Ask HN: SEO impact of HN's URL “itemid” vs. an actual title?

51 points by a_small_island ↗ HN
What is the reasoning behind only an itemid in the URL for HN? Is there an SEO impact?

For instance, a url on reddit may be:

https://www.reddit.com/r/askscience/comments/4ea7ee/what_would_the_horizon_look_like_if_you_were/

while a link on HN looks like:

https://news.ycombinator.com/item?id=11465163

53 comments

[ 2.8 ms ] story [ 125 ms ] thread
HN wasn't intended to be used as a tool to improve your SEO. It may increase your visibility to actual people or you may sometimes, through HN, get noticed by other websites but don't expect it to directly lead to an improvement in the SERPS.
I think the question may be regarding indexing HN itself, not the sites-linked-to. If I wanted to find discussion about "what would a horizon look like ... " by searching those terms, would reddit be favored because of those terms in the URL?
It depends on a multitude of factors but a huge weighting will be given to the page with the largest number of high quality links pointing to the site. Notice i said 'high quality'...

The URL structure isn't always that important, but it helps.

> The URL structure isn't always that important, but it helps.

If that's the case then all else aside the answer to the question being asked is, "Yes, there is an impact" and perhaps "but it's not significant enough to justify changing HN"

HN has a search service that does not use Google. Use that instead. Or from Google use "site:news.ycombinator.com" after your keywords to restrict google searches to HN.
I'll concede that the goal of the asker ( a_small_island) is not crystal clear but I think you should re-read what's being asked.

I doubt it's "I wish someone could tell me how to search for things on HN" more likely it's "Why did HN do it this way? Wouldn't it have been better this other way? Or perhaps is there nothing at all to be gained?" IOW the goal is to understand the motivation behind the different approaches between reddit and HN.

(comment deleted)
There is an SEO impact, however HN is not that oriented on general public and it has no ads so it doesn't really matter.
> There is an SEO impact,

Any source for that?

Technically "no" the SEO goals for HN would appear to be unaffected by the choice :-).

And in this case its actually a good thing. If you scroll through the /new pages as I do you will see that a lot of people try to use HN like Reddit as an SEO tool to get more views to their web site. That can be facilitated by a link baity headline cum URI which gets indexed with the keywords of interest of the day.

By simply putting 'itemid' in the link text HN gives very little "link love" to keywords and so is not as easily exploited by "digital presence" folks (aka people who try to SEO their client's sites or products).

I can think of a couple of impacts:

- Shorter urls are easier to copy and share.

+ Keywords in the url may increase search engine rank.

+ Full title in the url helps readers know what they're clicking.

There are advantages to having the full title in the url for both SEM and readers. However, as others mention, HN wasn't designed for SEM and contains no ads to profit from it.

>Full title in the url helps readers know what they're clicking.

Almost every website I've seen lets you re-write the url. For example,

https://www.reddit.com/r/worldnews/comments/4eaqjb/something

and

https://www.reddit.com/r/worldnews/comments/4eaqjb/anything

both go to the same link.

What is the reasoning behind only an itemid in the URL for HN?

It's easy to code

Is there an SEO impact?

Probably, but as SEO is search engine optimization, if a site doesn't care about search engines, then it also probably doesn't care about optimization of those searches.

To be fair putting a title and an item id is nearly as easy to code, just the "overhead" of a rewrite rule.
It's not exactly a huge amount of work, but validating and sluggifying text is also part of it.
I'm pretty sure HN doesn't want urls to provide an SEO boost either for themselves or the people submitting - and if so, I would tend to agree with them.

It's well know that pg doesn't want this site to have mainstream appeal, so having HN articles list high in search engines would probably be a problem they would want to avoid, but also, submitters shouldn't have an incentive to use this site to boost their own SEO by submitting low-quality linkspam.

If it were me, I would go even further and route every link through a dereferring proxy just to mess with their analytics as well, and block everyone except maybe IA through robots.txt. For a site which is meant to be about discussion and thought-provoking stories and not content aggregation for the sake of ad revenue, I think SEO is a cancer.

A site doesn't need to be selling something (ads) to implement good SEO

What is the reasoning behind only an itemid in the URL for HN?

