stick to native apps as much as I can and I'm thinking about getting a new phone.
If you don't mind sharing, what kind of phone is it and how much did you pay? I have a 1st gen Moto X which is now over two years old and the tasks you mention feel pretty much instantaneous. I know that the Moto X was a high end phone when it came out but I bought it used last summer for $100 and I imagine that the specs were comparable to inexpensive new phones at the time.
It's a Samsung GT-I8260. I got it on contract (I was really low on cash and with a broken phone) and it was marked at 160€. It's not unlocked and not rooted. Cynogenmod is not supported for the device (as far as I know).
What phone? I have a Nexus 4 running Android 5.1.1 (Lollipop). It's three years old and still works just fine. Battery life has decreased a bit as expected. It has some quirks but overall performance is on par with when it was new.
Using a phone for search is like sucking in information through a straw. Even if you do find something, nobody really wants to setup copy-paste infrastructure between their phone and desktop/laptop. Yes, I know about AirDroid. Loading almost any page is going to be painfully slow, render poorly, and then break completely. When I have loaded a page, I use ctrl+f to search for whatever I was looking for, and I use a large screen to render all the content so my eyes can lock on to whatever I am looking for. Going through 10 different results is just painful on a phone, and on a laptop I can do it in less than a minute. Small screen and touch doesn't benefit me enough to switch from desktop search to phone search...
Actually, if you just type your search phrase in the URL bar and scroll to the bottom of the suggested URLs that pop up the local page search results are there as well.
Yes, but I am often looking at 10 or 20 things at the same time. I don't need to "read later", I need to read all of them now. I'll probably lose like half my trains of thought if I have to wait multiple seconds to switch back to page rendering one of the other umpteen search results...
Well, yeah, at the scale of "real research" it doesn't really work. But you wouldn't want to do that on a tiny screen even if copy/paste was up to it. You probably wouldn't even want to do it on a laptop screen. That's just a tool/task mismatch.
I assume the idea is that you're searching on your phone when you're not at your desktop. Obviously if you have a "full" computer in front of you, you're not gonna switch over to your phone to search. It's for all the times when you aren't at your desk or on the couch with a laptop.
Although, I do search on my phone while at my desktop in exactly one scenario: playing games. It's less hassle to do a quick search for where to find item X in some random game than to pause, alt-tab out, fire up a browser, search, try to remember results, close browser, and alt-tab back to the game. Otherwise, yeah, full browser when I'm in front of one, phone for the rest of my life not spent in front of a computer.
My only thought when I read this was that Siri, Cortana and Google Now are getting used more. This means a lot of searches that used to be done on the Google search app are now being made through other means.
The author needs to clarify what portion of total revenue Google derives from ads, and then to break those ads out by service (Youtube, Gmail, Sponsored search), platform (desktop vs. mobile), and geography. Those stats probably aren't publicly available, but it would be good to have a rough estimate so we could have a more nuanced conversation around relative values rather than just one big pot of "ad revenue".
Otherwise, you're really just fumbling around in the dark, saying that monetizing mobile searches with such a low average daily click rate are a big problem for Google when the reality could be that they've never made much money from mobile search ads, especially in developing countries where they are adding the most users whose ad clicks are worth less because they have less disposable income than the average American or European. Similarly I'd wager that mobile search ads are a very small portion of Google's overall mobile business, and that Google never planned to make much money from mobile search in the first place, understanding that building out the app ecosystem and getting Android onto a ton of phones was more important in the long run.
People don't explore on their smartphones: they consume.
This as a cool thought experiment that could have been improved with more insight into breaking "ad revenue" into meaningful chunks to see whether or not Google actually cares about the slice considered in the headline: mobile search ad revenue.
"The author needs to clarify what portion of total revenue Google derives from ads, and then to break those ads out by service (Youtube, Gmail, Sponsored search), platform (desktop vs. mobile), and geography."
The author (me) would love to, but Google makes very little data available. The only geographic data is total revenues broken down by US/UK/Rest of World. There's also revenue from "Google [search]", "platforms" and "other". Estimates about YouTube etc revenue tend to be calculations based on various private metrics.
