The 6 Best A|B Tests for Mobile Performance Campaigns
August 28, 2016
Ampush Analysts combine our industry and platform expertise with good old fashioned testing to drive serious results for our partner. Across our entire portfolio of clients, spanning more than 10 verticals, here are the 6 most impact advertising AB tests for mobile performance campaigns
1. Each Audience’s Preferred Creative
Each of our partner’s audience groups has different motivations for engaging with their product or service, and we market to each group accordingly. Ampush Analysts A|B test creative iterations for numerous audience groups to find out which elements resonate most, per group.
Recently we’ve found that desktop users prefer comedy show ads while mobile Facebook users prefer drama tv show ads, men prefer action sporting images for ticket sales, and that women prefer lifestyle clothing images while men prefer the “laid out” product style.
2. Website Audiences
When retargeting website audiences, Ampush Analysts compare a number of website behavior variables to find the top performing for each partner. For some partners, recent visitors to their website outperforms time spent on website audiences- for others, website visitors who interacted with a specific page outperforms website visitors who visited the site over a specified date range. With Facebook’s pixel, Ampush Analysts can create custom website audiences from numerous variables, such as landing page visits and actions they took on each page. This gives Ampush Analysts the opportunity to set up multiple A/B test to find the highest quality users for each of our partners. Every brand, product or service will have unique website behavior that most strongly correlates with high-value customers or hot leads. Read more about advance website audience targeting here.
3. Value Prop
There are many different reasons why a Facebook user might be interested in our partner’s products and services and we A|B test until we find the top performing one! Ampush Analysts work closely with our partners to understand their user personas, and develop tailored strategies to appeal to each segment. This includes customized product pitches and creative.
Ampush’s approach to campaign management emphasizes ROAS, so our analysts hone in on our partner’s customer groups with the highest LTV and focus on understanding their most preferred value prop first.
Cities, regions, states, or a custom grouping? Geography targeting can be a key lever to unlocking performance by increasing the efficiency of your audience targeting (bigger is better) and by providing you with more options for a creative strategy (copy and creative specific to a geo typically out-performs generic imagery). Recently, Ampush was able to take geographic targeting strategy testing to a whole new level by creating “tiers” for individual neighborhoods, then grouping like tiers together. While it may seem odd to target geos not in proximity, it helped us prioritize top performing geos in a way no standard geo targeting strategy had done before (check out the case study to learn more).
5. Broad UX vs Tailored UX
Ampush compares the performance a standard user experience (most popular ad creative leading to a general landing page) against a tailored user experience (ad imagery designed to correspond with a specific landing page). The latter creates a more unified and tailored user experience that can lead to improved conversion rates and scale abilities. The combination of a cohesive user experience and landing pages tailored to a specific group drives improved click-throughs and conversions- likely boosted by the natural outcome of “self-selection sampling”. We’ve found users prefer a product that more exactly matches their needs.
One of the top advertising AB tests across all Ampush campaigns is comparing different keyword strategies to improve campaign performance. Trimming under-performing keywords can boost overall campaign ROAS. Often times our partner’s campaigns leverage groups of keywords. Ampush Analysts use our custom Keyword Performance tool to identify and compare individual top performing keywords against each other to ensure we’re always running the strongest keywords possible for our partners. Recently, eliminating under-performing keywords in a “Yuppies” keyword cluster drove a 21% lower CPA, with no reduction in volume for one of our media/entertainment partners.