At Brighton SEO this year one of the talks opened with a rather bold statement:
“Delete Google Analytics”
Given that everybody in the conference hall was involved in digital marketing in some way, and how much of website visitor tracking is done through Google Analytics, you might even speculate that was a foolhardy statement and that the only thing that saved the speaker was the cordon of riot police brought in specially for this talk. But then the man on the platform was Ammon Johns – a man with almost 20 years of SEO experience who is recognised by the industry as someone with a huge amount of SEO knowledge and who speaks at some of the largest digital marketing conferences around – so the riot police were little troubled, although many eyebrows were raised.
It turns out that the main aim of the talk wasn’t actually to get everybody in the room to boycott Google, but to make us think. And that’s what I’d like you to do throughout this post – question the common wisdom that Google Analytics is the best thing since hypertext protocols and ask yourself whether it might actually be harming your business.
Why is Google Analytics so great?
It is a truth universally acknowledged that Google Analytics is brilliant for four reasons:
- It’s very easy to use
- Everyone else uses it, so it must be the best
- It integrates brilliantly with AdWords
- It’s free. Who can argue with free?
The big question is, are these really the right reasons for choosing an analytics tool? Does “easy to use” mean “easy to get actionable insights from” or something else? With Google being a hugely successful corporation, are they really giving me a huge chunk of data for free or am I paying in some other way?
Is Google Analytics actually easy to use?
Google Analytics is definitely easy to set up. It’s also easy to get data out of and it’s easy to get rid of data you don’t want. But spitting out data isn’t the point of a web analytics. The point is to provide insights that let you build testable hypothesis and so improve the performance of your platform.
We’ve all seen the Google Analytics home screen – now the Audience Overview screen – with its visitor graphs and its language breakdowns. But have you really studied it? Head over to Analytics, take a look at that Audience Overview screen and ask yourself “how can I improve my business with these data and these data alone?” I’ll give you a few minutes of thinking time.
Did you manage to find anything? I would be very surprised if you did. Now that’s quite a shocking statement: you went to the first – and so by definition most important – screen of a tool that millions of people use every day and I don’t expect you to have found anything useful. Ouch.
That’s because while Google Analytics is very easy to set up and it’s very easy to see the data it spits out, it’s actually very difficult to get real insight. Almost every valuable analysis requires creating a custom report. You want to use cohort analysis to determine whether you have taken the right approach on a channel? Custom report. You want to see which blog posts drive the most and best engagement? Set up JavaScript events then build a custom report. You want to integrate offline sales data from your CRM? No can do; you will be able to when you get Universal Analytics, but only using (all together now) a custom report.
So there are plenty of things in Analytics that could be easier. But how can we make them easier? The problem here comes not from the data being collected but from the way it’s displayed. One option is to suck the data straight in from the API to your own set of reports that can not only be branded nicely but will only show the graphs you want to see, set up in the way you want. It’s not actually all that difficult for a good developer to do, and if it saves you time each week or month then you can make a good business case for investing in such a solution.
If you can make the business case for building a custom interface for Google Analytics, though, it might be worth asking yourself the question posed at the start of this post: “is Google Analytics really the best solution out there for me or can I justify investing in something else?” Take a couple of hours to explore the web analytics ecosystem and see if you can find a solution that would make it easier to deliver real, actionable insight.
Just because everyone else uses it, is Google Analytics really the best?
I started the last section off with a challenge, so I’ll do the same here. Don’t worry, this will be a simple one with no trips off to Analytics. Ready? Define “the best”. Go!
OK, so that’s actually what a mathematician would define as “complex”: a question that’s easy to ask but difficult to answer. The reason it’s difficult to answer is twofold:
- This is probably the first time we’ve ever asked ourselves this question
- The answer depends hugely on context: who is asking questions of our data, why they want answers, who is going to do the analysis, and a whole range of other factors
The reason I asked the question is that if we can’t define what “the best” means, how can we say Google Analytics is the best solution?
