Always Have a Baseline
All too often, campaign successes can get lost in the data. Say your brand sees a sudden increase in mobile sales. Maybe you could attribute this to fresh, new content going live—however, there are so many other things to consider. External factors like above-the-line marketing, consumer sentiment, and even something as seemingly trivial as the weather can make a difference in your campaign performance. Good science always has a baseline. Good marketing also needs baselines, so develop yours before getting too far into your campaign.
Re-think Social KPIs
KPIs for desktop can include CTR and CPA, but mobile KPIs look at just a few specific types of metrics: awareness/unique reach, in-target demographics (e.g., Nielsen and ComScore reports), and social metrics, such as shares. Programmatic social companies like RadiumOne talk about metrics such as the "sharing economy," which measures user engagement and specifically sharing with other users. However, these type of "mid-tier" campaigns should not be measured on traditional desktop metrics, such as CTR.
Test Demographic Data
Demographic data is great starting point. Start by asking: "Which demographic is currently purchasing my product?" and, more importantly: "Which demographics are not purchasing it?" There was a great case study by a razor blade company whereby they realized that growing beards was trendy again. This appeared to result in a quick decrease in sales—therefore, they needed to pivot their campaign strategy to sell "beard maintenance" products to people who weren't buying their shavers.
Invest in Surveys
The best way to find something out about your potential clients is to ask them. Mobile programmatic media has created some terrific technologies that allow brands to "ask" questions to their potential customers. MobeSeek is a great tool that does just this. Brands can invest a small fraction of their overall media budget to ask a simple question to a wide audience. For instance, a travel agency could ask: "Where would you rather go on your next vacation: (a) Hawaii, (b) London, or (c) stay at home?" The respondents then fall into three categories. You could perform A/B testing against each category. This can lead to stunning insights into consumer sentiment about your brand.
≥≥ Need a Shortcut?
1. Mobile A/B testing is an always-on iterative progress, not an afterthought.
2. Always have a baseline—always consider all potential external factors' influence on your campaign.
3. If you really want to know how your consumer feels, try just asking them.