Since the day I learned how to set up and install Google Analytics I have added it to every website, every custom landing page, every microsite. I never once questioned it, and pushed it on clients without thinking about whether they needed it or not.
I blame and congratulate Jordan Moore for making me start to question why I was doing this. Then I started questioning what I gained from my own GA account stats. Then it hit me: nothing. Actually, I did get something: hours of lost time staring into the abyss that is GA dashboard. We’ll get to that soon. But, first, the tweet that started it all:
Analytics are now off (http://t.co/PyftLa5i), feeling a sense of liberation. Happy New Year!
— Jordan Moore (@jordanmoore) January 1, 2013
Why I removed Google Analytcs from my website
The following are in no particular order. All or some or none may apply to your situation. More importantly, this is my take Google Analytics in the hopes that it helps you question whether you need it for your site. And to honest, I still use it for other sites and projects where appropriate.
Slows down the page load
Fact: the page speed decreases every time the browser has to go to a server and back. No matter how fast that server is, the page is waiting because a pipeline is being used. Milliseconds, full seconds, a minute, it still has to wait to reach out, connect, download and process the files that make the website display, or in this case trackable. The analytics.js (or legacy ga.js) file is listed in Page Speed as a 135B download for a 36.49KB file, and is only cached by your browser for 12 hours before downloading a new one. I know, 135 bytes, but for those that place the GA code before their body tag (instead of being the last thing on the page) they still have to wait for that process to complete. Add that to all the slow-loading social sharing buttons, AdSense ads, externally sourced jQuery, and you could have one slow site.
NOTE: the latest version of GA is now loading asynchronously. If you are using GA, be sure to visit your admin panel in GA and update to use the latest code for GA.
Some people don’t want to be tracked
Fact: a growing number of users are nervous of being tracked. With news reports of Facebook and Google tracking your every move, and advertisers being able to follow you to your mobile devices from desktop usage, users are pushing back by blocking ads, trackers and even GA. At time of writing, AdBlock Plus is the #1 downloaded and installed Firefox add-on. And yes, you can block GA with it. I use and prefer Ghostery to block over 1,955 known ads, trackers, beacons and widgets, including GA.
Placing a badge/notice on your site announcing that you do not track users’ every move can attract users to come back and read/shop with you. (An edge case, but still a positive one.)
The data can be overwhelming and highly subjective (if not suspect)
Fact (sort of): using too much data you don’t fully understand to make decisions can be worse than having no data. In short, when I sat down and thought about how I used the data provided by GA, I found that it was just narcissistic. Like the hours I spent looking at the giant spike in traffic when Smashing Magazine shared my previous blog post. I wasn’t doing anything with it. My website is just a portfolio with a blog for articles and tutorials. On that note, what is a local bar going to do with their website’s GA data? Probably nothing but see that 35 people visited on Wednesday last week. Seriously, they will never in a dozen lifetimes do A/B testing, or modify page content for better ranking in the ‘ales’ or ‘darts’ keyword categories.
The data provided by GA is massive and can take months to learn how to use properly (let me emphasize ‘properly’ again.) A massive industry has been built around reading and understanding GA data, and being able to come up with actionable insights from it. And to this day, many cannot agree on whether Bounce Rate is a useful metric or not.
Let’s pretend that the data is 100% accurate (another topic that can fill volumes) and 100 marketing and SEO professionals spent exactly the same amount of time extracting, researching and processing your data. Out will come 100 very different reports and suggestions to improve your SEO, TAC, page rank and a dozen other industry-specific terms. They may overlap on popular topics being thrown around that week, but for the most part each will essentially be bringing more baggage to the table than you have data. Think of it on par with reading tea leaves: Google will not give out its secrets on how it ranks sites, so experts have devised a massive industry around the subject of guessing what Google is doing with the data. Let’s be honest here, there is no science to SEO since no one can prove the outcome was a direct result of the method. And let’s call it what it is: professional guesswork based on experience in hitting an ever moving target that is altering the rules of the game itself.
Think of SEO as an archer, and Page Rank as a target in the back of a speeding truck, but the target can control the wind and even deflect arrows at will. And worse still, if the arrows hits a bullseye, the target (GA) can reach up and throw the arrow back, if it so chooses and without reason or notice. How much time and money are you going to give that archer to gain the reward of hitting the target?
The reward of ‘hitting the target’
Fortune cookie: every journey has a reward, it just may not be the one you want. Have you ever asked yourself “what is the reward of being on the first page of Google results for X, Y, and Z keywords?” Is it that you think the user will click on the link and visit your site more often? Then what? Is it that you think your sales will skyrocket? Will they really? Or is it some personal triumph like climbing Kilimanjaro?
