Is it possible for goals to be dependant on eachother?
For example, if I was doing a simple button variant experiment then I might have three goals:
1. Click on the button
2. Visit to the trial signup page
3. Completed trial signup.
The problem is each of these goals are measure independantly, so users who come into the trial signup page directly, or via a different way than clicking on the button, are irrelevant to my button experiment and skew the data.
Can I essentially create a goal funnel, where the user must have completed each of the preceeding goals in order to be counted?
Thanks for any help!
Variation 1 links to "/signup?key=one"
Variation 2 links to "/signup?key=two"
Variation 3 links to "/signup?key=three"
You could then create a goal for each of those specific page hits.
You would also want to run an experiment (possibly the same experiment) that changes the URL that the successful signup click goes to (and create goals for each of those URLs as well).
Hi @SimonK - Can you let us know where the experiment is actually running? The reason I ask is because the visitor needs to be actually bucketed into the experiment in order to trigger any goals. So, if the visitor completed a trial without going through an experiment page, this would not increment your conversions. Only if the visitor first gets cookied with the experiment and variation, and *then* triggers a goal would your conversions increase.
@nap0leon's suggestion is a good one - it ensures that the visitor had to start in the correct location in order for those quert string params to be present in the URL. @MartijnSch is also totally right about integrating with GA or other platform so you can easily see the funnels. You can find details on how to do this here.
Finally, check out this discussion where a similar topic is highlighted. @Josh points out a technique to use drop a cookie on click of an event and then trigger the following goal using a custom event only if the prior cookie is present.
Do any of these sound like feasible options? Let us know if we can help!
Thanks for all your responses and apologies for taking a while to reply!
In this particular case the experiment is running across our entire site as the trial signup button is in the top menu, so it's not a specific landing page.
However, I think I can do something along the lines of @nap0leon's suggestion and change the link on the button to include ?source=button or something like that, and then use that URL as a goal so that people coming to the signup page via other routes aren't skewing the figures.
Signup completion is currently tracked with a custom event, so I would probably also need a different/additional event if the user signed up via the ?source=button version of the page, and essentially create my own funnel that links all the goals together.
Thanks again for all the suggestions. I'll get back to the "lab"
For reference, here is the Product Idea for setting up a Goal Funnel. Please up-vote it.
Let us know how your work in the "lab" goes and if you find a solution that works for you. I'm sure others would love to hear about it!