In the world of high interest rates, companies like Snap that have not figured out a business model in the 10 yeas of their existence are not likely to survive. Not surprising. Unfortunately there will be more of these from Snap and the likes of it.
Over the longer term that is true, but the world had gotten used to well over a decade of ZIRP or near-ZIRP after the GFC. It's an assumption that is now deeply ingrained in a lot of business models.
> It's worth pointing out that, historically speaking [0], we're not an in era of particularly high interest rates yet, just not absurdly low ones.
The question is what is the right level of interest rate in 2023 as compared to say 1960.
19060s were boom time with GI bills, lots of new industries started by veterans, a booming suburban household, booming number of children. People and companies of the time had very low debt aka there was room for them to take on more debt. All of this leads to rising credit, which requires a higher interest rates to keep inflation low.
2023 is an anomalous post-pandemic boom coming from trillions added to US government debt. There is no population boom, no business boom, no new tech boom (excluding the AI stuff going on now). Nobody can take on more debt as most people/companies are completely tapped out. This actually requires lower interest rates but we have high inflation so the FED is keeping interest rates artificially high.
This is to say, that companies like SNAP just cannot continue to exist in a higher interest rate environment. Neither can companies like Meta, Uber, Google, Microsoft without cutting costs somewhere or without letting the stock collapse.
A recession is the only way out because companies will NEVER let the stock collapse in favor of saving their employees.
> A recession is the only way out because companies will NEVER let the stock collapse in favor of saving their employees.
Only as stock in this case is a useful proxy for the finances of the business. It would be better to say that, outside of ridiculously comfortable and easy financial environments, companies will prioritise investing in activities that make money (directly, such as making products, or indirectly, such as security) over activities that don't.
> It would be better to say that, outside of ridiculously comfortable and easy financial environemnts, companies will prioritise investing in activities that make money (directly, such as making products, or indirectly, such as security) over activities that don't.
And these activities are designed to increase the stock price. Everything they do is for the price of stocks. That's how the system is.
That seems a jaundiced viewpoint though. Stock prices are a proxy for the business's value going up, which businesses should in general strive for, unless they have it so monopolistic that they don't need to compete with anyone else. This is just as true in a business without stocks even being available.
I'm not claiming that companies shouldn't strive to raise value of business.
All I said was their activities are designed to boost stock prices, given the prevailing economic conditions. Stock prices may be related to business value or financial engineering or whatever. Doesn't matter how they get there. Their goal is to just keep stock prices rising.
With this in mind, if it ever comes to choosing between employees, customers, product, or anything versus the stock price, they will choose the stock price.
I don't think there's any point saying this in this way, though. Not every company will do anything it takes to hit a stock market goal, even financial engineering. Many will just work on fundamentals, and either go broke or get outcompeted or succeed.
And there are plenty of companies that aren't public companies with any stock to buy.
It just seems far too simplistic. You seem to be critiquing the times that some company use slightly dodgy financial engineering to boost their stock price. But that's not many companies, and not all the time - it's never a permanent fix. There's certainly no need to attach stock price reason to these layoffs, which are much more likely just cost-cutting measures designed to save the company a load of money it thinks is more valuable than the work of those people.
The ECB aims for "2% inflation on average". Inflation and interest rate have a close relationship, so at least for them that's the new normal. I seem to remember that the Fed and other central banks likely talked about the 2% inflation target.
Since 2008, the interest rate was much lower and nobody _quite_ cared, now both inflation and interest rate are higher and the ECB widened the window they consider for their average substantially to straighten things out even though they're way about 2% now.
Still, there has been a shift in the finance world, and the target seems to be 2%. Having interest rates and/or inflation at twice that (or more!) seems to justify the label "high" to me.
Are you saying that central bank interest rate is targeted to be 2%? If so, I don't believe that is true. The target is 2% inflation and full employment, the interest rate is adjusted to whatever level achieves those goals.
It's about inflation (see https://www.ecb.europa.eu/mopo/strategy/pricestab/html/index...), but there's a relationship because their interest rate informs the interest rates across the region they maintain monetarily (practically the baseline, where every other lender adds their margin on top), and the interest rate is the strongest tool they have at their disposal to adjust inflation rate.
So the central bank interest rate isn't targeted to be 2%, but it will be somewhere close to it: too high, and they drive inflation well above 2% all on their own. They can stay lower for quite a while (see the past ~15 years) but that was already considered an emergency situation.
