CPE

Cost per engagement, or CPE, is your ad spend divided by the number of engagements the ad received. What counts as an engagement depends on the platform and format. On social platforms it commonly includes reactions, comments, shares, saves, and link clicks. On YouTube, Google's help page says a view of a skippable in-stream ad is counted when someone watches 30 seconds or interacts with the ad, whichever comes first, and for Shorts ads a like, comment, share, or tap on the call-to-action counts as a view. These definitions are the denominator, and they differ, so CPE from different platforms or formats is not directly comparable. Because the platform decides what counts, CPE is a flexible cousin of cost per click, and engagement campaigns usually report a lower CPE than traffic campaigns report CPC.

Because CPE bundles unlike actions together, it is only as meaningful as the mix. A comment from a prospect asking about pricing and an accidental tap on an image both add one to the count. Platforms optimize delivery toward whatever event you ask for, so an ad set optimized for engagement will find people who engage often, which is a different group from people who buy. Cheap engagement tends to come from placements and audiences that interact loosely, which is why CPE can look excellent while sales are flat. Engagement can still be valuable. Comments and shares send signals to your audience, and people who engage can be retargeted with a stronger offer. Compare it against the same campaign's cost per landing page view.

CPE interacts with objective, bidding, and creative. In Meta, an engagement objective sets the result to engagement events, so your cost per result in that ad set is effectively CPE. Creative that asks a question, runs a poll, or offers a giveaway lifts engagement and lowers CPE without telling you about intent. If your actual goal is leads or sales, CPE is a diagnostic at best: it tells you whether the creative stops the scroll, not whether it brings the right customer. Engagement also feeds later retargeting audiences, which is a legitimate use. Watch for bot or low-quality engagement in the comments. A burst of emojis from accounts with no history is cheap, and it will not buy anything.

The common mistake is treating a low CPE as audience affinity or brand strength. A giveaway can post a very low CPE and attract people who want free things. Another mistake is comparing CPE with CPC: a click is one specific action with a clear next step, while an engagement is a mixed bag. A third is optimizing for engagement to build social proof, then spending on boosting posts without any follow-up path. Check what counted: if most engagements are reactions and few are link clicks, the ad entertained people rather than moving them to your site. For a local business, a handful of engaged neighbors who comment, tag a friend, and then visit is worth more than a thousand idle reactions.

To use CPE, break it down by engagement type in your reporting. Look at the ratio of link clicks, shares, and comments to reactions. A healthy engagement ad for a service business has some comments that are questions and some link clicks. This week, pick one ad set optimized for engagement and compare its CPE with the cost per landing page view or per lead from the same audience. If the cheap engagement does not produce any downstream actions, move budget to a goal closer to revenue and use engagement only as a creative test. Record which engagement types dominate each month so that shifts in the mix do not get mistaken for shifts in performance.