"I'd wager… that Google never planned to make much money from mobile search in the first place"
Possible. As the article says, and has been repeated multiple times, Android was a defensive move to stop Microsoft cornering the mobile search field.
"People don't explore on their smartphones: they consume."
I must introduce you to some networks called "Instagram" and "Snapchat" and "Line" and "Weibo" and "WhatsApp".
I wonder what's up with those annual 15%-or-so dips. Maybe a religious holiday? I'd love to see an X axis there. And does it continue to this day? (The photo is from 2011.)
People use mobile phones in addition to desktops. They might be doing a price check on a mobile phone and then actually purchasing on their desktop.
They might be reading the news, which is roughly 50%-99% PR placements depending on the publication, which pushes them down some sales funnel or another. They're just browsing higher up in the conversion funnel. Searches higher up in the funnel tend to be less valuable in terms of costs per click than searches which are closer to the purchase.
Everything in the news is marketing anyway, and it's just as paid for as it is when it's an ad -- often times it's more expensive, because access to the news is rationed by publicists and PR firms rather than by the advertising department. Further, the bulk of social media is the effluvia of PR as it reverberates through people who talk too much on the internet. As far as Google is concerned, that propaganda consumption is just likely to stimulate more searches later on which can be advertised against.
The flow might be: Check Facebook App -> Click Gossip Rag Article -> See fabulous new shoes in celebrity photo -> Order lunch on Seamless App -> Go home from work -> Remember to Google "buy red high heels kim kardashian" -> Click ad -> Price check, bounce (Google gets paid) -> "red high heels discount" -> Click ad (Google gets paid again) -> Buy shoes.
When they are ready to buy, they will pay for the $10 click -- maybe multiple times across different advertisers -- on their desktop or $15 app download on their mobile phone.
The headline that this author uses is also really loosey goosey. Google's ad products on mobile are also increasingly more biased towards generating paid phone calls. When you search for a locksmith or auto repair, you want to get a guy on the phone right now, and are likely to need to buy immediately.
So, this isn't really Google's "growing problem" -- Google as an organization is set up to use machine learning to figure out how to make money off of mobile search behavior. The company is still learning what it has to do to make money off of mobile search behavior and how it differs from behavior on desktops.
Also, as far as Google's customers (advertisers) are concerned, if a desktop search results in people subscribing to an e-mail list that they read on their phone, the advertiser earned measurable results from the money that they gave Google. If their customers don't like to buy on mobile or they behave differently, campaigns can be configured accordingly.
My search pattern changed when I got a useful phone. The most useful was my old Droid 4. It had the real keyboard, and I used the crap out of it. The phone itself was no winner, but it could render search results reasonably, which is what made the whole thing worth it.
Those changed a little with a Note 4. It's a much more powerful phone, and I do still search on it, but touch input is a modest PITA. I cut back on searches, but I have added some voice use cases too.
In my experience, people with powerful phones use them more. A lot of phones out there aren't so powerful, and they keep it to social, MMS, etc...
I actually disallow FaceBook on my phone. It's a battery hog, and I've better things to do. FB once a week on a laptop is enough for me.
It would be super ironic and fascinating if Google's low-cost mobile operating system was what wound up killing their search business. How long until they start inserting ads into their app store?
What I got from this article is that Google is making money every time I open a web browser on a desktop, type in "facebook", and click on the top search result. Now on mobile devices, there's an actual app for Facebook, and opening the app is cutting Google out of the loop (and the Money). Is that right?!
If that's the case, then it seems pretty shitty from the advertiser's point of view. Especially since Google came up with the all-in-one search and url bar "feature" in Chrome.
245 comments
[ 2.9 ms ] story [ 244 ms ] threadI mean, when I'm on social media, I'm rarely cross-checking information, let alone doing something useful with the information I find. So I don't need to search. It's basically pure consumption.
Perhaps that's just the general thing about mobile - that people aren't productive while not near their computers. Personally, I never use my smartphone for anything then note-taking, and even that is a tedious task. That being said, I actually Google more often since I got an Android phone, due to the fact that the Google-search bar is on my front-page, and so I don't need to open my browser manually.
