VP of Marketing and Growth
Test. Learn. Iterate. Repeat.Granted, it’s not quite as sexy as “Eat. Sleep. Rave. Repeat.” But that is marketing in a nutshell- not quite as sexy as you thought it would be.
Testing is the foundation for effective and fast optimization and those that forgo the strategy portion of testing often miss most of the upside as well. Fundamental to driving fast and effective tests is the ability to create a sound testing strategy with measurable KPIs, clean data and understandable/ actionable conclusions. In other words, know what you are testing, why you are testing and what you are going to do with the info, or don’t waste my time.
The first step is to build a strategy based on what you are trying to achieve and the benchmarks that drive those goals. Then, you must have an effective method of measuring the results. This is where testing platforms and tools come in. There are a lot of great paid and free tools available to a/b (multivariate) test everything from imagery, to messaging, day parting, content, landing experience, funnel experience, offer, etc.
There are also tools you may currently use to run ads that can be employed to test and even self-optimize. Most ad servers allow ad testing and can optimize on the fly, based on CTR, conversion, bounce rate, or engagement. Platforms such as Facebook and Google (GDN) can also be set to self-optimized based on this criterion. Imagines in Facebook carousel ads for example, can be set to display in the most optimized order based on CTR or CVR.
For more complex testing such as sales price, a more complex testing methodology may be needed. Some time ago I tested offering discounts on an annual subscription product. I compared a deep discount vs a slight discount, considering initial conversion rate and renewal rates. Intuitively you know that conversion rates on the deep discount should be higher, and they were. But I wanted to understand the LTV and overall revenue effect. It turns out that customers that purchased on the slight discount were significantly more likely to renew (at full price) than the deep discount customers. This fact combined with their higher initial order size made their LTV close to twice that of the deep discount customer.
Another important issue is building meaningful cohorts. You can create these groups based on almost anything with the data that is available from first and third party sources, so selecting criteria that is meaningful is critical to success. For example, if your product or service is not seasonal then building cohorts based on initial visit month may not be meaningful.
This also brings up the subject of prioritizing testing based on potential ROI. Do not run tests on the color of the “Buy Now” button when you know that testing the length of the funnel may have a much greater effect than red vs. green! Certain channels will always be more lucrative for certain products; if your high volume comes from the affiliate channel then prioritize testing there before you go down the rabbit hole of testing paid search ads.
This article could go on forever, just as optimization testing has no end. But I would be remiss to end without mentioning RESULTS. Being able to report on actionable results is the most essential element in this cycle. The “learn” part is what we are all after. What did we learn that is going to help us drive more revenue, create higher engagement, improve retention, increase AOS, and build the brand. Don’t lose focus and get lost in the exciting world that is Marketing Testing.
A creative intuition is a powerful tool in driving marketing success, couple that with actionable data from strategic tests and you have an unstoppable marketing force at your fingertips. Let me know if you think I should have tested this content more before publishing.