Like someone previously mentioned, easier to code, less thought required around site architecture and optimization

Is there an SEO impact? Yes, from a basic standpoint, descriptive URLs are easier to crawl, index, and rank accordingly. They help readers find info better when searching for questions + answers

From a more highlevel standpoint, descriptive URLs and an optimized site structure helps in many ways including SEO, analytics, accessibility, and more. Reddit does it well

> What is the reasoning behind only an itemid in the URL for HN?

The reason is because it's the minimum information required to do the job. HN was built to function by PG who built it old school as a demonstration of his custom programming language, it's clearly not been optimized for SEO and putting a slug in the URL adds nothing functionally and thus the engineer had no reason to add that feature. It also uses tables for layout, has embedded style information, and saves all state to files on disk using no database. It's hardly a "best practices" type application.

> Yes, from a basic standpoint, descriptive URLs are easier to crawl, index, and rank accordingly.

Why are they easier to crawl and index? Most people do shoot themselves in the foot with some duplicate content if they try to implement those speaking URLs.

And "rank accordingly"? Like keyword tags help to rank better?

Why does the URL text matter for SEO? Wouldn't a crawler be looking at the page content?
Frankly, the usage of having keywords in the URL is a very minimal factor. If everything were equal for a URL with the keywords in the URL and one without, there wouldn't be much more of a benefit to the one with the keywords in the URL.

However, if your overall site structure is one that has topics and subtopics or categories and subcategories, it would help the user see that site structure. For example, in the reddit example above, users can get directly to the askscience subreddit by removing part of the URL and going directly to https://www.reddit.com/r/askscience/. Setting up URLs like this is a good practice, the URL would then follow the site's breadcrumb trail.

For H/N, I don't really see any benefit, at this point, for using keywords in the URLs. There's just no SEO benefit.

When I Google for "SEO actual title" this page is the first result.

https://www.google.com/search?rls=en&q=seo+actual+title&ie=U...

That's a fairly specific search term though. I'm not sure I'd use that as the only evidence that it doesn't matter.
It's really not that specific. It's not the only evidence that it doesn't matter, but if google can pick up on a 3 word query (one of them being "seo", an extremely competitive keyword) in <30 mins then it's safe to say hn is doing fine.
Google polls sites that provide primarily discussions regularly. Doing an exact query match makes the most sense to do as step 1 of a query. So even if HN was trying to actively prevent this I think Google would still have an easy time.
exact query match

You keep saying exact query match, but commenter above used 3 non-consecutive words out of the 12 words in the title, and did not use any sort of quoted search text.

Indeed, it even works with just two words out of the title:

https://www.google.com/search?q=seo+actual

The point being that HN's use of just the id in the URL has a minimal, if any, effect on search ranking.

> You keep saying exact query match

Only said it once =/

> The point being that HN's use of just the id in the URL has a minimal, if any, effect on search ranking.

A single data point absolutely does not indicate whether the id in the URL affects SEO or not. Your Google search is context specific (to you) so I would certainly expect it to show up near or at the top. But what about those who have never touched that page or even have gone to HN? The more specific the title the better overall but none of this gives us data about the id in the URL being good or bad for SEO.

Likely I think it's more of a UX than an SEO thing. But still don't go off a context specific, single data point to decide whether something is true in general or not.

Your Google search is context specific (to you) so I would certainly expect it to show up near or at the top.

New VM at aws, no previous Google searches. Still #2.

As the person who wrote the code for the SEO part of the reddit URL, I can tell you that there is definitely an impact. It made a huge difference at the time, because reddit wasn't really on Google's radar. Today I suspect it would have less impact for reddit.

For HN, I get the impression that they don't really want to be all that optimized for Google, so it probably hits their goals just fine, but it probably does hurt them a little bit. But since the words are in an H1 right at the top, probably not all that much.

Edit: The code in case anyone is interested:

https://github.com/reddit/reddit/blob/cfd979fa0119191257eadc...

Are you sure that the impact was caused by changes in the URL part and not any other changes which were made at the same time.
It's impossible to be sure without asking Google, but there weren't many of us at the time and so there weren't many changes to conflate it with.
To me your comment demonstrates why it's really hard to figure out anything worthwhile about SEO.