I remember owning a Treo way back and at that time, a phone was something I could easily compose text and be productive on. Ever since my first smartphone with a touch screen keyboard I've thought of the phone as something to read or play games on. Entering text has become enough of a chore that I stopped thinking of a phone as a productivity tool. It's good at keeping me entertained on the subway or being a music player, but that's it.
TLDR: They all suck. Terribly.
Assuming your users will fall back to voice-search is not an assumption you should make without having the data to support it.
Plus the voice recognition's accuracy is absolutely stunning for me; even in a crowded bar.
https://www.youtube.com/watch?v=sAz_UvnUeuU [NSFWOS]
Captures both the failure of the technology and the resulting anger!
I'm also learning Chinese using a mixture of tools on my smartphone, including speech recognition. I can't tell how good this is in general, but it is clear that the transcription works when I pronounce something correctly.
The worst performance I get is when I try to speak in my home (American English) dialect. "It can't see the forest for the trees" becomes "Kansas City forest with trees." So yes you may need to code switch.
And yet, I refer to my phone as "he".
>so much that he understand almost everything I say.
By contrast, I use VoiceAttack to tie voice commands to keyboard macros on my desktop, and it does what it's supposed to most of the time (but I bet it wouldn't if I had an accent.) Voice control definitely isn't an unsolvable problem, but it sure seems like one whenever I try to use it on my phone.
Thanks for that! Never heard of the program; I'm going to try it at work tomorrow :).
For longer text, voice dictation also doesn't fix the problem. Sometimes it works but often you have to seek to arbitrary points in your sentence/paragraph and fix errors. I preferred a physical keyboard.
I also do everything I can to avoid the need for a lot of text input on mobile and save it for when I have a real keyboard. I agree with other users that the decline of physical keyboards on mobiles has had a terrible impact on their real-world productive use, but speech-to-text helps alleviate that a little bit at least, if you don't mind homonyms getting mixed in there occasionally.
I want to just pull my phone out of my pocket, hold a button (or ideally hold a button on a smartwatch) and speak a command. Using the button release to finalise would also speed things up, because the phone doesn't have to wait for silence to know you're finished. It would also work better as a keyboard replacement. Select text box, hold button and speak to type.
It just seems so obvious to me that I don't know why no one's done it yet.
If I don't use regular Google search due to privacy and tracking concerns, why would I use Google Now?
While I like not having a physical keyboard, aside from text messages, I loathe typing anything of length on a smartphone.
I also find the UX of the iPhone's autocorrect to be really backward and frustrating; when I type a word, and I get a suggestion, I shouldn't have to lift my fingers from the keyboard to kill the suggestion. If I keep typing -- then keep what I am typing.
And yet, it's the most prominent feature in view whenever you see a "glamour shot" of a new Android device. It's a Google OS, so of course it's got to have a prominent search bar! But I never once used it—I suppose reinforcing the findings of this article. The ugliness (both visually and metaphorically) of a big Google search bar on the home screen of my phone is one reason I dropped Android.
I understand that you no longer use an Android device but hopefully this is helpful to anyone else who happens to read the comments and has similar wants. There are other launchers but I haven't kept up with them in a while since finding one that works best for me.
Seriously though, how is this:
http://i.imgur.com/kazES2a.png
supposed to be better than this:
http://i.imgur.com/RIKz2qe.png
(in fairness, I had to throw a bunch of icons onto the Google Launcher desktop because it's not my default and I wanted to show how space is used. Also I put neutral-ish wallpapers on so as not to distract but they're not my usual ones.)
This is how it looks: http://imgur.com/a/ce72A.
Those are fresh shots I just did, no editing or setting things up. That's how it looks on my phone every day.
There are some display bugs, particularly with the system flashlight widget, but I got used to them and I can't live with anything that has bigger / less dense icons. I also find a simple black background to look cooler than a wallpaper (having a screen that displays black well helps).
I was a die-hard iPhone supporter since I got my 3G, but switched to Android when I started at Google back in June. One of the things I like best about Android is the search bar on the home screen - I use it constantly. Half the time what I'm looking for isn't a web page per se (directions, apps, email, etc), and having search integrated into the home screen is extremely convenient.
