This blog was written before Pardot was renamed to Marketing Cloud Account Engagement. You can read more about the name change and what it means here.
We should all be aware of how important it is to get your landing pages right. They are the last thing your prospect sees before conversion, and may well have a direct effect on whether your visitor becomes a prospect or not, which is why the value of Pardot multivariate testing simply cannot be overlooked.
You are undoubtedly spending a lot of time and resources on getting your landing pages set up, why not take that one extra step to make sure they are optimised for those all-important conversions?
Getting to grips with multivariate testing in Pardot isn’t too difficult but in my experience, there are a few frequently asked questions. This blog exists to answer your questions and help get you started.
What is Pardot multivariate testing?
The multivariate testing functionality in Pardot is used to measure and compare the performance of different versions of the same landing page over a set period of time.
It’s effective, simple to set up, and provides some really good insight into page performance.
It’s essentially A/B… testing for your landing pages but it does differ from typical A/B testing in some ways. Confused? Read on…
A/B testing vs. multivariate testing
Cast your minds to A/B testing on emails, multivariate testing is very similar.
For those who aren’t familiar, A/B testing involves creating two versions of the same email with one variable aspect (subject line for example), then defining winning criteria for which email should be used going forward.
Pardot multivariate testing is essentially the same thing, but with landing pages.
Let’s say you have a Whitepaper download page – will visitors respond better to a centred form? Or perhaps a form to the left with a snapshot of the Whitepaper will perform better?
The CTA colour is another great one, do big bright attention-grabbing buttons get the nod over subtle toned down ones? These are the kinds of questions you can begin to ask of your landing pages with multivariate testing.
As marketers, we can never underestimate the impact of design on our performance. Elements such as:
- CTA button
…can make a big difference.
A particularly great aspect of multivariate tests is that unlike A/B testing, you can create multiple different versions of your landing page.
This means your testing is a little more in-depth than you might expect from a simple A/B test.
A word of advice though, whilst you’re not restricted to just the two versions, we recommend not getting silly with this – remember, you want to test one particular variable of the page so you can make a truly data-driven decision on the landing page moving forward.
Can I control how often each version is shown?
When setting multivariate testing this up, we have some control over how many visitors should be subject to each version. This is achieved by weighting the priority of each version.
For example, let’s say you have two2 pages and you weight the priority of each evenly. Every time someone new visits that page, they have a 50/50 chance of seeing each version.
What is really important about multivariate testing is that the next time that person visits the page, they will see the same version based on the cookie stored on their browser.
Remember – If you want to test out your shiny new multivariate test, make sure you clear the cookies on your browser, or use different browsers. Once the cookie is stored on your browser, you will land on the exact same version each time.
It’s then a case of seeing which version performed better.