We'll have to see where things go from now, but the recent shift of their language towards "2% in the _medium_ term" indicates to me that we'll stay in the higher end of the spectrum for a while instead of quickly trending down again.
> So the central bank interest rate isn't targeted to be 2%, but it will be somewhere close to it: too high, and they drive inflation well above 2% all on their own.
Isn't it the reverse relationship? Increased rate decrease the velocity of money and decrease inflation?
(At the risk of causing a recession)
Central banks have been raising interests to fight inflation so far.
As long as we’re pointing things out, we should also point out that unsustainable companies have not quite yet collapsed due to interest rates. We’re only seeing some pain, but once we due reach a proper high interest rate era these companies will be gone.
It doesn’t. But if it did, I would still say the collapse wasn’t due to interest rate hikes but rather from increasing regulation banning these scooters from major cities and from a flood of cheap competitors.
Why wouldn't SPACs count? What allowed those financial scams, spacs, crypto, zombie companies, was this low interest rates environment, for over a decade. Of course SPACs count, they are a consequence of LIR.
I think SPACs are a bit different. They have been around since the 1990s, and for a long time there were maybe a dozen SPACs in a year. In 2021 there were about 600.
I think they will stick around and go back to their previous volume. Grift-o-currency is a scam and I hope it dies soon.
This is a slick visualization and agree with your point. I feel ZIRP for such an extended time period was nuts and it was responsible for the asset bubble, let weak players survive and supported more risk taking.
As I was looking at the chart, I noticed there was a big bump from 1994 to 1995 (like from 3% to 6%). What happened then? Was it inflation? I recall rates were about 6ish since when I was in undergrad. I guess Greenspan/Y2K/dot com days.
When I finished undergrad, the most desirable tech company to work for was Microsoft. They paid 55-60K (starting comp) I recall? Interviews were nothing like today (with the insanity of leetcode). What is mind boggling is that I'd have a better quality of life in those days (houses were super cheap) and work was not as competitive/backstabby as today.
The million dollar compensations (when counting stock growth) and intensity/stress around interviews and promotions is what ZIRP wrought for the elite in tech.
Would be interested if people who were mid-career in the mid 90s can comment on their perspective.
> Interviews were nothing like today (with the insanity of leetcode).
Would that mean it was the era of "Why are manhole covers round?"
I don't want to sound like someone who supports leetcode (mostly because I don't), but it seems like it's at least an attempt at measuring something related to programming skills.
Eh...interviews were a mix of brainteasers like the manhole question and coding questions. IIRC some of my MSFT interviews back when allowed pseudocode but mostly wanted to fill the role of e.g. FizzBuzz. Then IIRC there was some validation of C or C++ competency.
Leetcode is FizzBuzz on steroids, but IMHO is not on its own likely to be more predictive of success at a company than the old MSFT way.
Being able to reason about possible design decisions and constraints of things you encounter sounds much more related to programming jobs than being able to code a red-black tree from the top of your head.
Looking back, the highest quality of life I ever enjoyed was when I was working my first software development job out of university in the 90s. $45K/year comp and I bought a condo for $45K. Imagine: being able to buy an actual house for a mere 1X your yearly salary. Unheard of now. It's been all downhill from there in terms of the cost-of-living:compensation ratio. And that's for software development, one of the more financially lucrative careers out there. It's been even steeper downhill for non-techies.
Was there 96-2000 and agree completely. It was a paradisical time and I knew it, because this was nowhere near my first job. Best work situation I ever had. People were great, and bureaucracy was big enough to be useful but small enough to be intrusive.
> This is a slick visualization and agree with your point. I feel ZIRP for such an extended time period was nuts and it was responsible for the asset bubble, let weak players survive and supported more risk taking.
I think it might also end up being important to remain competitive in a global market where the other economic superpower's government is willing to invest trillions into building whatever it wants and needs despite profitability[1].
Some things we want/need won't be immediately profitable, or even profitable in the long run. They might not even be things we realize we need until something unprofitable is researched and developed.
Not saying that Snap is something we need, but if the US is forced to strictly rely on market forces to compete, "free" capital via low/no interest rates is a way to kind of do it.
How much of ZIRP-inflated capital went into chasing fanciful ventures like UberEats, Wework, Doordash, Airbnb, crypto and Twitter? State-run capitalism channeling funds into strategically important areas feels like a breath of fresh air compared to handing it to venture capitalists.