The one thing I really do miss is being able to hold down the home button to activate Siri. Saying "OK, Google" is only successful for me about 50% of the time (and I have a very standard American accent) - when it doesn't work I feel like Scotty in Star Trek IV: The One with the Whales: https://www.youtube.com/watch?v=LkqiDu1BQXY
For no discernable reason, I have never gotten into using voice commands for everything.
(I'm grateful my natural inclination has proved out to be a good use of time, but now I don't even feel guilty about it, which is nice.)
If we were robots or Vulcans, perhaps socialization is a net loss, but for humans in general it is not. If you are trying to optimize your productivity, removing social contact is not necessarily moving in the right direction; the productivity optimum is not "zero social contact".
Of course "social networks" may be low-grade or negative socialization. Perhaps the fact it is "low-calorie" social interaction causes people to spend too much time on it when they could do better by other mechanisms. I don't know, really. I'm just pointing out that assuming social == not producing value is an oversimplification, if you're a human being.
Phone calls, SMS messages, IRC chats or even emails are way more healthy than whatever happens on social networks because they are bidirectional. And I'm not even speaking about the magic that happens when shaking somebody's hand or when receiving a smile.
So it goes both ways.
- broadcast but pull (not push) communication; everyone (from the group that can see a particular post) can join or leave a discussion at any time, and reading it does not require commitment to active participation
- asynchronous communication; I can drop a comment and read replies at my own schedule, and there is no expectation for immediate response
And no, you can't really disconnect it from the other side of social networking, i.e. posting cat pictures and news articles - they provide a social object, around which a discussion can form[0].
Complaining about social networks as "inferior" form of social interaction seems like complaining about potatoes as inferior food because they weren't around in Europe 500 years ago. Yes, living only on potatoes isn't the healthiest thing, but they are a decent food and even if you make it the core of your diet, you won't get sick.
[0] - http://gapingvoid.com/blog/2007/12/31/social-objects-for-beg... curiously, this text now starts with an image that wasn't there last time I read it, and that perfectly sums up my opinion on advertising
Just as with fast food that depends on subsidized corn, ignorance and fossil fuels, social networks like Facebook represents the industrialization of our social interactions: fast, cheap, shallow and designed for mass consumption. And make no mistake, just like how the food demand is inelastic because our stomach doesn't grow bigger, so is our attention span as there are only 24 hours in a day, so the same strategies are employed by both sides, with Facebook being the new TV (how ironic, almost as ironic as organic food being industrialized).
And going back to the potato, we Europeans may have started eating it only 500 years ago, but the potato has been domesticated 7,000 to 10,000 years ago in South America. We simply discovered what other people have been eating for thousands of years. No such thing happened with social networks. Or with present day processed foods for that matter.
And you know, industrialized food and the modern lifestyle has been making us obese and gave us diabetes. We might also discover that this modern lifestyle is also bringing with it other gifts, like autism and I wouldn't be surprised at all if some conditions from the autistic spectrum are also caused by humans having shallower social interactions.
Are you sure you are not just experiencing confirmation bias?
A lot of very high-octane code has been written by devs who eschew social contact (any distractions, really) for the duration of the push, so I suppose it just depends.
I have a hard time believing that any "study" can really tease out optimum productivity for all cases.
Take running. Trying to get from point A to B as fast as possible requires different strategies depending upon how far apart they are. A very short distance is best done by going all out 100%, but if you do that for a marathon you'll be beaten by someone who paces themself.
"I did a thing" -> "Study comes out which says the thing is a good thing" does not conform to any useful definition of "confirmation bias" I can think of. (I can stretch it, but the result is not a useful definition.)
I can't speak for everyone in the group, but I feel a little recharged after a flurry of funny gifs or esoteric programming jokes.
(See my comment here: https://news.ycombinator.com/item?id=10419034 for my reasoning)
https://en.wiktionary.org/wiki/Grauniad
Communicating, managing projects, taking notes, reviewing/signing documents, sending payments, drafting designs, outlining documents, finding/ordering kit -- are these not productive?
Yeah, you're probably not doing hardcore document creation, or real research, away from a keyboard, comfortable chair, and a nice big screen, but that's not the shape of all productivity.
