What is lookalike stacking in Facebook & Instagram, advertising?
Lookalike (LAL) stacking is where you combine more than one predefined LAL audience in the same ad set. For example, instead of purely utilising a LAL of people who have purchased in the last 60 days, you might build something like:
Read the full article here.
LAL of people who have purchased in the last 60 days PLUS a LAL of people who have engaged with the Instagram page in the last 100 days PLUS a LAL people who have added 4+ items to a cart with no purchase.
The more data you provide to the algorithm, regardless of which digital advertising platform you’re utilising (across Meta Ads, Pinterest Ads, Google Ads and so on), typically the better results you should expect.
- On average, LAL stacks had a cost per result 70% lower than a 4% LAL audience.
- In 3 campaigns out of 11, a solo LAL outperformed a stack. Two of these three were lead gen campaigns and one was eCommerce. Of the 8 that had a stack as the winner, ALL were eCommerce clients/campaigns.
- All campaigns tested had a statistically significant amount of spend/impressions/clicks behind them to warrant trends being extracted from.
Whilst no two campaigns, businesses or industries should ever be treated the same, it is interesting to see a vast majority of stacks greatly outperforming a single LAL.
Are we saying you should immediately switch to stacks? Nope! We recommend it absolutely with a test for your campaigns, but a few things you will need to consider before investing in a test:
- The strength of the data feeding a LAL audience. Is it based on a manual database upload you haven’t updated in 2 years? We wouldn’t expect great things!
- The amount of data fuelling your LAL. Is it based on a couple hundred people when you’re trying to now target 5x capital cities? Whilst it’s better than nothing, it isn’t the strongest base.
- Everything plays a part. We say this often, but you could have the best LAL audience in the world, but if your creative, timing, call-to-action and one of four hundred other elements aren’t in check, you may incorrectly label the result of the test down to the audience.
Need a hand with your audience targeting? Get in touch with us and we’ll be happy to help!