I agree, hence my comment about the US choosing to rely strictly on markets to allocate capital. If the only lever the country is willing to pull is to adjust interest rates, it's sort of a means to that end, however inefficient.
There is a nice Youtube channel called asionometry that shows that this was done time and time again (maybe more in Asia and Europe) in things related to electronics and semicon manufacturer, and it failed most of the time. TSMC was a shocking exception though I'm sure there are others.
In the old days, these investments were tiny compared to what is being currently spent on the chips act. I think for STEM a better approach would be a govt sponsored entrepreneurship thing .. imagine the same YC deal but we give this to any PhDs (or maybe even Masters) once in their career. I'm from Canada and when I see the cash wasted on other initiatives, I can't help but wonder why this is done. Not to say Masters or PhDs are geniuses. It is just the current system means only the well-to-do or well-pedigreed can get into places like YC (and maybe elite schools like Berkeley) while intelligence, grit and ambition are more widely distributed.
But how many businesses that might have been profitable just fine and that will be wanted/needed at one point have disappeared or just never happened because their market was killed by one of those post-profitability unicorns high on ZIRP-fueled market cap? Up to what point is it still market capitalism and where does it become barely more than an elaborate hoax benefiting those who lack neither collateral nor the will to leverage?
Technically, they were trying to innovate a new business and failed. Your comment is effectively blaming them for trying to "figure out a business model", which I find odd. Damned if You Do and Damned if You Don't...
How is a social network supported by ads a new business model by the time Snap came along?
By that point you already had:
1. Facebook, obviously
2. Google+
3. Twitter
4. Tumblr
5. Foursquare/Swarm
6. MySpace
7. Friendster
8. Orkut etc
The only real innovation they brought was the ability to parlay teens sending each other legally dubious dick pics by deleting them almost immediately into venture funding and then an IPO.
Snap also innovated by figuring out how to drive obsessive-compulsive behavior among teens with snapstreaks. And even getting them to pay to restore snapstreaks if they missed a day. This is what the brightest minds of our generation are working on.
IMO, the overlooked big innovation of Snap, and the thing that got them their initial popularity, was their absolutely arcane interface. They've done a lot over the years to make the app easier to use, but in its early days, there was an almost toy-like quality to trying to figure out the various features and functionality. That acted as a selling point to the teenage demographic: not only did friends showing each other how to access some practically secret part of the app act as viral advertising, it also kept your parents off the app.
It also acted as an overt anti-Facebook, in an era where that was a growing trend: No, and then limited, broadcast communication versus FB's "tell everyone everything" defaults, ephemeral messages in stark contrast to FB's "hey, remember this inane thing you said 3 years ago", and a culture of communicating asynchronously with a face attached.
This (perceived) lack of a panopticon made Snapchat a much more appealing form of social media. IMO, it's still one of the better ones.
They have lots of revenue. They would need to pare back their operating costs quite a bit to break even, and who knows if they can do that without overly impacting revenues.
If Evan sold Snap to Facebook when Mark wanted to buy it, Snap would have been turned into what is Instagram today. Mark & Facebook really only have one trick when buying companies - turn them into a newsfeed and make everything public and push celebrities and public figures to use the platform.
Instagram and Snapchat today are two fundamentally different products, and Snapchat is estimated to capture more of the younger demographics in more markets.
Sure, you can say that Instagram today has about 100 billion valuation compared to 15 billion for Snapchat, but the race is far from over. Snap recently exceeded a 100 billion valuation as recently as 2021.
when you need to re finance the loans, the new interest kick, also high interest means lower valuation so is higher interest + higher insurance premiums (% debt/valuation)
Acquired is doing some incredible work there to cover up one billionare who ran his mouth in a legally actionable fashion and got sucked into a business deal that literally everyone on earth, himself included, thought was a bad one.
Also X is now worth 4 billion so apparently it's at the very least, not a growth industry.
The challenges at Snap seem to have nothing to do with interest rates and more to do with a challenging ad market, which has affected pretty much all the major players, including Meta & Google.
I don't think it's so much that Snap hasn't figured out a business model - they make a billion dollars in revenue per quarter, how many companies can claim that? - it's that Snap hasn't developed a rational balance sheet due to excessive SBC. If it wasn't for SBC, they would be making money hand over fist.