I think it's active vs. passive. You search to find something vs just consuming whatever Zuckerberg thinks you want to know about.
If those same people had a 27" screen during that same slice of time and space, they'd probably be doing the same thing.
How is that not search?
When you're a journalist ;-)
I think it's easy in theory (and probably hard in practice) to solve this problem. Google just needs to reduce the friction of web searches. I think they've gone a long way to achieving this goal on Android 6.0. For the uninitiated, they added that weird hold to search which automatically picks stuff out from the screen to search and often presents results in a way where it's really easy to switch back to what you were originally looking at.
Imagine if all you had to do was think about something to initiate a google search. Okay, that would probably get pretty dystopian but it would let you do many more searches even while your screen has been abducted by an app.
I think all google needs to do is reduce the frictions to search.
Literally all it takes to search is one click and you can start typing or talking.
I've got a motorola so I can also just start talking with my keyphrase and it will search...can't really get any easier.
The problems they have is usually the small screen that makes it hard to view much content at once and that text input is really burdensome. Some tasks require these and are thus not suitable for the mobile devices. They however also have advantages to be usable in many situations computers aren't (for using a computer while standing you need a special desk while you can use the mobile devices in almost every position and place) and touch input may be better for some tasks.
For stuff that would be possible both for computers and for mobile devices equally good there's the barrier that it's already been developed for computers and that everybody needing it for work already uses a computer, so you would need to put a unbearable mass of work into it to bring it to already is with computers today.
On mobile I feel "caged in" psychologically, but that may be influenced by having had the desktop experience as my first and also my primary experience in search.
What could help, is if there was an easy and smooth way to pursure browsing started on the mobile on the desktop, and vice versa to continue reading a document started on the desktop on the mobile.
One very good thing that Youtube does, is to record in the history the point at which we're in a video, so that if you browse to the history on another browser on another computer, it will continue the video from the point you left it. This allows you to switch between desktop and mobile and vice-versa. Cats can be watched on mobile, but if browse to a conference with slides and you want to take notes, you definitely want a bigger screen and a desktop computer.
It's similar for the browsing (including the bookmarks, and the tabs). If they were shared smoothly, one switch more easily between the two devices.
Apple has that; they call it "Handoff". It works reasonably well.
https://support.apple.com/en-us/HT204681
Although, I use exactly your desktop pattern on my phone. Go anywhere there is a search bar (they're all over on Android), and search "wikipedia foobar" and go straight to the first wikipedia hit. Works better than wikipedia's own search usually.
And their mobile site is actually pretty nice. I like it at least. I'll check out the app though.
One interface feature I like with the app is within-article navigation: if you swipe inwards from the right edge, you pull out a table of contents, and can easily jump to article sections by tapping one. This is useful enough that I find myself unintentionally trying to do it in mobile Firefox, hoping I can jump around a blog post or documentation's headings that way (I wonder if mobile FF could implement this? do enough pages use HTML headings in a semantically meaningful way?).
This is more a stylesheet choice, but the app also more aggressively collapses things other than the main article text, like infoboxes and tables. On the mobile web view I find I have to scroll down too much to get past big infoboxes and such.
On slower phones and slower data connections it's also speedier. Especially on longer pages and pages with more non-text stuff, my phone has a noticeable pause to render the page after it loads, while in the app it's nearly instant. The data transfer via the app is also low enough that it's usable even on slow connections (I'm semi-frequently on T-mobile's free global roaming, which is rate-limited to 2g speeds, ~128 kbps). Since the interface is already preloaded, all it needs to do to load an article is grab the article text through the MediaWiki API, which is usually in the range of 2-20kb. It then lazy-loads images later (which you can disable if you really want to be lean on data).
1. 50% of people do one or more searches on mobile. That's awesome from Google's POV!
2. The number of people who are using mobile for their Internet has exploded in only a few years, many of those people still use their desktop for search.
3. Many, many people still don't use their phone for anything other than texting and phone calls.
But here is the kicker:
4. Charles Arthur is extrapolating a measure of central tendency in a rather foolish way. He's taking the mean average of the number of searches from mobiles per day for an entire month, then he's extrapolating that none of those mobile users do any searches per day for an entire month.
