Recently I read a tweet by Chris Coyier rejoicing in the ease at which you can set up Google Analytic Events on your website and it reminded me about a post I had drafted but never published around setting Goals & Events.
I had begun to write up this post 2 years ago after I posted a question on Twitter and got a message back from (of the Web Analytics 2.0 fame).
I sent through more details about what I was trying to achieve for the client, the details are included below. His first response was:
The best option for complex GA requests is to hire a GAAC to go through the requirements and validate and recommend the right path. You’ll find a list here: www.bit.ly/gaac
For simpler questions this guide is normally helpful: Event Tracker Guide
I emailed Avinash back again with thanks for helping out. I explained that I was going through the Google Analytics University (a very worthwhile way to spend some spare time). Twenty minutes later he emailed back with the following:
I had a dear friend do me a favor and share some guidance, please find that below. I hope you find it to be of value.
Awesomeness
The questions
Below are the initial questions that I had discussed with the client, a traditional printing company trying to move across into the digital world. I had been brought in to help improve the online side of their business. The site had been running for around 18 months, but unfortunately the site had no analytics to begin with, so the first thing I set about doing was to work out what was happening.
- How many people are clicking through on their home page using the icons?
- What portion of people are using that rather than the Products menu bar option?
- When people land on the product page are they clicking on Order Online or Request a Quote?
- How many people click on one of those buttons and complete their transaction (get to the thankyou page).
Below are the questions again as headings with my original ideas that had been sent through to Avinash in italics. Below those are the replied answers from Avinash’s friend in blockquotes.
How many people are clicking through on their home page using the icons?
My thoughts for #1 were to set up Event Tracking (I’m using Google Analytics here) like so
onclick="_gaq.push(['_trackEvent', 'Product Icons', 'Clicked', '', 1]);"
That way I can see all the Home page products and which ones are working best. the will get switched out for each of the different products. I’m not entirely sure what to do with “Clicked” or the value, but I’ve put what I have for now.
Yes. You use event tracking for this. What you have below will work. Though I recommend keeping things a bit cleaner using:
onclick="_gaq.push(['_trackEvent', 'homepage_clicks', 'product icons', '', 1]);"So all your special homepage tracking is under “homepage_clicks”
What portion of people are using that rather than the Products menu bar option?
I’m not sure how to exactly get #2, but I assume they can then compare the total views of the page and work out the % that have come from the events.
The simplest way is only track the links in the tool bar when the user is on the homepage, then you can use:
onclick="_gaq.push(['_trackEvent', 'homepage_clicks', 'menu_bar', '', 1]);"Keeping product name consistent will allow you pivot between location and product.
When people land on the product page are they clicking on Order Online or Request a Quote?
My thoughts for #3 is similar to #1.
onclick="_gaq.push(['_trackEvent', '', '', '-', 1]);"
Here product name is the same title as what I’ve used in #1, and the is either “Quote” or “Order Online”. For the opt_label I was thinking about combining the and , and again I’m not sure what to do with 1.0
You can also use an event to track the “click”. What you have will work but keeping things cleaner:
onclick="_gaq.push(['_trackEvent', 'quote_clicks', '', '']);"you don’t need a value.
You then create an advanced segment matching on landing page and apply to this report. Or you can configure each event as a separate goal, then create a custom report with landing page and both goal conversions. This is a better way to compare by landing page.
Note that this is giving you “when somebody lands on page x, did they ever click on either link within their session”
How many people click on one of those buttons and complete their transaction (get to the thankyou page).
I’m sure #4 is possible, I just need to do a bit more research (or point me to blog posts if you can).
The answer is not apparent.
Here you can just create an advanced segment for people who get to the thank you page. This matches all sessions which have been to the thank you page once. Then apply the segment to the event report. This gives you total events for those sub-sets of sessions.
Of course to get “users” you can create a custom report and use the Total Visitors metric.
Results
After 3 months we looked through the results of the event tracking. Truth be told I looked at it every other day because I was interested, but it wasn’t worth looking at it from a decision point of view until more data had been collected.
We found that what the company had thought was the most popular products were in fact wrong. Based on the results, the home page icons were re-ordered to push some of the less known products. We noticed that of the total visits to the products that around 80% of them were via the home page icons.
Most of the people that landed on the products were willing to order online, however only 2 of the products had been set up for an online ordering option. The work to make more products available for online ordering was then moved up to a top priority.
One of the unexpected actions to come out of the testing was that 35% of people were selecting a “No Artwork” option on the home page (it was one of the icons) which we turned into offering the internal graphic designers as an online service.
Things to consider
Google analytics is awesome. That is all.
Okay, it’s not all. Google analytics is pretty awesome, but it’s not just about collecting the standard page views and working out how well you’re doing with your target keywords. Every website needs to have a goal, a purpose, it’s raison d’etre if you will.
Understanding the websites purpose also means understanding what it takes for that website to succeed, to fulfil it’s purpose, and in turn the types of analytics you will need to be aware of to know whether or not the site is achieving those goals.
I’m going to contradict myself here a little bit, but don’t just start adding in tracking events and goals willy-nilly. You need to understand exactly what you want to report on and ensure that everything you apply to the site works together and provides you data that you can use.
(Contradiction time) Get going by adding event goals to your site and see how they work. You won’t get it right the first time so keep in mind that you’re going to be ignoring these goals in months to come (so you should probably prefix them with zz to exclude them from future reports). Once you begin to understand the results and how somethings do and do not work you’ll be in a much better place to get it right (the second time)