Facebook Conversion Variable Analysis -Beauty CPG + Ecommerce $Case Study - Campaign
A category opening beauty brand launched with wild success on YouTube and wanted to replicate their progress on Facebook.
Variable Media analyzed all the data available from their AdWords account across all channels (video, search, and display). We combined this effort with deep analysis into audiences, platform, as well as a variety of other first and second level variables. Moreover, we were able to provide projected outcomes on a 3, 6, and 12-month strategic plan which including an optimization path unique to the Facebook platform and user engagement cues specific to the environment.
The true challenge would be the replication of success. Could we take what we have learned on what platform and apply similar methodology and tactics on a vastly different platform with similar audience tools, reach, and scalability.
Secondary challenges would include a large emphasis on user behavior specific to platform and device. Also, audience testing and segment scalability. Below will be our core areas of focus for this case study which helped us learn, optimize, and ultimately achieve success:
- Platform & Device
Where and how users engage with ads are critical. Desktop and Mobile couldn’t be more different experiences and having a clear understanding of how users interact on each can yield increased results. Moreover, performance metrics such as click-through rate and click to conversion rate generate human behaviors that can be manipulated and optimized for success.
- Creative Length
Vastly important to any campaign however even more important for video media, a variety of creative lengths are necessary to maximize results in the current campaign as well as future efforts. This strategy is even more important on Facebook due to the algorithm, quick attention spans, no sound, and primarily mobile viewing experience. Modifications in this can in some cases be more impactful than many optimization variables inside the UI
- Audience Comparison & Optimization
This strategic area for optimization will be able to rely heavily on the previous YouTube campaign results. In situations with a previous understanding of audience we apply a multi-variable analysis across both entities to ensure our best segments match, and if they don’t, we optimize accordingly.
With the performance of the campaign being the primary focused we held to strict monthly cost per acquisition (CPA) goals. Thus we tested moderately keeping our engagement (click-through rate) below campaign average while making specific tweaks to platform, creative, and audience to enhance our user conversion rate.
Below you will find our strategic marketing learning and optimization to generate increased performance.
- Month 1 & 2
Uniquely a broad audience proved to be the most effective in comparison to our other targets which included category, enthusiast, previous customer match (e-mail) and retargeting. Additionally, we gained an understanding of platform, placement, and device which proved user behavior, engagement, and conversion propensity
- Month 3 & 4
Additional audiences were tested against the optimized variables from the previous months and proved effective, improving our CPA 20%. Our vanity metrics CPC and CPV both increased causing us to pay a premium for users however our conversion rate increased 2.6x resulting in better performance.
Creative testing helped produce improved results across multiple lengths at fifteen, thirty, and long form 2+ min. With minimum spend associated didn’t have a huge impact on results, conversion rates were much higher yielding better overall performance.
- Month 5 & 6
Based on our experience and optimization efforts it was time to test our scalable segments which we helped drive below our projections. Performance is more often than not a product of optimization, but what we wanted to answer was if our solutions were scalable to meet the increase in budget if possible.
To achieve scalability with improved performance, we settled on a broad target with a small retargeting bucket limited to geo and time length. Additionally, we leverage the creative learning in month 4 promoting shorter video length across the best performing platform, device, and placement areas.
This campaign produced over two thousand conversions and just over one million views. Each month the campaign CPA improved due to our investigative audit, optimizations and creative length testing. Click-to-conversion improvement helped drive our CPA performance, which increased from 1.8% in month one to 6% in month six, which equates to a 3x increase resulting in a 40% decrease in CPA.
Our efforts proved valuable and scalable, the numbers & effort below:
- Generated best performance in latter testing months with increased spending levels demonstrating optimization improvement and scalability
Platform & Device
- Both Mobile & Desktop improved each month in parallel. Despite paying more than double for a desktop user, this environment yielded a 35%+ click to conversion ratio 15x higher than mobile
- News Feed proved to be best performing thus sourced the bulk of the budget driving the most conversions
- Audience Network & Instagram while less expensive produced less than average results and optimizations pushed solely to Facebook
- The shorter the version, the better the CPA performance, conversely the click through rate proved much higher on the longer form
- Multivariable analysis demonstrated desktop to be 3x better performing on shorter form ads than mobile which yielded another optimization round to increase results
- Similar audiences (10) proved our broad efforts were generating better than average results
- More testing can still help improve the campaign further specifically against retargeting and look-a-like audiences
- Retargeting yielded the highest click to conversion ratio above 6%. However, the cost premium placed on this small audience increased our CPA in comparison to broad. Broad also yielded more halo opportunities by reaching more people at a higher CTR engagement percentage