Counting Ghosts: A case for abandoning web analytics
Comments
Perhaps we shouldn't be taking advice on how to build better websites from an absolute garbage of a website from a usability perspective.
This sounds more like a case for using analytics intentionally, rather than abandoning it. Which is great! We should do more things intentionally.
One of the reasons I wrote a Perl script to analyse my access logs was to find discussions of my content that I can learn from and participate in. Hard to do that without some sort of analytics!
Interesting thesis. When used as intended, web analytics should be only one input to an overall strategy -- not the sole driver for all decisions. But I can see how an organization with KPI's tied to web analytics will lose objectivity -- especially when someone's job performance is tied to the analytics. It's a slippery slope.
I love this quote at the end: "Caring about quality is the heart of craftsmanship. Until you're hooked into those outcomes, micro-optimizing the individual parts is pointless."
Perhaps, as we move towards a web dominated by AI agents, quality will supersede metrics.
This site needs analytics if not just to see how much trouble (abandonment) the page scrolling issues cause.
Can't scroll or read the article from my phone...
When google analytics ended for the version I was using I switched it off from my 21m pageviews per month web site. I now only focus on the metrics that count for me and my business, and I already have that data in my database.
Beyond that, Cloudflare gives me some extra data, which I really don't need but is fun to marvel at from time to time.
What a godawful page design. Who thought it was a good idea to break and article into pieces you have to swipe horizontally to scroll between and then also scroll vertically to see each page?
Apparently someone who calls themselves a "product engineer", whatever that means.
Seems to be a personal website though, so some leeway should be given, although I also found it a bit annoying. Especially the part when you go beyond the article contents and gets sent to the "homepage", but scrolling back doesn't go back to the article contents anymore.
I'm not even sure what I just read. I thought it was going to explain why they don't use any analytics anymore and all I got as a 10,000 foot answer that could be summarized as, "Well, because!"
I don't use analytics on any of my services simply because I don't like analytics and people tracking me, so why would I do it to others?
Does it mean I don't track my business metrics? No. I still measure general conversion rates from sign up to payers. I measure things like sign ups per-month. You don't need analytics to track that. Basic metrics combined with a "CHANGELOG" file with dates/releases/fixes is plenty for my solo business. Want to know what I did in January to spike sign ups or more payers? Look at my change log.
As I understood it, the idea is that analytics are the wrong answer, or rather, if they are an answer, at all, you're asking the wrong questions. Which I, in turn, interpret as a signal for more opinionated approaches.
(Meaning, if you follow these kind of metrics closely, you may actually miss crucial opportunities, as these will always bind you to the perceived mainstream. Moreover, these analytics are still sparse and there may be hidden variables. E.g, you may have changed something in January, but what happened else in January, elsewhere, which may have had some impact? Finally, these metrics are always about intermediate goals and partial results, but never about the entire product or mission. Metrics for these are found elsewhere.)
Disclaimer: I abandoned all analytics for my own projects some years ago, and I do not miss anything. So I may be somewhat sympathetic to this.
I don't get it - it is a nice sentiment, but the crux of the argument seems to "go with your gut." The reason web analytics has been so popular is that you can actually improve your product significantly without as much subjectivity. Before web analytics, going with your gut was the only option. It turns out, threading events through a funnel that show you lost 90% of your users because your button doesn't show up on certain screen sizes is a meaningful capability. The question from there is how far do you go with it, but I don't think it makes any sense to both reject it and go on to propose such a fuzzy, underdefined alternative as "you'll gain intuition about if you're making connections."
Having built a number of apps, I don't know how to gain such intution once the userbase becomes too large to interact with directly enough to grok it. It's very easy to be fooled by a vocal minority of people who complain about your app sucking when the majority are getting value from it. Without data and analytics, it's easy to be pushed around by your most vocal users, because your gut is saying you should listen to them. Sometimes you should, sometimes you shouldn't, but the difference between the two can often be determined by looking at the data to get a bigger, less biased picture of what is actually going on with your app.
> improve your product significantly without as much subjectivity
It's fake objectivity that often leads to a race to the boottom.
Between "engagement" as a terrible proxy for user happiness/success and the idea that it's more important and better if nobody can fail to use your app's "UX" than it is if it to enables a smaller number of people who take the time to learn to do really great things, all wrapped up in the VC driven dream of hypermegablitzscaling, all a reliance on analytics does is guarantee mediocrity.
I want software created by people who care enough not to look at the data.
Analytics data are like emotions: a valuable source of information which should not be used as a direct controlling signal. It should be first thought through.
The pendulum has swung too far. We quantify everything now, chasing numbers to the point of losing our sense of quality and craftsmanship. There's a huge difference between "going with your gut" and the ability to assess, qualify, and reason about something. To understand how and why something works or doesn't. Whether from first principles or experience.
Numbers provide a happy path for consensus. If the "why" of your project is to please as many people as possible, then web analytics probably works great. I'm not arguing they're categorically useless, only trying to surface the unspoken costs I see bear out in our industry from relying on them so heavily.
