A/B Testing vs. Multivariate Testing with Pardot [Infographic]

Share this article...

“Do you test what messaging drives the most engagement in emails or landing pages?” Which parts are a success? What’s a miss? How can you make it better?

Optimising your marketing is one of the most exciting parts of the process. Intentional, well-planned testing is key. Pardot has two options for testing marketing assets: A/B Testing and Multivariate testing. What are the differences? Which should you use when?

Thanks to Casey Cheshire for this content from his recently published book: “Marketing Automation Unleashed” – where he shares his Pardot and marketing automation expertise told in his energetic style.

Why Should You Test?

It’s estimated that you find the real ‘winner’ in 1 out of every 7 tests.

The Challenges of Testing

Marketers find barriers to running intentional, well-planned testing because they are:

  • Too busy: “monthly campaign, rinse, repeat…” attitude that leaves no space for experimentation.
  • Skipping ahead: No lead source tracking in place.
  • Confused: with no direction in their testing, they can’t explain the results.

Strategic Hierarchy for Optimisation

The hierarchy ensures that you first test the factors that promise the biggest results.

  • Value proposition: “What’s in it for me?’, where the ‘me’ is your buyer.
  • Layout: are you highlighting the right points? Or is it busy, with competing words?
  • Credibility: does your marketing asset build the buyer’s confidence, eg. thought leadership, expert quotes.
  • Social proof: answers when the buyer asks: ‘what did my peers choose? Were they happy with that decision?’

How to Test: 5 Steps to Follow

  1. Determine your goal: increase email open rates? Improve conversion rates on your landing pages?
  2. Create a hypothesis: Select a factor from the hierarchy above. If you make that change, what do you guess will happen?
  3. Create alternatives: copy and change something about your current assets – these will be your test subjects. Don’t feel you need multiple!
  4. Confirm your measurement: make sure your tools are set up properly for the metric you chose in step 1.
  5. Execute your test: Test alternatives with live traffic! Be sure to choose the right method: A/B or multivariate.
A/B Testing
Multivariate Testing
Compare one changeCompare multiple changes
Quick SetupLonger Setup
Requires smaller sample sizeRequires larger sample size
For testing emails in PardotFor testing landing pages in Pardot

Add Comment