In other words, he's saying that half of all users don't do ANY searches at all, whereas all that he can say is that on some days, some users don't do any Google searches on their mobile device.
And if that's all his statistics are showing, then I'm not exactly blown away with this insight.
This is me, well except the texting part. I only have a smart phone for the occasional time I need navigation. The vast majority of days I use it to make one phone call.
That navigation part makes the cost to upgrade every few years worth it though.
From the article: "who does what searches isn’t fixed; so someone who did zero searches yesterday might do 10 tomorrow. But equally, the 10-searcher yesterday does none or one or four today. And so on."
So yes. But on any given day, half of people don't do a mobile search. And the numbers from Google suggest that the average across a 30-day period is less than 30. If you think there should be a different model for how many people do how many searches than a Pareto distribution, please offer it.
How do you know the spread of data isn't that there are 20% of users who don't use their mobile device for the Internet for 10 days in a row, but then do a lot of searches in a short space of time?
Maybe I'm missing something here, but by taking an average over 30 days, you seem to assume that users do similar numbers of searches every day.
Fewer navigational searches would increase the proportion of monetisable searches; wouldn't that be a good thing for Google?
In which case, do these lost searches actually reduce revenue in terms of valid as click throughs? I would argue that most people searching for Facebook for example are simply going to click on the first Facebook link the results.
If Facebook really are going to advertise (paid) in Google for the search term Facebook, then obviously Google miss out on a relatively pointless ad, but in most cases, I can't see the loss here. Why would Facebook advertise for the term "Facebook"?
Search the name of virtually any large company, and their website will come up as both the first real result and the banner ad at the top of the results.
"More people perform half a search every day, than any other amount." -- What?!
I guess maybe that's more like "the mode over 30 days, divided by 30"?
That you would publish him... Well, that's your judgement but I consider it to be lacking.
Thus neatly showing how comment threads turn "pretty toxic". If you have better analysis of the data, that would be great to see.
And it is. According to his numbers, the mode is precisely 0. The most common number of searches is zero... more people do 0 Google searches in a day than do 7 or 1 or any other specific number.
This is not a USEFUL statistic, but (according to this data) it is accurate.
In order to get the mode, you'd need to know how many searches each person does each day. You can't get the mode from the figures he's been provided.
But in the article, he wrote:
"I tried modelling what search activity probably looks like on mobile: I used a mean = 0.925 (as per Singhal) and mode = 0.5. The mode must be below the mean because of the long tail of higher values; 0.5 is a guess, but moving it around doesn’t have a large effect. This gives a median of 0.94, close to the mean, which you’d also expect."
He made some guesstimates which he used as input parameters to form a Pareto Distribution.
As for the mode always being less than the mean, that's wrong too.
I was referring to the fact that the graphs - e.g. https://theoverspill.files.wordpress.com/2015/10/screenshot-... - are using using mode = 0.5
My career has spent a fair bit of time working on search engines... google's moat is that everyone uses it, it has the audience. Providing competent search is something a company the size of Apple could replicate with some effort.
I thin the threat to google is real... I think google knew this a decade ago, and that's why they bought Android Inc. and developed the android OS.
I used to joke that smart phones are the equivalent of Penny's "computer book" from the Inspector Gadget cartoons I watched as a kid. I remember being in grade school watching that show and thinking how awesome it would be to have a powerful computer I could carry around with me all the time and use to find any info I wanted.
I've gotten a job because an ad for a job search website that looked interesting, I've found out about new products and (after some of my own searching) purchased them, and i've definitely gone back to products i've left because of newly advertised features. (Microsoft Azure being one of them)
I don't remember if these were specifically on a search page (chances are they weren't) but they were ads on the web.
The goal of many ads isn't to make the consumer say "wow I need to buy that now!". It's so when the consumer is presented with a handful of choices, they say "yeah, I know this brand." And is certainly is measurable, in brand awareness studies.
The hardest part of relying on brand awareness in ads is you have to repeat the ads over and over and over, and hope people start to recognize the name before they get sick of seeing the ad.
