Useless online metrics you might be tracking

  • Previous
  • 1 of 4
  • View as single page

Measurement gone wrong

The good news about digital advertising is that you can measure everything. The bad news is that you're often asked to try. Everything cannot tell you a story that makes sense, just as I cannot tell you the story of how I met my wife in a way that includes all the words in the Oxford English Dictionary. What matters is that you choose the right words, or in this case, the right metrics that will let you know if you've spent your money in a way that achieves your goals.

The problem is that we often select metrics very broadly, or for reasons that don't have much to do with the campaign goals at hand. Sometimes we're just trying to compare to benchmarks, spending goals, or even just reaching into the grab-bag of the familiar. We also have a tendency to try to measure a spectrum of different tactics with one rigid set of metrics, like deciding that length is the thing that matters most and then measuring water with a yardstick.

Choosing the right tools to determine the success of a campaign is crucial. Below are a few examples of how metrics and strategy can become unhinged -- and how we might line them up again.



Sam Diener
Sam Diener August 13, 2010 at 12:27 PM

The other thing is that viewthrough shouldn't be tag based .... but that is a whole nother can of worms.

Sam Diener
Sam Diener August 13, 2010 at 12:25 PM

Very interesting piece.

I think the other issue is that with branding ads (the non-ctr ones) your goal is NOT necessarily to get conversions but to reach people. Sometimes the picture isn't about just the internet... sometimes it's bigger. You have to look at the quantitative effect of all ads....

Click fraud is rampant --- but easy to combat.

Corey Kronengold
Corey Kronengold August 11, 2010 at 11:17 AM

Great piece, Matt.

I love telling people, "You can't measure how much someone weighs with a yardstick." Too many metrics, too many tools, not enough real, actionable insight from them.

Completion rate of pre-roll is one of my biggest pet peeve of a metric. It tells you nothing about the ad or the user's intent, other than how much they wanted to see the content that comes after it.

I often tell clients that they need to look at metrics in their proper context. Advertisers need to look at whether or not they moved the needle that they are trying to move, not an arbitrary benchmark for a vertical category that is irrelevant to the specific creative execution. Why measure CTR if the ad isn't designed to be clicked on?

At Innovid, we're looking at a new breed of metrics that are specific to the ad executions. Instead of looking at the data, we focus on what the data is telling us. We look at the creative and determine what each of the actions or engagements is a proxy for, and lay out a narrative that explains how users are interacting, not just what they clicked on.

The most important point you mention, though, is that all too often, the metrics are too easy to manipulate. And while a higher number may make a client happy in the short term, playing data games is bad for the industry overall. We need to give advertisers reasons to spend more money online, not reasons to go elsewhere.