Trivializing Engagement Is A Fool’s Errand

adamkleinbergData-Driven Thinking" is written by members of the media community and contains fresh ideas on the digital revolution in media.

Today’s column is written by Adam Kleinberg, CEO at Traction.

Math wonks are quick to deprecate the value of engagement. For those who believe spreadsheets and algorithms never lie, this makes perfect sense. If you can’t prove the effectiveness of your advertising with math, it must not be working, right? Some have described engagement as a “red herring” because there isn’t any “solid proof that engagement drives brand love.”

Sad news: Digital metrics are imperfect. They can’t measure everything. Basing 100% of decisions on the things backed by “solid proof” is a fool’s errand. Reach and frequency are comforting fallbacks because they are things we can count, but the fundamentals of the game have changed.

Not all impressions are created equal. The most important impressions are those that make one. Advertising is now a function of reach, frequency and impact.

Brand love is a description of emotional impact. It is manifested when creative touches a human being. Engagement is a far from perfect or consistent measure of the degree of that impact, but to relegate it to an afterthought is to relegate the value of creative as an afterthought.

Perhaps marginalizing creative works at the bottom of your funnel, but not when we’re talking about branding. Doing so represents a fundamental misunderstanding of humanity. Human beings feel stuff. When you make them laugh or cry or think or pump their fist in the air, you make an impact.

You can’t put a pixel on somebody’s nose to see if milk comes out. But having been a creative for 20 years, I can tell you unequivocally that when you present work to a room full of people and they crack up, other people who see that work will crack up when they see it too. And that laughter has impact.

Engagement is hard.

Last week I was at an industry conference and sat in a roundtable session with about 40 seasoned digital marketers on how to measure engagement. The big takeaway? Everyone seems to know engagement important, but nobody's sure how to measure it.

That doesn't mean you shouldn't. Too often, we measure what's easy to measure, rather than what's right to measure.

The challenge exists because just as not all impressions are created equal, neither are all engagements. Someone watching a 15-second video is not the same as someone watching a 60-second video. Someone sharing content, spending time on a website, viewing multiple pages and clicking "like" or "favorite" can all be defined as engagements. Each has a value, but that value is not the same across the board.

Consider a blended cost-per-engagement model as a method for measurement. Marketers should have a conversation with their media planning team about what they value and assign a weighted engagement multiplier to each engagement in the campaign. Perhaps a "favorite" or 15-second video view gets a 1X, while a 60-second view or someone who clicks on a banner and looks at multiple site pages is a 3X. Then you have something to optimize toward.

Some may say this approach is based on arbitrary assumptions of value. Well, it is. But we make assumptions all the time and at least this one is based on what is valued.

I'd prefer making an assumption based on what I value, rather than make generalizations about consumers that are based on assumptions about the meaning of data. What has meaning to you? The answer lies in the heart, not a spreadsheet.

Just like brand love.

Follow Adam Kleinberg (@adamkleinberg), Traction (@traction) and AdExchanger (@adexchanger) on Twitter.


  1. "The answer lies in the heart, not a spreadsheet." - Mmh, what would your CFO say to this?

  2. Interesting perspective, but I disagree with two key arguments: 1) engagement metrics (clicks, likes etc) are much more prone to ad fraud than solid end-of-the-funnel conversions (=sales)! There are click farms, but not conversions farms.
    2) Adam, you mention the important link of engagement and brand love. The only reason brand equity is of interest is because it is supposed to be linked to sales as well - it's what makes people still buy from your brand even if there was no advertising (your foundation = being known and preferred by consumers). So everything is tied to sales -as it should be. Because even though it may be nice to have a million people liking you or engaging with your - if they get stuck in the funnel and you don't sell anything, there will be no pay cheque - and no one can live from love alone.

    • MarketersRStupid

      Everything is tied to sales? Net Promoter Score is tied to loyalty. While you could say that is tied to sales, it's very indirect. Revenue is a terrible barometer for success because people don't purchase the way they have in the past because mobile has obliterated the "funnel." Again, what you think is right, is totally wrong in our new world. I think that's what Adam is trying to say here. Engagement is a very good predictor tied to future revenue. If you don't track that then you don't utilize an ongoing relationship management model which is where sales is moving.

    • Your point about fraud is valid, NN. It poisons a lot in this industry. Still, you're point about sales is simply wrong. If all we do is focus on the bottom of the funnel, we are doing a dramatic disservice to our brands. The funnel is not a new concept and I certainly did not invent it. The premise of your argument is that you shouldn't justify any advertising that isn't directly attributable to sales. If the way most customers purchased products was by surfing the web, clicking on banners, and buying now, you're argument would hold weight, but it's not. People build emotional relationships with brands over time. Ultimately, that makes them buy more from those brands. Great brands sell more than crappy brands. That's just a fact.

  3. Hi Adam,

    I'm flattered that you've spent 9 months stewing on this. On the other hand if you want to go after wonks, you shouldn't waste time on minnows like me. Instead why not go after Archimedes or Newton and their outlandish claims based on maths?

    Clearly having people spent time with your brand is worth something. I am proud of the many campaigns we've built that leverage digital interaction to create real and measurable brand value. However chasing an arbitrarily weighted engagement objective without regard to anything upstream of it (like checking we're connecting with a decent number of the right human beings, and not on a porn site) is a weak plan.

    I don't see much "data driven thinking" here. I think an agency owes it to their clients to make assumptions based on more than what their CEO values personally. Yes, digital metrics are imperfect, and yes, creativity is paramount. That doesn't absolve us from paying proper attention.


Add a comment

XHTML: You can use these tags: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>