I've personally never been in a situation where analytics drove change. Even mere A/B testing seems like a ghost in (most?) organizations, where management or designers decide to force through their personal preferences.
That said, I would love to join such an organization. Any suggestions on where to find data driven jobs?
Amazon?
The second half of this article explains Amazon's process around metrics, well worth a read:
> Any suggestions where to find data driven jobs?
My experiences suggest a large measure of caution. Many organizations describe themselves as data-driven and many actually do try. However humans are humans and many will be “data driven” when it aligns with what they want to do, otherwise it’s off for “more data” or a “new perspective”. In other words sometimes data driven means use data when it supports my already staked out position. I hate to be so cynical but I can’t help it.
One might argue that your cynicism is (drum roll please) data driven.
It may be less used in driving website design, but I assure you it is very very widely used in marketing to determine which campaigns are working.
At my previous job, A/B testing was widely used to check which of the small UX tweaks works better.
You could hire a decent QA department.
Then you'd run the app on all screen sizes before each git push and notice that the button doesn't always show up, or isn't always clickable.
> Before web analytics, going with your gut was the only option.
The analytics people insistent denial that usability research ever existed seems to have reached a new level.
What I wrote wasn't an insistent denial, try steelmanning next time. Extrapolating the magnitude of problems from user research is hard, since you're still sampling a distribution.
What you wrote was just denial. The insistent part comes from every time those discussions going this same way.
Anyway, and FYI, one of the most robust findings in user research is that the magnitude of problems isn't really something important to know.
If the already-released version of your product has so many problems that you have to use computational techniques to extrapolate their magnitude, then you have already lost.
Absolutely, and a good reminder that your web analytics is only one source of data you can use when making decisions. Sites and apps generate so much interesting quantitative and qualitative information that is important to compare and validate assumptions.
I build websites for small business, and since the beginning have used Google Analytics. What used to be a valuable tool that I would eagerly tell my clients to utilize, has now become an overly complex nightmare I loathe to install and rarely recommend to my clients. I recently had a client ask to use StatCounter.com (my reaction: "they're still around?!?") and it was refreshingly simple, much like the original Google Analytics.
I'm sure the new GA is super powerful, but most people just want to see how many "hits" they got, what content is most popular, and where the visitors came from. Google would do well with a simplified Analytics service just for small business.
Once you go down the optimization route, if your Google Analytics isn't super valuable it's an easy bloat to eliminate.
Another vote for plausible here. After reading what you described, I thought that’s exactly what you need. It’s cheap, especially if you have a lot of sites. It tells me more than I need to know and is wonderful as far as page weight and performance.
GA4 is actually the simple version of GA. Lots of stuff is hidden. So it looks lol Google did exactly what you wanted )‘simplified Analytics service’). However, if you want an even simpler solution: Matamo is totally free as well.
I switched to plausible.io for my small business and couldn't be happier. Maybe take a look at them
plausible imo is the way it should be. Embed the script, done. Bonus: No cookies & cookie banner. They even have instructions on how to circumvent adblockers.
Using a different analytics service misses the entire point of the article.
The point is that we shouldn't be using analytics _at all_.
I'm not sure if I agree with that or not yet - I have to think about it more. My gut says that's right.
I totally thought this was going to be a different article on phantom clicks or botnet visitors.
Been thinking about that case lately, and it makes a much better argument for me in terms of analytics.
I realized that from my own perspective as a private citizen, I'm not sure if there's any way to prove that Google and Facebook are not falsifying my clicks. I face such an information disparity in bargaining. I literally depend on the company for the tools to monitor the company. "How many people downloaded my app? Arbitrary."
Its a bit like trying to win against the NSA inside America, except they can already listen to every phone line (still finding stuff, yet another NSA leak today).
It's just too easily Machiavellian. "We've got so many new user signups, and they all love your videos...post more content like that. All the rest of you are failures, because we can just bury your content and you'll never know."
Whatever it is about, I refused to read because of this silly scrolling/paginating
[dead]
It was fine using the ← and → keys on my keyboard for page change. But it's a bit counter-intuitive...
impossible to read and to follow the narrative
I can’t even figure out how to open the article beyond the intro. There are some things websites can just leave to the browser.
A case for not holding my scrollbar hostage
And reader is greyed. Literally and figuratively there nothing there. You didn't miss anything. The irony is not lost.
*Update: I removed the pagination and scrolling effects. The entire post can be read on a single page now.
(And I didn't need analytics to understand something wasn't working lol.)
But you did need user feedback, which analytics would help with
User analytics would not have provided feedback in the form of "I don't like the scrolling on this site."
damn, came to post same aside: what is this weird storybook/scrolling-in-all-directions UI
Came to say exactly this. I can read it if I try hard enough to scroll exactly as the author intended, but it feels super weird.
Sometimes it scrolls down, sometimes it pages left. And if you want to go back using the browser you will have to click click click through each page!
I thought it was neat, but scrolling broke at page 4 so I had to use the widget along the left edge to get to page 5.