Snap & PINS are two interesting case studies in how companies in a similarly challenging ad market can have potentially different strategies and outcomes. Both have very depressed stock prices, but both have a clear path to having a sustainable and highly profitable business.
Again, interest rates seem pretty irrelevant to the fate of both companies, only as much as low interest rates encourage economic investment and can fuel the digital ad market.
Article had another interesting tidbit: "advent of generative AI has made it easier for companies of all sizes to create try-on experiences for their customers and made it harder for us to differentiate our offering.”"
Some other large companies (can't mention names) have also done layoffs targetting genAI recently. I was surprised by this.
I'm not sure this is quite relevant here, although I do love the link.
Meta's open source play I think comes back to regaining public sentiment about them after scandals/metaverse, and also making it difficult for established competitors to have much moat.
Given how much Snap pays their employees it was probably the right call.
Meta released a lot of high quality AI research and models (Llama) and tools (PyTorch) as open source. So instead of paying OpenAI for access to their large language model, you can just download a LLM from Meta and use it for free. So in effect, they are flooding the market with free alternatives to paid Microsoft/Google/OpenAI offerings.
> Well, which is it? Is it so trivial that anyone can compete, or is it too costly to build it?
I think that's the wrong "either/or" framing in this case. It's not that building competing tech would be trivially cheap, but it's something around which Snap would not likely have a moat or any sort of competitive advantage. It's also in a business area that's pretty unrelated to their core offering.
> They tried something, and it failed, and they don't want to say that.
The CEO's exact quote was "Leading in augmented reality means that sometimes we will fail, and I am proud that our team dared to build this business even if we did not succeed." So I'm not sure why you think "they don't want to say that."
> And they didn't know these things six months ago?
Given that the initiative was originally announced in March, I'm assuming the whole project was greenlit at least 3-6 months prior to that. And ChatGPT was only just released on Nov 30, 2022, and since then it hardly feels like a week goes by where I'm not mindblown by some new AI announcement. So yeah, I think it's entirely reasonable to think that the landscape when this ARES project was first started is very different than the one we find ourselves in today.
It's not really just startups - honestly I think 99% of those startups will fail.
The problem with generative AI as a business is basically what that leaked Google memo said, which was something along the lines of: "OpenAI doesn't really have a moat, but neither do we."
There are a few large companies who are creating these successful models (OpenAI, Microsoft, Google, Meta, perhaps Amazon with their Anthropic investment), but there are tons of other startups that are trying to come up with "individual products around AI", and I think it will be incredibly hard for them to differentiate. That is, for a lot of the "dedicated" early-stage products I've seen, I've thought "I can pretty much do this all with ChatGPT, maybe ChatGPT with plugins - why would I want to download your special app?"
I don't think all of these companies are doomed to fail (for example, I think Harvey AI was smart to focus on a very specific, profitable niche: lawyers have a lot of money to spend on this stuff, and they have specific needs e.g. around compliance, auditability, privacy, etc. that you could see a specific legal-focused AI tool addressing), but I think a lot of them are.
> there are tons of other startups that are trying to come up with "individual products around AI", and I think it will be incredibly hard for them to differentiate.
Even more fundamentally, they're putting the cart before the horse. They're making the classic mistake: Taking a tool/technology, and then searching for a problem to solve with it, rather than starting with the problem. I agree a lot of these companies are doomed to fail for the same reason companies founded to "Figure out what to do with blockchain" were doomed to fail. They'll no doubt suck down plenty of funding while failing though.
There’s a lot to unpack, this is from my perspective doing work in advertising innovations for these customers on social media.
Snap’s biggest problem is it’s a crummy place to work with too many… let’s say personalities that are not commensurate to what they deliver. Same could be said of Amazon, though they have a far higher revenue ad product than Snap ever had, so shows how much those two are related.
To this specific quote: Generative AI meaning image generators are anticipated to replace creative agencies for display ads.
They are not good ads right now. Mondelez and Unilever who are trying this for example have not found a big breakthrough. Part of this is that neither of the two latent diffusion models create sufficiently creative creatives. You need programmers with art backgrounds to master these and the brands have neither of these. They’re super rare generally, you cannot just pull them out of the professional recruiting pool, and they most definitely do not want to work for brands or Snap.
However the appeal is in the race to the bottom costs. Hard to compete with “free.”
These AR experiences still require opinionated creative development which Snap, its agencies and the brands lack. So that’s really what ARES in particular failed.