You wrote that code almost 8 years ago[1]. Google and other search engines have changed a lot since then. Who knows if this has any impact today?

Actually did you even A/B test it at the time? Or just turn it on along with a bunch of other changes?

It's hard to tell, and there's so much SEO (mis)information out there based on old anecdotes, and all the players doing A/B testing for this sort of thing on a big scale keep their data to themselves.

1. https://github.com/reddit/reddit/commit/353ad2a

We did not A/B test it, but we didn't release many changes at the same time either, so it's hard to say for sure, but we were pretty sure it made a difference.

Also, SEO is not supposed to be easy, Google purposely makes it hard. They don't want you to be successful (with good reason).

They've always said, make good content that easy for a human to find and Google will find it too.

That's probably the best SEO advice you can get.

I'm pretty confident that no one A/B tests URLs for SEO purposes. How would you A/B test this anyway? You can't exactly serve Google different URLs and see what works (in fact, this would ruin both approaches).
You're definitely wrong about that - there are whole businesses built on building "case studies" that try to decipher the inner workings of Google's algorithm.

These days most tests are run by buying two domains that are a combination of random letters/numbers, setting up two nearly identical sites, and doing identical linkbuilding to both - then seeing which one ranks higher.

It's not exactly scholastic research quality, but repeated enough times it gives you an idea of what's working "right now".

This method wouldn't work. To make both results comparable you would have to put the same content on both sites, which would result in one of both pages getting hit by the duplicate content problem.
It actually works great; the "duplicate content problem" only impacts pages on the same domain, not identical content across multiple sites. It's possible to make a lot of money by taking the content of mild authority sites and putting it on a high authority domain - can outrank the source sites in short order.
We could have only done the special URLs for 1/2 the content and then see if the inbound requests from Google were more heavily weighted towards the ones with the special URLs.

Not a perfect test, but with enough links, it would have been a good indicator.

In large localized sites you do it by geography. (E.g. Craigslist, Trulia, Auto Trader, etc) You analyze your traffic, find two geographies that are statistical neighbors and that have sufficient traffic. One is your control, the other is your experiment. Run a test with 20 control and experiment locations.
> find two geographies that are statistical neighbors

You would probably have to clone the earth to find that.

HN doesnt really care. Even if slugs didn't provide any SEO benefit, I would implement them purely for usability reasons. xyz.com/learn-seo looks much better to a user compared to xyz.com/46765475
I don't think search engines are the main source of readers for HN.
I never understood why a slug should be considered a signal.
These might be a few:

- Though it's not much more difficult, it does show you spent a little more effort on the website to not just leave the pk there. I would think this would be more relevant for sites not using a framework like wordpress (which we absolutely know from search results filters that google can categorize sites in this way...)

- Legacy... older websites that are pure static pages will likely have good keywords in the file name, since it would be maintained primarily by humans looking at directories of files.

- Social significance. The link would be more likely to have a better social impact because the content of the page could be determined whether its an anchor tag, posted in irc, or sent in gmail. You might argue that just because the page name has the keywords doesnt mean the content is about that, but I would argue google can quickly detect and demote those kind of things (aka the page title is only really relevant to your seo if it's also relevant to the content... otherwise it's essentially just a random primary key).

Just my thoughts...

I find all these 3 are contrary to the definition of a URL, and the whole idea of using slugs just screams "game me".
My understanding is that originally [and perhaps currently] Hacker News uses the file system for storage and that the id number is the name of a file on disk. A few years ago, I recall a discussion about a reorganization of the files from a single [or few?] directories into a more broadly branched tree. The basis for doing IIRC so was to improve performance.

My impression is that this would be a simple way to produce a RESTful interface. When the resource is a file on disk, how complex does application layer routing have to be?

Anyway, my guess is that general SEO is not particularly high on the list of features to implement. On the other hand, if adding Algolia and the API then it's another story. That was also a substantial improvement to the feature set. It might even turn up the aforementioned discussion about the file system hierarchy [I think PG wrote the post].

Good luck.

As a user, I would find it helpful to have a more human readable permalink / slug for the "at a quick glance" purposes. As someone w/ some knowledge of SEO, it would most likely add value as well.
Isn't the link text human readable enough?
Doing 'SEO' is like punching smoke. Build for those visiting and using your website.