To the author: may I suggest you make page order vertical instead, so it is more natural to scroll; and maybe just use CSS scroll snapping to achieve the similar effect? (Also, when I scroll past the final page I don't want to be redirected to homepage automatically :-)
The idea is neat, but the implementation is a bit too over-engineered and hard to use. Hope this helps change that!
Yeah idk what to do on mobile, and don't care enough
What the actual heck is going on there!
Maybe they should run some analytics and check the bounce rates!
gottem
> I have a theory about why businesses lose their soul as they grow.
sighs deeply
https://www.marxists.org/archive/marx/works/download/pdf/Cap...
Articles like this always seem to miss the ways that the big analytics platforms are used in real life by marketing and product teams. If all you are doing is measuring page hits, then Google Analytics is overkill and you should be using one of the privacy focused alternatives. However, if you care about connecting your marketing tactics to larger strategic goals and measuring the impact of your work, a properly configured analytics stack is still really important.
I really like https://www.portent.com/onetrick/ for anyone looking for an intro to how analytics enhances marketing work.
Retvrn to grepping access.log
This is one of the most obnoxious websites I've ever seen. Why did they need to try to reinvent scrolling instead of just making their point?
I love the groundbreaking UX /s
Scrolling goes to the next page unless I want you to scroll to read, then it goes to the next page when you're done scrolling the content. Beautiful.
Exactly why most sites don't do this. But it's special!
Actual title should be "A case against scroll hijacking."
I’ll raise you one better I came across yesterday
https://logz.io/learn/how-to-run-opentelemetry-demo-with-log...
If it’s not obvious, try scrolling a bit then try to go back to this page
Did they erase my browser history?
Just burn your browser and get a new one.
I was trying to look up what instrumentation with OTLP looks like and got so much rage yesterday I had to quit and do something else.
OMG WTF. When parent says "a bit", do a tiny amount of scrolling, or open it in a new tab. That behaviour is downright malicious.
Thanks for all the constructive criticism on the pagination. It's an experiment I'm playing around with on my personal site (which is what they're for right?). Definitely sounds like I should have tested it on a few more devices, lol. Will work on a more readable view.
In the meantime, a few helpers:
- Scrolling "past" the end of the page will move to the next page.
- Left arrow, Right arrow will nav back and forward.
- On mobile, swiping will do the same.
- Space and enter key will nav to the next page.
> Scrolling "past" the end of the page will move to the next page.
Except when it doesn't, was my experience.
In general, for a webpage that is mostly text I would like to be able to read it with a text-only browser (e.g. Reader View on Firefox). Having all the text on one page, or at least navigation links at the bottom, would really help.
I am sure readers with disabilities (motion or vision impaired, for example) would appreciate this too!
Would also be cool if it worked with e.g. Safari's Reader view.
I have to hit right arrow anywhere from 2-4 times to nav forward; left arrow 3-5 times to nav back. Seems to be fine if I wait longer than a second or so after the page transition - otherwise I see a blue border around the edge of the tab.
Anyway, as somebody that builds things with no trackers, I agree with you and appreciate you posting this! The one thing I want to know is if/when a page or website is picked up somewhere, I'd just like to see where the traffic came from so I can respond or engage with people. I find absolute traffic numbers, and most of the data you get from analytics tools a distraction and a net negative, especially given the privacy and performance issues that come with their JS.
If you don't have any analytics, then how do you know your audience size? Is it 10 people per month? 1000 people? 1,000,000 people?
No offense but I can't accept "it doesn't matter."
Vanity metric. The only thing you need to know is how many people paid for your product.
For my no-code website package I built a simple non-tracking analytics tool (I'd rather call it 'visitor stats') and I definitely like to see what language my visitors use on their machine (guesswork for where they might come from), what screen size they have (on all my sites it's steadily more mobile than desktop, which is to be expected nowadays but was an eye-opener for me when I got back as a webdev a couple of years ago), what referring sites they come from and which pages are most visited.
For me personally, the visitor stats of my own websites are a little bit as 'likes' on social media (nice to have for a general idea about my content being liked), but then without big tech in between.
For those on mobile: scroll right (that is, swipe left) to read the article.
Phew, good thing horizontal swipes aren't used for other functionality on mobile. Otherwise it would be an unusable mess of a design that is hard to read.
Uhh it's a messed up site. I think they need some analytics lol.
That said I feel like so much of analytics this day and age are measuring what we want instead of accurately measuring. In the mess of data we generating conforms already to an existing belief. Everywhere I've worked a product guy or a ux person has an idea and makes use of analytics to prove their point and for some reason their point is always proven... I think analytics are like damned statistics.
I agree with the central point, that web analytics lacks rigor. But I don't think it's wholly true that the problem is that our own motivations cause our analysis to lack objectivity. I think the other reason the author gives is the really key one: analytics data tantalises us with the idea that it may be coherent, complete and realistic; in fact it is wildly diffuse and represents only a tiny slice of the lived experience of users.
I wrote a post on a similar topic this week. https://mechanicalsurvival.com/blog/failing-to-visualise-web...