Delete everything on your Google Analytics dashboard. Now.

Just what should your website analytics program be doing for you? I’d like you to forget everything you know about those little blue graph lines of visitor numbers, and think again.

On Tuesday, we held a fascinating web analytics seminar and training day. There I listened as the day’s main presenter, Niall McKinney of UTalkMarketing, made this point. And I think it was an eye-opening moment for many of the audience, and worth expanding on here.

Your web analytics exists to measure the outcomes from your website. However, the key point is that you need to define what those outcomes are. Where we all go wrong is to just look at the “pre-set” outcomes which the website analytics program gives us (visitor numbers, number of pages viewed) and assume that’s what’s important. Worse, we feel that we ought to find all of these useful. People ask me: “Chris, what should my bounce rate be?” …to which the answer is: “10%, preferably. Or perhaps 90%. Or something else. How long is a piece of string?”

I think the first page you see on Google Analytics would be more helpful if – instead of having all those graphs – it just said: “Right, what do you want to measure?” Nothing else. Just a white page which said: “Right, what do you want to measure?”

That would force us to think about what we want to measure. Even “visitor numbers” may not be particularly significant. Imagine you have the biggest and most eye-wateringly expensive exhibition stand your company has ever invested in. What would make it a good exhibition? 10,000 people seeing your company looking so affluent? 1,000 people taking away your brochures? 100 people asking for a sales call afterwards? 10 people placing orders on the stand? There’s no right answer. But you’ll want numbers on some, or probably all of these things in order to be able to judge.

However, looking at Google Analytics as it is presented to you is the equivalent of just having the show organisers’ statistics afterwards. Apparently, 20,000 people came to the show. Yes, but what did we get out of it?

Look at your website and ask yourself: “What are the outcomes I need to measure in order to judge the success of the website?” Maybe they’re the things which Google Analytics just happens to offer you as its initial examples, like “visits” and “time on site” and “bounce rate” (whatever that means). I bet they’re not though.

Even if the overall number of visitors is one of the outcomes you want to measure, it’ll only be the potential prospects which you need to know about. The numbers you really want to know about will be to do with people reading product pages, people downloading data sheets, people emailing you for more information. These are the things which you need to set up your website analytics program to measure.

So have a think. Are there any neat dividing lines between website visitors I’m interested in, and those I’m not? If so, can I just measure the good ones? What actions am I trying to get these people to take on the website? Can I set up my website analytics program to measure these actions?

At the moment, you probably have Google Analytics set up with the standard “dashboard”, and from this, you’re getting the very unexciting data that in November, you’ve had 2100 visitors, so this month you might well beat October’s 3000. It looks like of those 2100, half came from Google, 200 came from Widget World Monthly Online, and 150 you don’t know about, but they came in a big group on the day you sent out your last newsletter. And quite frankly, they’re all just numbers.

Wouldn’t it be better to know that in November you’ve had 800 visitors who might be interesting, because they looked at more than one page and they’re from Western Europe (which is the only place you sell)? And to know that of these, 130 downloaded one or more datasheets, 90 signed up for your email newsletter and 45 requested product information through your enquiry form? How does that compare with last month? And wouldn’t you know it, those visitors from Widget World Monthly Online turn out to form the bulk of the new email newsletter signups, most of the datasheet requests came from your newsletter readers, and the traffic from your neglected Google AdWords campaign didn’t seem to do anything at all!

If your Google Analytics “dashboard” had all that information on it, how much more interesting would it be?

And guess what? It can. Join our Insider Programme Pro for 2010 and we can help you set up these sorts of things – and a lot more besides.

Discussion

  1. Ravi

    Hey Chris, very nice article. This is exactly what we suggest to our client that dont look at the bare nos. It can hurt you.And that is exactly why we put together nice little tool which extracts google analytics data into excel. This way people would be less keen on looking fancy metrics and more keen on doing number crunching in excel !

    Cheers for the article

  2. Andrew Reynolds

    Chris, nice article, food for thought (and action) as always. I thought I’d share an occasion where GA had quite a different use for me the other day…About a year ago I allowed myself to be convinced into paying for a listing on an online directory (doh!) – The rep called me the other day to renew the (annual) subscription. He said “let’s visit the directory site to check your stats”. Now being quite a fan of GA (and not trusting sites own stats) I politely refused and instead went to our analytics page while the rep was still on the phone. The quantity of referrals was very disappointing and the quality worse. Naturally he suggested I hadn’t configured the analytics correctly, so we made an experiment – he clicked through once, and so did I. “We’ll talk again next week” he said (full of optimism that neither click would register and therefore he’d have proved GA was wrong).

    Well the scores are in and guess what? – for the day in question, there are two referrals from his site listed on GA (one is definitely mine because I know which product I clicked on) so I assume the other one is his.

    It’s often so difficult to determine which elements of marketing actually achieve the results we strive for but in this case I could prove with clear evidence that this particular service wasn’t providing value for money – hooray for GA!

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.