I expect a flamewar about whether or not people actually click on ads. I'll be happy to participate, because I feel the need to vent out some of my issues with the advertising industry ;).
Oh, and to answer your question - personally, no, not directly through an ad.
If I were an investor, I might be worried about some aspects of Google, but this isn't one of them.
I'm probably biased, though. I am constantly "ok googling" to learn stuff. Having young children will do that to you.
I have a 1 year old, low/mid range phone I bought because I just didn't have the cash for a flagship. I have only the "most common" apps installed. Mobile web is unusable. Opening chrome takes 2 minutes, doing the actual search a minute more, rendering the site ... well, let's just say I have JS turned off. Unless I don't have access to the internet some other way and it's really crucial, I'm just not going to bother.
Is $200 unlocked not what you mean by low end?
1: http://www.walmart.com/browse/cell-phones/shop-all-no-contra...
$200 to someone living in poverty is more than a month of food, while $20 up front + $40/ month for a baseline limited text and call plan seems far more manageable.
It seems that there are lots of ~$100 (or less) phones shipping though, and they are quite a bit less capable than the one I bought.
You get reasonable (unlocked!) Android phones for as little as $100, and entry-level ones for a lot less. Low end is really, really low nowadays.
Edit: The cheapest I could find: http://www.alcatelonetouch.com/global-en/products/detail?mod... Sells for less than 50€ unlocked.
So much amateurish bloat. I'm baffled by those engineers.
You are not the customer they want.
That's true. Then I'd say I just don't like the engineering they are doing.
I have a Samsung Galaxy Note 10.1 2014 edition that I bought in February, and it lasts nowhere as long as my wife's 2012 iPad3. Such a disappointment. I will sell it, last Android device I buy.
I have Samsung Galaxy Note 10.1 but older edition, battery is still fine, the OS got slow, even on factory reset, so, I decided to root it and runs much faster - pen use was not that great before, now it's perfect.
Note that the act of rooting didn't have anything to do with the speed up, after installing a custom OS with GPU optimizations for drawing graphical elements it worked great.
You can always install a very minimal version of the GApps later if you really need them.
Look for a Nano or Pico version. I can't remember which is the smaller of the two, but it basically just comes with Play Store and the Framework to make it work.
Especially in mobile. Sure, your site loads fast when your in a major downtown area, but try using your 30 different java-scripts when your in Nowheresville Alabama
Just to be clear, I've experienced similar degradation with Apple products as well (specifically an older iPad), but never as severe as with the Note 3. My old iPad is still definitely useable; it's just not as smooth as it once was. The phone on the other hand had just become complete junk.
Meanwhile my girlfriend's Nexus 5 phone has been on a roller coaster of battery life changes. Not too long ago, an update caused her battery life to drain to 0 in roughly half a day of light usage. Then it was fixed. Then it was ruined again in the update to Marshmallow. She's going to have to get a new phone as well because hers too has become nearly unusable.
I've used Cyanogenmod on my smart devices for as long as I've been buying them. CM has had sandboxing for apps for a few years now, and I was running 5.1.1 on my S3 months before it was available for my SO's S5 through her carrier. CM13, announced recently, will be built around Marshmallow.
For older devices, you can either try the newest CM versions or go back to earlier stable major releases. And they're all bloatware and CarrierIQ free.
Fairly easy to do if you buy an opener kit for a few dollars.
It's a hardware problem.
Completely unintuitive, I love Apple's design team, but it makes me aware that Apple designers are mortal after all :-)
Under the "Find on this page" heading.
For tagged bookmarking I use jotmuch: https://github.com/davidlazar/jotmuch
For more context, I almost always use my phone with this emergency pocket keyboard: http://www.amazon.com/ZAGG-FOLZKFLEXSLV-Zagg-ZAGGkeys-FLEX/d... .. so I somewhat normal, high input rate.
Although, I do search on my phone while at my desktop in exactly one scenario: playing games. It's less hassle to do a quick search for where to find item X in some random game than to pause, alt-tab out, fire up a browser, search, try to remember results, close browser, and alt-tab back to the game. Otherwise, yeah, full browser when I'm in front of one, phone for the rest of my life not spent in front of a computer.
