Messy Landing Page Test and Campaign Testing
So we can get messy with our marketing campaigns, can't we?
A big org I am apart of did a bunch of landing page tests on a series of pages that I send A/B tested emails to. I don't have much control over the pages, and had no idea the tests would be coming down the pipeline (problems of bigger orgs).
Well, I'm pretty confident it messed up my campaign results for the past 2 weeks. If I had known about them, maybe I could have planned better for it.
Since this testing will be regular on the landing pages, and I'm going to be regularly sending A/B test campaigns through email, what processes/procedures do you recommend for testing both simulananeously? I considered stopping the email tests, but they said they wanted to continue with "business as usual". I just want to make sure we are attributing wins appropriately.
We use a shared Excel sheet that goes into our collaboration software (Confluence). This way people can see if there are any active tests on a page. If you know Excel, you could probably add logic that shades a new entry if the target matches an existing one. Hopefully, that would alert the new poster the page was already in play.
I'll admit, it's a manual process. However, it has prevented part of the problem you mention.
Send 1/2 of your email 1 with links to landing page 1
Send 1/2 of your email 1 with links to landing page 2
Send 1/2 of your email 2 with links to landing page 1
Send 1/2 of your email 2 with links to landing page 2
That way, you can determine which email works best with each landing page. If they end up deciding on landing page 1, you know what email to run next time. If they still have not picked a new landing page when you run another similar email, you can calculate which email was the best performer overall.
I really like @prodport's suggestion of a shared calendar or roadmap of your tests. Socializing testing in all marketing channels will help your business gain meaningful data faster and make smarter business decisions.
I would also encourage your landing page testing team to socialize the results of their tests with you and the larger org - have them help sell the story for you that testing works and figuring out the right message for the right customer at the right time is key for your business.
I've also suggested that large orgs look at their day-to-day processes for marketing campaign design, execution, QA, and approval to see where efficiencies can be gained. For example, if you're working with an email service provider, ask them what their product offers in the way of A/B testing to minimize the amount of versions you need to create for them. Since Optimizely can allow you to create numerous versions of your page in one test, you can simply track a (set of) query parameter(s) from email clicks to the landing page and filter your results by the value(s). The test can go out faster and get results on all your meaningful segments.
How are others in this Community marrying email testing with landing page testing?
Solutions Architect | Optimizely, Inc.