The product people want is “Ads people like.” That’s expensive to make, which is the opposite of the incentives for creative production in the ecosystem.
That said I have found success creating and supporting interactive, instant streaming games as branded destinations for Meta and TikTok. Try the one I made for Hyundai: https://appmana.com/watch/virtualtestdrive - tap drive and see for yourself. This isn’t meant to be self promotion so much to prove that I dogfood my opinions into a proven, factual reality. You can see more at AppMana.com for huge brands like Nike, VISA, etc.
The AR experiences people make are too “2D” IMO; also they are not that fun. They might hit 15s of engagement on average, and most of that is spent loading, so it’s kind of a fake number compared to that virtual test drive which has instant loading and average engagement measure in minutes at million-visitor scales. Video ads have a median zero watch time and mean of 2.1s so there’s lot of opportunities for “marginally” better. Ultimately it comes down to costs.
87 comments
[ 2.9 ms ] story [ 394 ms ] threadIt's worth pointing out that, historically speaking [0], we're not an in era of particularly high interest rates yet, just not absurdly low ones.
https://fred.stlouisfed.org/series/FEDFUNDS
The question is what is the right level of interest rate in 2023 as compared to say 1960.
19060s were boom time with GI bills, lots of new industries started by veterans, a booming suburban household, booming number of children. People and companies of the time had very low debt aka there was room for them to take on more debt. All of this leads to rising credit, which requires a higher interest rates to keep inflation low.
2023 is an anomalous post-pandemic boom coming from trillions added to US government debt. There is no population boom, no business boom, no new tech boom (excluding the AI stuff going on now). Nobody can take on more debt as most people/companies are completely tapped out. This actually requires lower interest rates but we have high inflation so the FED is keeping interest rates artificially high.
This is to say, that companies like SNAP just cannot continue to exist in a higher interest rate environment. Neither can companies like Meta, Uber, Google, Microsoft without cutting costs somewhere or without letting the stock collapse.
A recession is the only way out because companies will NEVER let the stock collapse in favor of saving their employees.
Only as stock in this case is a useful proxy for the finances of the business. It would be better to say that, outside of ridiculously comfortable and easy financial environments, companies will prioritise investing in activities that make money (directly, such as making products, or indirectly, such as security) over activities that don't.
And these activities are designed to increase the stock price. Everything they do is for the price of stocks. That's how the system is.
All I said was their activities are designed to boost stock prices, given the prevailing economic conditions. Stock prices may be related to business value or financial engineering or whatever. Doesn't matter how they get there. Their goal is to just keep stock prices rising.
With this in mind, if it ever comes to choosing between employees, customers, product, or anything versus the stock price, they will choose the stock price.
And there are plenty of companies that aren't public companies with any stock to buy.
It just seems far too simplistic. You seem to be critiquing the times that some company use slightly dodgy financial engineering to boost their stock price. But that's not many companies, and not all the time - it's never a permanent fix. There's certainly no need to attach stock price reason to these layoffs, which are much more likely just cost-cutting measures designed to save the company a load of money it thinks is more valuable than the work of those people.
Since 2008, the interest rate was much lower and nobody _quite_ cared, now both inflation and interest rate are higher and the ECB widened the window they consider for their average substantially to straighten things out even though they're way about 2% now.
Still, there has been a shift in the finance world, and the target seems to be 2%. Having interest rates and/or inflation at twice that (or more!) seems to justify the label "high" to me.
So the central bank interest rate isn't targeted to be 2%, but it will be somewhere close to it: too high, and they drive inflation well above 2% all on their own. They can stay lower for quite a while (see the past ~15 years) but that was already considered an emergency situation.
Before 2008, the ECB moved between 1.5% and 3.75% (https://www.ecb.europa.eu/stats/policy_and_exchange_rates/ke...) with at most 16 months at a time above 3% which was followed by a bump down to 1%.
We'll have to see where things go from now, but the recent shift of their language towards "2% in the _medium_ term" indicates to me that we'll stay in the higher end of the spectrum for a while instead of quickly trending down again.
Isn't it the reverse relationship? Increased rate decrease the velocity of money and decrease inflation?
(At the risk of causing a recession)
Central banks have been raising interests to fight inflation so far.
This was valued at nearly $10B at the peak.
and yes I realize the majority of SPACs were pre-product scams.
I think they will stick around and go back to their previous volume. Grift-o-currency is a scam and I hope it dies soon.
