New Engen's Approach
In the first 3 weeks, our team (comprised of a Facebook marketing expert, a designer and a copywriter), researched niche competitors, influencers and cosmetics interests. We ran initial tests of multiple broad creative directions to plan our next move.
The partner didn’t believe digital marketing could work for them – they were running at an extremely low scale. We drove ideas from the very beginning. Our research and introductory tests revealed that competitor and lookalike audiences were the most promising directions to dig into to find success.
Making Up for Lost Time
This new partner had fallen behind while working with agencies. We aimed to help them succeed as efficiently as possible, with a range of cuts, styles and techniques. Our team began testing into audiences, by launching with their core line of 10-12 products, each in several variations. The engine learns and iterates as we do, so winning with this audience testing strategy informed the development of our audience creation and refinement tools, which empower rapid, broad audience testing.
By our third month, nearly every product was on back order.
As our relationship evolved, we pulled inspiration from other brands, particularly success we’d seen from native posts. After we launched with a creative testing cycle of 3-5 ads each week, we saw results compound. Creative and audience matches emerged and we scaled up to daily creative updates. Our team continued to drive success as new creative optimized on our learnings. We used our innovative engine to revisit previous iterations. Utilizing ideas that had previous failed, our tech and audience strategy found success.
By our third month, nearly every product was on backorder. Prior to New Engen, this partner was finding just a few dozen conversions a week, with CPAs well above target. We scaled quickly, drove CPAs to record lows and accelerated growth.
We maintained rapid creative iteration and steady growth on Facebook and Instagram. Our software and our team leveraged metrics from Facebook to determine why particular ads and products were successful, informing both creative and bidding strategies.
Showcasing Individual Products
To draw focus to their unique and extended product line, we started running PLAs in late 2016. We initially launched with 10-12 products from their core line, some of which have 5-6 different colors and styles. We quickly expanded to all of their 400 products, and allowed our algorithm to prioritize spend for their top products while still capturing additional volume.
We created and organized new shopping campaigns directly from our proprietary technology platform, pulling specific items from our shopping feeds in an intuitive and organized campaign view. The algorithm populated one product or product group without manual product groupings.
We also established a shopping hierarchy to optimize bidding and volume. The campaign prioritized to differentiate between non-branded terms and branded terms, sale items and non-sale items, etc. to help drive efficiency.
We've ramped spend, kept the CPA below target, and continued to optimize.
We found great success after splitting campaigns into branded and non-branded feeds. This increased efficiency since our algorithm can bid separately for branded and non-branded search queries. In the 10 days after this switch, spend reached record levels while CPAs hit record lows.
This partner believed that digital marketing was a dead end, but our team dramatically increased revenue by 7X in 3 months. We continued to iterate on new ways to showcase their line of products, alongside the most efficient way to connect to their audiences and gain more customers. The technology we’ve built for Facebook leverages both rapid creative iteration and our innovative bidding engine to fuel growth. As a Facebook partner, the tool allows us to move more nimbly with creative at scale, and we were happy to show this partner what our tech and our team could achieve.