About Pivads

Creators

Anton Cogood
Polly Chekhova

Idea

We wanted to create a database of great Facebook & Instagram ads you can piggyback off or inspire from. All the ads are handpicked and showcase various "templates" that tend to greatly perform in late 2020. By template, we mean UGC for e-com, screencasts for SaaS, fake gameplay for mobile games, etc.

Qualitative metrics we used to decide on what ads to include in the Pivads collection.
1. We can personally vouch for a few "templates" e.g. "wall of reviews" for retargeting (meaning we ran them profitably).
2. Most ads (90%) come from either hugely profitable brands (meaning it's a fact their ads work great) or brands publicly working with industry known advertising experts.
3. We also included some ads we believe must convert well. It's based on our professional intuition and experience of growing several edtech, gaming, and ecom startups.

Who is it for

Marketers, product managers, founders, and anyone who wants to promote their business on Facebook and Instagram.

How to use it

  1. Start with applying filters to narrow down the scope of ads you want to dive in. Industry and Niche filters might be paired (AND) or used independently.
  2. Your main goal is to break down a specific ad creative you liked into smaller parts. E.g. great ads usually start off with a scroll stopper (achieved via pattern interrupt or a hook). Once you identified the key parts try to ideate how to recreate them for your product.
  3. Note that ad copy might be as important as the ad creative (that's why the ads are captured in full). Again, consider piggybacking off the copy of the "example" ad.
  4. Each ad has a link to the brand's Facebook Ad Library page. Always check other ads this brand is running.
  5. When observing the brand's Ad Library page look closely at the ad publishing date. Pay more attention to the ads that have been running for at least a week. 7 days+ are usually enough to turn off the ads with the bad performance so those remaining are likely to perform well.