As I was looking at the chart, I noticed there was a big bump from 1994 to 1995 (like from 3% to 6%). What happened then? Was it inflation? I recall rates were about 6ish since when I was in undergrad. I guess Greenspan/Y2K/dot com days.
Edited: found the answer and it is intriguing: https://markets.businessinsider.com/news/bonds/federal-reser....
The stock market craziness continued even with high rates .. wow .. didn't expect that.
The million dollar compensations (when counting stock growth) and intensity/stress around interviews and promotions is what ZIRP wrought for the elite in tech.
Would be interested if people who were mid-career in the mid 90s can comment on their perspective.
Would that mean it was the era of "Why are manhole covers round?"
I don't want to sound like someone who supports leetcode (mostly because I don't), but it seems like it's at least an attempt at measuring something related to programming skills.
Leetcode is FizzBuzz on steroids, but IMHO is not on its own likely to be more predictive of success at a company than the old MSFT way.
"how many balloons can you fit under this table"
I think it might also end up being important to remain competitive in a global market where the other economic superpower's government is willing to invest trillions into building whatever it wants and needs despite profitability[1].
Some things we want/need won't be immediately profitable, or even profitable in the long run. They might not even be things we realize we need until something unprofitable is researched and developed.
Not saying that Snap is something we need, but if the US is forced to strictly rely on market forces to compete, "free" capital via low/no interest rates is a way to kind of do it.
[1] https://tnsr.org/2022/12/chinas-brute-force-economics-waking...
In the old days, these investments were tiny compared to what is being currently spent on the chips act. I think for STEM a better approach would be a govt sponsored entrepreneurship thing .. imagine the same YC deal but we give this to any PhDs (or maybe even Masters) once in their career. I'm from Canada and when I see the cash wasted on other initiatives, I can't help but wonder why this is done. Not to say Masters or PhDs are geniuses. It is just the current system means only the well-to-do or well-pedigreed can get into places like YC (and maybe elite schools like Berkeley) while intelligence, grit and ambition are more widely distributed.
So the current hike shouldn’t have been surprising at all.
By that point you already had: 1. Facebook, obviously 2. Google+ 3. Twitter 4. Tumblr 5. Foursquare/Swarm 6. MySpace 7. Friendster 8. Orkut etc
The only real innovation they brought was the ability to parlay teens sending each other legally dubious dick pics by deleting them almost immediately into venture funding and then an IPO.
https://help.snapchat.com/hc/en-us/articles/7012318024852-I-...
It also acted as an overt anti-Facebook, in an era where that was a growing trend: No, and then limited, broadcast communication versus FB's "tell everyone everything" defaults, ephemeral messages in stark contrast to FB's "hey, remember this inane thing you said 3 years ago", and a culture of communicating asynchronously with a face attached.
This (perceived) lack of a panopticon made Snapchat a much more appealing form of social media. IMO, it's still one of the better ones.
Snap (and others) just continue to bet on the wrong things
Their cloud commits from their S-1 is still an all timer for me. Billions of dollars.
People think Zucc just acquired his way to the top when Meta has probably been the top 3, if not best, run company in the past decade.
Had Evan sold Snap to Facebook back in the day, I'm fairly confident it would be valued at least 10x what it is now.
Instagram and Snapchat today are two fundamentally different products, and Snapchat is estimated to capture more of the younger demographics in more markets.
Sure, you can say that Instagram today has about 100 billion valuation compared to 15 billion for Snapchat, but the race is far from over. Snap recently exceeded a 100 billion valuation as recently as 2021.
Also X is now worth 4 billion so apparently it's at the very least, not a growth industry.
I don't think it's so much that Snap hasn't figured out a business model - they make a billion dollars in revenue per quarter, how many companies can claim that? - it's that Snap hasn't developed a rational balance sheet due to excessive SBC. If it wasn't for SBC, they would be making money hand over fist.
Snap & PINS are two interesting case studies in how companies in a similarly challenging ad market can have potentially different strategies and outcomes. Both have very depressed stock prices, but both have a clear path to having a sustainable and highly profitable business.
Again, interest rates seem pretty irrelevant to the fate of both companies, only as much as low interest rates encourage economic investment and can fuel the digital ad market.
Some other large companies (can't mention names) have also done layoffs targetting genAI recently. I was surprised by this.
Is it really startups eating their lunch???
Meta's open source play I think comes back to regaining public sentiment about them after scandals/metaverse, and also making it difficult for established competitors to have much moat.
