The term "attribution" is becoming increasingly ubiquitous in measurement conversations these days. Read our No-Bullsh*t Guide to Attribution and PR Measurement to get a better understanding of what attribution really means.
If what attribution means to you is showing your impact on business goals, then here are a few suggestions of ways to show attribution. You still won’t be able to tie a particular press release to a sale. But you will be able to demonstrate that your work definitely contributes to the bottom line.
1. Use GA4 to show attribution
If you haven’t heard, the version of Google Analytics known as UA, or Universal Analytics, will disappear next year. In its place is GA4: Google Analytics 4. (Read more about the measurement advantages of GA4 in our article “Why PR and Comms People Need to Get On Board with GA4 and Measurement Now!”)
This isn’t an upgrade, it is an entire rethink of how we show attribution. It takes into account that we now access content on a wide variety of devices, from desktops to phones to TVs to cars. It also acknowledges the abject failure that “last click” attribution has always been. Rather than just assigning credit to the last thing that a buyer clicks on before purchase, it allows you to choose your own attribution model and weight it accordingly.
Your options include first click (what earned media is assumed be, i.e., at the very top of the marketing funnel) and last click, as well as:
- Cross-channel first click. Looks across all channels and then gives all credit for the conversion to the first channel that a customer clicked (or engaged view through for YouTube) before converting.
- Cross-channel linear. Distributes the credit for the conversion equally across all the channels a customer clicked (or engaged view through for YouTube) before converting.
- Time-decay model. Attributes 40% credit to the first and last interaction, and the remaining 20% credit is distributed evenly to the middle interactions.
- Position-based model. Gives more credit to the touch points that happened closer in time to the conversion. Credit is distributed using a seven-day half-life. In other words, a click eight days before a conversion gets half as much credit as a click one day before a conversion.
Whichever model you use, your attribution data will be far more accurate and robust than anything you could get out of the old Google Analytics.
2. Google Tag Manager
Many of us have been using Google Tag Manager for a while to add tags to our content—be it a press release or a white paper or anything else—to ensure that we know the specific source of a conversion. GA4 integrates Tag Manager into its attribution process to further identify sources of your results.
3. Pre/post surveys
Unlike GA4, which is perfect for counting actions on a digital property, a survey identifies a shift in attitude, belief, consideration, preference, or intent among real human beings. So, if your goals are to change beliefs, perceptions, awareness, or other attitudes towards your cause or brand, a pre/post survey is essential.
A pre/post survey first measures the level of belief, awareness, or perception prior to the start of your campaign. It is conducted before you launch that campaign. You then survey the target audience again after the campaign is underway or after it’s finished. Then compare before to after to see if attitudes have changed.
For media to claim credit, you need to make sure you ask the audience, “Do you recall seeing this brand in the news in the last X days/months (or however long your campaign is underway)?” That way you can compare results between those who have seen something in the news with people who have not, and contrast the difference(s) in how they responded.
4. Use existing marketing data to analyze what triggers a purchase
Most marketers and advertising folks have loads of historical data that tells them what to put into their ads or marketing materials to get people to buy a product. I learned this years ago when I was presenting to a crowd of PR folks along with product managers and marketing managers at Procter & Gamble. I was going through the latest earned media results and reporting on the percentage of stories that contained a desirable visual, a key message, and a recommendation.
The lead product marketing manager asked me how many people that represented. Since we had reliable circulation figures for the key media we were tracking, I did the math and told him. “You’re telling me that 33 million women in our key markets had the opportunity last quarter to see a desirable photo, a key message, and a recommendation from a stylist?” he asked. I responded in the affirmative. He responded that based on that figure, he could calculate how many cases of hair products were sold as a result. What he knew and I didn’t was that those three elements were necessary to persuade a woman to buy their products.
So, if you are a B2C company and have data on what moves your products, start tracking your earned media for whatever elements it takes to make a sale. You’ll then have a valid technique to show attribution.
5. Marketing mix modeling
MMM has long been the methodology used by most marketing and advertising folks to understand how many widgets they can or will sell with a particular campaign. What most of those models lack is any data from earned media.
I’ll never forget a packed room at a conference where a 3M data scientist presented data about a campaign to celebrate the anniversary of the birth of the Post-it Note. He claimed that the paid digital campaign was responsible for xx millions of dollars in sales of Post-it Notes.
Since I had measured the earned media for that campaign, I asked him if he had taken the earned media data into account when he did his calculations. He said no. I asked him why not? He admitted he hadn’t thought of it. I informed him (and the entire audience) that his conclusion was flawed because his data was incomplete; he needed to go back and add in earned media.
So, Lesson 1: Make sure that any marketing mix models that are in use include earned media. Also, make sure you have sufficient data and the right data, i.e., analyze your media for those elements that drive purchasing. ∞