At Wish, we were by far one of the largest advertisers, often at times spending millions per day. To add on top of this, Wish had an extremely lean marketing growth team due to its Engineering-First approach to marketing (read my first newsletter!). This lead to some very interesting optimization challenges and philosophies that traditional advertisers likely don’t encounter. Here is a short list:
Narrow targeting wasn’t worth it, because we’re already reaching the entire world. Most advertisers today rely on interest targeting, keyword targeting, lookalike targeting, and a myriad of technologies to enhance the relevance of their ads so that they can be seen by users most likely to convert. This works fine when you have a small budget and a niche product. However, this totally breaks down when your marketing budget is essentially large enough to reach the entire world, you are already effectively reaching and exhausting all of these users already through broad targeting. As some back of the napkin math, $1000/day == 50,000 users reached a day. $1,000,000/day == 50,000,000 users reached a day * 365 days == the ability to reach 15BN users.
Resurrection was a huge portion of our marketing portfolio. Often, smaller advertisers with niche products focus on acquiring new customers. For Wish, because of the scale that we operated, we had churned through and already acquired a large portion of our TAM. It became critical at that point to develop smarter strategies and measurement methods for optimizing our spend against re-acquiring users and retaining users.
We utilized nearly all Marketing Products an advertising channel had to offer. What I mean by that for example on Facebook, was that Wish would diversify its portfolio and utilize a large number of objectives and optimization goals in our campaign setup. We didn’t just focus on optimizing for offsite conversions or installs or clicks or reach, we ran campaigns for all of these optimizations regularly, often all at once.
We utilized nearly all Ad Formats available that an advertising channel had to offer. This meant running video ads when video ads didn’t perform as well as images, and running display ads when they didn’t perform as well as search ads, etc. The idea here was that each ad format would help us reach an incremental audience segment with a higher affinity for specific ad formats. While not all ad formats were easy to scale, we worked hard to continuously experiment even on underperforming ad formats like video.
Optimizing the Wish app and post-click ad experience became a larger lever than optimizing our ads. Running ads at scale is a numbers game, and there is an upper-bound of efficiency you essentially get capped at because it just becomes impossible to optimize spend on high value users that nearly every other major ecomm advertiser is also targeting. The best we could do was continuously iterate on our creative strategies like deploying overlays on our image assets and optimize our product catalog while also balancing our spend across placements and ad channels as optimally as possible. Once we found the optimal set up and balance for a given budget, there was limited room to improve and thus ad-level optimizations played second fiddle to the opportunities that lay ahead in the app and purchase experience itself to retain our users, maximize our margins and improve conversion rates in the app.
It became difficult to scale ad channels other than Facebook, Instagram, Google and YouTube. With these ad channels you are essentially already reaching the bulk of the world’s population, and thus we found little incrementality and tons of difficulty scaling other ad channels. While we advertised across 10-20 acquisition / marketing channels at a time, the bulk of our effort was spent on the big ones.
View attribution was an enigma. This is less of a problem for smaller advertisers focused on acquiring new customers, where every ad impression counts as a new opportunity to expose a user to your product and brand. For Wish, we had such a large presence and market saturation that views meant very little vs clicks and had much higher overlap across our entire suite of tools for engaging a user.
If you are ready to begin scaling your ad spend to the next level, this list might be helpful for you consider of challenges and opportunities that lie ahead. Good luck and happy marketing.
If views mean very little, how did you factor in Wish spending so much on "view only" marketing such as sports team sponsorships? Was this the right decision vs scaling online ad channels even more?