Given how much Snap pays their employees it was probably the right call.
Will be interested to see down the line how any startups nowadays will fare with Meta.
> The company made the decision that building up ARES would take “significant” investment and it couldn’t continue to fund those efforts.
Well, which is it? Is it so trivial that anyone can compete, or is it too costly to build it?
I think that's the wrong "either/or" framing in this case. It's not that building competing tech would be trivially cheap, but it's something around which Snap would not likely have a moat or any sort of competitive advantage. It's also in a business area that's pretty unrelated to their core offering.
The CEO's exact quote was "Leading in augmented reality means that sometimes we will fail, and I am proud that our team dared to build this business even if we did not succeed." So I'm not sure why you think "they don't want to say that."
> And they didn't know these things six months ago?
Given that the initiative was originally announced in March, I'm assuming the whole project was greenlit at least 3-6 months prior to that. And ChatGPT was only just released on Nov 30, 2022, and since then it hardly feels like a week goes by where I'm not mindblown by some new AI announcement. So yeah, I think it's entirely reasonable to think that the landscape when this ARES project was first started is very different than the one we find ourselves in today.
It's not really just startups - honestly I think 99% of those startups will fail.
The problem with generative AI as a business is basically what that leaked Google memo said, which was something along the lines of: "OpenAI doesn't really have a moat, but neither do we."
There are a few large companies who are creating these successful models (OpenAI, Microsoft, Google, Meta, perhaps Amazon with their Anthropic investment), but there are tons of other startups that are trying to come up with "individual products around AI", and I think it will be incredibly hard for them to differentiate. That is, for a lot of the "dedicated" early-stage products I've seen, I've thought "I can pretty much do this all with ChatGPT, maybe ChatGPT with plugins - why would I want to download your special app?"
I don't think all of these companies are doomed to fail (for example, I think Harvey AI was smart to focus on a very specific, profitable niche: lawyers have a lot of money to spend on this stuff, and they have specific needs e.g. around compliance, auditability, privacy, etc. that you could see a specific legal-focused AI tool addressing), but I think a lot of them are.
Even more fundamentally, they're putting the cart before the horse. They're making the classic mistake: Taking a tool/technology, and then searching for a problem to solve with it, rather than starting with the problem. I agree a lot of these companies are doomed to fail for the same reason companies founded to "Figure out what to do with blockchain" were doomed to fail. They'll no doubt suck down plenty of funding while failing though.
Snap’s biggest problem is it’s a crummy place to work with too many… let’s say personalities that are not commensurate to what they deliver. Same could be said of Amazon, though they have a far higher revenue ad product than Snap ever had, so shows how much those two are related.
To this specific quote: Generative AI meaning image generators are anticipated to replace creative agencies for display ads.
They are not good ads right now. Mondelez and Unilever who are trying this for example have not found a big breakthrough. Part of this is that neither of the two latent diffusion models create sufficiently creative creatives. You need programmers with art backgrounds to master these and the brands have neither of these. They’re super rare generally, you cannot just pull them out of the professional recruiting pool, and they most definitely do not want to work for brands or Snap.
However the appeal is in the race to the bottom costs. Hard to compete with “free.”
These AR experiences still require opinionated creative development which Snap, its agencies and the brands lack. So that’s really what ARES in particular failed.
The product people want is “Ads people like.” That’s expensive to make, which is the opposite of the incentives for creative production in the ecosystem.
That said I have found success creating and supporting interactive, instant streaming games as branded destinations for Meta and TikTok. Try the one I made for Hyundai: https://appmana.com/watch/virtualtestdrive - tap drive and see for yourself. This isn’t meant to be self promotion so much to prove that I dogfood my opinions into a proven, factual reality. You can see more at AppMana.com for huge brands like Nike, VISA, etc.
The AR experiences people make are too “2D” IMO; also they are not that fun. They might hit 15s of engagement on average, and most of that is spent loading, so it’s kind of a fake number compared to that virtual test drive which has instant loading and average engagement measure in minutes at million-visitor scales. Video ads have a median zero watch time and mean of 2.1s so there’s lot of opportunities for “marginally” better. Ultimately it comes down to costs.
https://investor.snap.com/news/news-details/2023/Snap-Inc.-A...
Of course it's mostly gone. My mother is a painter so I always appreciated the artform, but I found a new way to appreciate it that day.