There are some things it does brilliantly. Tracking visitor flow, aggregating data over multiple pages and channels, letting us look at engagement. But there are some questions it simply cannot answer. For example, what would your reply be if your boss asked:
- “The average time spent on this landing page is two minutes. Is that because they were reading the copy or because they were comparing our product to our competitors?”
- “How well are the videos on our site engaging visitors?
- “People jump from their mobile, to their work PC, back to their mobile on the train home, then onto their home computer. How can we track this happening to get a real picture of cross-device behaviour?”
- “What happens if people have cookies turned off?”
Hands up all those who said “ermmm”.
There are tools out there that can do these things:
- Crazy Egg gives you heatmaps showing what proportion of people have scrolled down a page and how many have clicked links on a given page (I personally love Crazy Egg. No affiliation, they just make a great product).
- Digital Analytix from comScore lets you track individuals across devices. Universal Analytics will bring in this behaviour to some extent, but only for people who sign in to their Google accounts while browsing
- While you could cobble together a video analysis using time on page, JavaScript events, and a pinch of salt, Digital Analytix gives you data on browser behaviour during video streaming
- Piwik is an open source (read “free and fully customisable”) analytics tool that doesn’t use cookies, so doesn’t give you the problem of not being able to track people who have turned off cookies
A screenshot from Crazy Egg used on the Optimizely blog. When a CRO tools company starts using a web analytics tool it could be interesting to take a look (Image credit: Crazy Egg)
For a lot of people those are some pretty fundamental questions that can’t be answered. But some people know enough about JavaScript – or employ people who do – that they can set up event listeners to get a portion of this data. And some people are not asking these questions. But think about whether Google Analytics has ever not given you the answer to a question, or even if you haven’t asked a question because you know it can’t be answered; if this has happened a few times then it might be a good time to head off and do that research into other providers.
Anything free is amazing. But is Analytics really free?
Now I imagine that a lot of people reading that heading have straight away thought “of course it’s really free, we don’t give them a penny”. But think about this: in using Analytics you give Google all of the data. That gives them knowledge about you and your customers, and knowledge, as we all know, is power. So you might not be paying Google cash, but you are definitely helping them keep their position as one of the most powerful companies on the planet.
But more than that, if knowledge is power and power is money then surely gaining knowledge about data and its manipulation is a great learning opportunity and one that will make you a fair return one day. As Ammon said in his talk, “Using Google Analytics doesn’t make you good with data, just with Google Analytics”. Because if you just accept what Analytics pukes out at you, are you really asking the difficult questions that will help your business to improve?
One last thought: the data that Google Analytics gets is yours for free anyway. It’s your information about people coming to your website and interacting with your services, not Google’s. Lots of companies are moving towards data warehouses now, keeping all of their information within their own domain instead of giving it to third parties. And if you have any concerns about privacy following the recent revelations about the NSA and GCHQ then you might consider them pretty sensible people.
When is “Good Enough” good enough?
This was actually going to be the title of this post, but I don’t quite have Ammon’s nerve (and it’s a great topic for a project management post so has been filed away for later use).
As we’ve seen, Google Analytics is not the best solution out there. It’s not even the best free solution out there for some people. But what it is is “good enough”. It’s good enough to get some profound insights out of if you work with it, and like Excel, even better if you can build a custom dashboard. It’s good enough if you value those insights over privacy. It’s good enough if you can’t invest the time to learn a new tool that will give you similar insights. It’s good enough if you ask it the right sort of questions.
It might be for him, but is it for you? (Image credit The Meme Wiki)
But – and it’s a big but – for you that might not be enough for you and your company. Do you work for a “data-driven organisation”? Do you want to ask hard questions, make big changes, and get big improvements as a result of the data in your hands? Do you want to stand out from all of the other companies and agencies out there who do analytics in the same way?
If “good enough” suits your needs, dismiss this post with a wave of the hand. But if you think that you might need more than “good enough” in the future, or if you really want to be a properly data-driven decision maker, or if you think that big changes will give you big results I urge you to think about your choices. Explore the options out there; even if you go back to Google Analytics, you’ll come back with more knowledge than you had before. But if you don’t go back, you can look forward to a long, exciting, and rewarding journey.