What I can tell you is that the vast majority of website owners and SEO/Marketing have only discussed ‘getting on page one’ but have not given a single moments thought to what that really means for the website. Or, better yet, for their users.
For my wife’s skate shop, it meant rising above the skate manufacturer websites that don’t sell direct, and other local skate shops that do. It meant lending credence to their shop as being an established and knowledgeable place for skates in the community. These are laudable goals, whereas ‘selling more stuff’ is not going to be a direct result of being on page 1 of results. Users still need to feel safe and secure on your site, have a good experience (UX), and feel that they are going to be taken care of after the purchase.
Being on Page 1 and having more customers click through is great, but if they are coming to a shitty website or get shitty service, they will leave and never come back. Your Page 1 rank may be a paper tiger, unable to withstand the rain. Meaning, Google will depress your page rank if users continuously bounce, regardless of how good your SEO strategy is.
When I asked myself this question, I found I didn’t care if I was on page 1 for X, Y, and Z keywords. The web industry in San Francisco alone is massive, and not one I can afford to go up against. Secondly, I have a full-time job and currently do not want to do freelance work, so that’s no longer a priority. Lastly, my site is not selling anything, there is no advertising, and the blog posts are meant to be conversation starters, free downloads or tutorials. If people find them, great; if not, no big deal.
It’s more for Google’s sake than my own
Fact: Google profits and benefits more from our analytics than we ever do. You could spend 40 hours a week on monitoring and crunching your GA data, and you will never even come close to benefiting as much as Google does. That data isn’t private to you alone, and they do store it all and crunch it endlessly. User bibi21 on this forum brought up a good point that a high bounce rate might be used against you. bibi21 claims that an increase in traffic also meant an increase in Bounce Rate, which brought the website’s Page Rank down. Without GA installed Google won’t be able to track a visitor’s Bounce Rate or % Exit on your site. (Make your own judgment call on this.)
The data can affect your goals
Jordan Moore said “…a one off post that happens to gain a bit of momentum and out-performs the other posts on my blog; the statistical results of that article would affect my writing intentions for the next piece. In other words, my writing wouldn’t be honest, it would be trying to become something it isn’t.”
For years I pounded the desk yelling “See! Tutorials and free downloads attract more traffic! I am a genius!” But, then, thinking back on it, I never once received a paying gig by anyone following a tutorial or downloading a plugin. Over the years, thousands of people visited, downloaded, and left. And, odd as it sounds, that is exactly what the primary purpose of the tutorials and free downloads is: give something back to the community. When I saw huge spikes in traffic because of them, I became misguided in thinking that traffic would convert to paying clients. Instead of writing on topics I wanted to explore, I focused on thinking up tutorials and more free downloads instead of writing what I was interested in. And, professionally, that is a big mistake.
Fact: when using CloudFlare, I noticed 25-33% of my traffic was non-human (bots). Bots are a major part of the Web. There are good ones (these are mostly used by search engines), and there are bad ones (brute force bots, email scrappers, form/spam bots, and more). Even if they were all good, they are major part of the traffic to your site. Depending on the popularity of your site, you could have upwards of 25-40% (or more) coming from bots of all kinds. Popular sites that also have a commenting component (blog or articles)—especially WordPress and other open-source CMS powered websites—will see an inordinate amount of bots hitting their pages and admin logins to post spam and exploit known security holes. With GA, this all gets recorded as a real live visitor. So, go ahead and just lop off 25% of your site’s visitors to account for bots, since really, you should only be counting human visitors. Better yet, try using CloudFlare to block known threats and bad bots, or use a blackhole (or Project Honeypot) to trap and deny bad bots from crawling your site.
To sum it up: use it or lose it
Google Analytics was a time waster for me, and my website doesn’t need it. That bar site I mentioned earlier may or may not need it either. If they are planning on selling ads or paying someone to manage it, then yes, you would need to know if 10 people or 1,000 people are using it daily to see if the cost is worth the time and effort. GA is good for using as a baseline or starting point in projects (but know that the data can be highly subjective.) GA can also help ecommerce sites to know what products customers searched for to get to your site, and what they viewed when they got there (where Bounce Rate matters.) But, if you don’t know what to do with that data, then the time spent looking at it could likely be better spent on your business, improving the performance and user-experience of your site, or hugging your kids.
Feel free to weigh in on this below. Or maybe you have more reasons for removing Google Analytics from a website?