Imagine you’re watching the Indy 500. It’s lap two of the race and you’re thinking about placing a bet. The problem is all you know is how the drivers performed on lap one. You don’t know which drivers will get better over time, or how much they will improve. Can you make a smart bet?
As sophisticated marketers we like to avoid “betting.” We try to “invest” our budgets typically with the goal of optimizing returns. Certain platforms make the calculation of returns quite challenging. In this post we will discuss the optimal way to measure and report on your Pinterest Promoted Pins (i.e. paid advertising) data.
Calculating attribution across a multitude of acquisition channels can be challenging, but at least the data is measuring outputs on the same scale. For example, if you look at your results today for last week’s performance across all channels you can make the appropriate decisions for budget allocation. On Pinterest however there is something called the Aging or Baking effect, and misunderstanding it could be like trying to bet on a driver at lap two of a race. It will be hard to make a smart decision about where budget should be allocated because current data doesn’t necessarily indicate future performance. But when you correctly optimize your campaign for this effect it can have a major impact in your ability to reduce Cost of Customer Acquisition (COCA) on the channel. In fact, missing the opportunity to properly optimize for aging is one of the largest issues we find when assessing Pinterest promoted pins accounts.
Normally ad channels like Facebook and Google have a “click now or click never” mentality meaning if the ad gets turned off, the ad is gone forever in a few seconds. The only way to get it to show again is to pay the good folks at Facebook and Google more money. It’s something of a symbiotic relationship and somewhat straightforward to track results. But Pinterest ads can either be clicked (like Facebook or Google ads), or they can be pinned. The difference between a click and a pin? Pinned ads are essentially saved for later, and can be clicked weeks later, either by the original pinner or someone else who browses the pin on anyone’s public board. Average ads are pinned and clicked at a 1:1, according to our data, so pins are an important aspect to driving performance.
The problem you will run into is tracking this effect because there is a timing element.
Say you are comparing a four week old ad spend that has four weeks of free pin clicks with one week ad spend that has only one week of free pin clicks. It wouldn’t be a useful comparison and you might be shutting off the wrong ads or adding budget to the wrong ads because you simply don’t know where you should be spending. It’s like betting on race car drivers on lap one of the Indy 500. All you know is where performance is right now. You don’t know how much better one driver will get over time. Similarly you don’t know how much better one ad will get over time. And that’s just betting. We already established that betting isn’t a game we are into as sophisticated marketers.
So how do we solve this?
The key is to compare four weeks ago ad spend as it looked one week later (i.e. 3 weeks ago) with the last week’s ad spend as it looks now. Now you are comparing apples to apples.
Once you have the right data points to compare you can use the difference between the data from four weeks ago (as viewed one week later) and four weeks ago data (as viewed now) and extrapolate what last week’s data would like like three weeks from now.
As you can see in the chart above, a conversion that cost $100 after one week decreased to $80 after four weeks and further decreased to $60 after eight weeks. Once a statistically valid trend is established, we can then optimize our short-term decisions by projecting estimated aging effect. In this way we are able to make changes faster, maximizing the value our clients can capture.
The moral of the story? There’s no need to place bets on your Pinterest promoted pins. When you take into account the aging effect you are able to accurately project how a campaign will convert over time and make more-informed, data-driven decisions.