> guessing that 50% of those corporate machines, ie 450m, can’t make Google searches
Otherwise, you're really just fumbling around in the dark, saying that monetizing mobile searches with such a low average daily click rate are a big problem for Google when the reality could be that they've never made much money from mobile search ads, especially in developing countries where they are adding the most users whose ad clicks are worth less because they have less disposable income than the average American or European. Similarly I'd wager that mobile search ads are a very small portion of Google's overall mobile business, and that Google never planned to make much money from mobile search in the first place, understanding that building out the app ecosystem and getting Android onto a ton of phones was more important in the long run.
People don't explore on their smartphones: they consume.
This as a cool thought experiment that could have been improved with more insight into breaking "ad revenue" into meaningful chunks to see whether or not Google actually cares about the slice considered in the headline: mobile search ad revenue.
The author (me) would love to, but Google makes very little data available. The only geographic data is total revenues broken down by US/UK/Rest of World. There's also revenue from "Google [search]", "platforms" and "other". Estimates about YouTube etc revenue tend to be calculations based on various private metrics.
"I'd wager… that Google never planned to make much money from mobile search in the first place"
Possible. As the article says, and has been repeated multiple times, Android was a defensive move to stop Microsoft cornering the mobile search field.
"People don't explore on their smartphones: they consume."
I must introduce you to some networks called "Instagram" and "Snapchat" and "Line" and "Weibo" and "WhatsApp".
They might be reading the news, which is roughly 50%-99% PR placements depending on the publication, which pushes them down some sales funnel or another. They're just browsing higher up in the conversion funnel. Searches higher up in the funnel tend to be less valuable in terms of costs per click than searches which are closer to the purchase.
Everything in the news is marketing anyway, and it's just as paid for as it is when it's an ad -- often times it's more expensive, because access to the news is rationed by publicists and PR firms rather than by the advertising department. Further, the bulk of social media is the effluvia of PR as it reverberates through people who talk too much on the internet. As far as Google is concerned, that propaganda consumption is just likely to stimulate more searches later on which can be advertised against.
The flow might be: Check Facebook App -> Click Gossip Rag Article -> See fabulous new shoes in celebrity photo -> Order lunch on Seamless App -> Go home from work -> Remember to Google "buy red high heels kim kardashian" -> Click ad -> Price check, bounce (Google gets paid) -> "red high heels discount" -> Click ad (Google gets paid again) -> Buy shoes.
When they are ready to buy, they will pay for the $10 click -- maybe multiple times across different advertisers -- on their desktop or $15 app download on their mobile phone.
The headline that this author uses is also really loosey goosey. Google's ad products on mobile are also increasingly more biased towards generating paid phone calls. When you search for a locksmith or auto repair, you want to get a guy on the phone right now, and are likely to need to buy immediately.
So, this isn't really Google's "growing problem" -- Google as an organization is set up to use machine learning to figure out how to make money off of mobile search behavior. The company is still learning what it has to do to make money off of mobile search behavior and how it differs from behavior on desktops.
Also, as far as Google's customers (advertisers) are concerned, if a desktop search results in people subscribing to an e-mail list that they read on their phone, the advertiser earned measurable results from the money that they gave Google. If their customers don't like to buy on mobile or they behave differently, campaigns can be configured accordingly.
Those changed a little with a Note 4. It's a much more powerful phone, and I do still search on it, but touch input is a modest PITA. I cut back on searches, but I have added some voice use cases too.
In my experience, people with powerful phones use them more. A lot of phones out there aren't so powerful, and they keep it to social, MMS, etc...
I actually disallow FaceBook on my phone. It's a battery hog, and I've better things to do. FB once a week on a laptop is enough for me.
Look up a movie? 'imdb spiderman'
Want to know how many calories is in something? 'cinnabon calories'
What's the football score? 'piggers score'
If that's the case, then it seems pretty shitty from the advertiser's point of view. Especially since Google came up with the all-in-one search and url bar "feature" in Chrome.
The latest available number is from 2012: http://searchengineland.com/google-1-trillion-searches-per-y...
That's 3 years ago.