For years, Amazon has only offered one way to measure the success of ad campaigns: last-touch attribution.
Under a last-touch attribution model, only the final touchpoint in an ad campaign gets credit for a sale. It doesn’t matter how many of your ads a customer saw along the way—the conversion only goes to that one ad.
Since the release of Amazon Marketing Cloud in 2021, however, it is finally possible to apportion credit to each of the ads that a customer interacted with along their journey.
This is called multi-touch attribution, and it is a model I know firsthand. Before starting Intentwise, I was part of a group spending significant time building attribution models, including multi-touch attribution, and I have the battle scars to prove it.
Many brands I speak to are excited to try out multi-touch attribution on Amazon—but my advice to them is to pause before jumping in. Multi-touch sounds flashy, but it doesn’t make sense for everyone. Here’s how you’ll know if it’s right for you.
What is multi-touch attribution?
The promise of multi-touch attribution stems from the fact that the average shopper interacts with several ads before ultimately purchasing a product.
Multi-touch attribution provides a more dynamic way to track how your ads influence that journey. Rather than attributing a sale to the last ad clicked, with multi-touch, all of the ads get some credit.
At its best, multi-touch attribution lets you discover new wins and losses from your ad campaign. Let’s say you run a connected TV ad on a streaming service. When you use a single-touch attribution model, your conversion rate on the CTV ad is probably going to be quite low. Let’s say it’s .01%.
Most likely, the bulk of your conversions came from a different ad further down the funnel. Let’s say you ran a series of Sponsored Products ads in Amazon search results that received conversion rates around 10% each.
By just comparing those two conversion rates—.01% vs. 12%—you might say that the CTV ad was a failure and the Sponsored Products ad was a huge success. But what if you knew that 75% of the people who bought from the Sponsored Products ad first saw your CTV ad?
Suddenly, that CTV ad looks like a critical touchpoint in the shopping journey. It gives you a new hypothesis to test: Just how important was that CTV ad?
What can multi-touch attribution unlock for your brand?
Multi-touch attribution allows you to:
- Test the success of your top-of-funnel ads. If you’re a heavy spender on top-of-funnel ads, like DSP or CTV ads, you know that these ads tend to function primarily as awareness drivers. With multi-touch, you can more clearly give these ads credit for starting the customer journey.
- Find hidden ad successes. When you use multi-touch, you might be surprised to discover that an ad you had written off as ineffective was in fact influencing many shoppers—and you just couldn’t see that with a last-touch attribution model.
- Identify highly successful combinations of ads. You can also use multi-touch attribution to track how different ads work in combination with each other. You can answer questions like: Which series of ads are most likely to convert a new customer, and what is the best order in which to serve them?
For our Amazon Marketing Cloud solution, Intentwise has built SQL queries that allow you to identify patterns in your shopper journeys. If you run this query, you might discover, for instance, that a certain combination of 1) a DSP ad, 2) a Sponsored Brands ad, and 3) a Sponsored Products ad together converted the highest share of shoppers. Knowing that this pattern is so successful, you might put more money behind that series of ads.
So wait, why shouldn’t everyone just use multi-touch attribution?
Yes, multi-touch attribution provides a more sophisticated way to track the success of your ads. But the reality is, switching to a multi-touch attribution model doesn’t make a huge difference for every brand.
Ask yourself: Do you have multi-touch? Before investing in multi-touch attribution, you first have to be sure that a large share of your customers are clicking multiple of your ads before converting.
For instance, if only 5% of your shoppers experienced multi-touch, it is not worth pursuing a fully-fledged analysis. On the other hand, if that number were 50%, you have significantly more data to work with, and the story gets more interesting.
Is there anything surprising here? My perspective is that multi-touch is only worth your investment if it makes you rethink the success or failure of a particular ad campaign. Let’s say a campaign is underperforming when you look at it from the last-touch lens but performing well through the first-touch lens. That’s a great example of the value of multi-touch. It implies that you are undervaluing the first campaign due to the attribution model and could be making less-than-optimal optimization decisions.
In practice, however, I’ve sometimes found that, when adding multi-touch attribution, a brand’s understanding of which of their campaigns are succeeding—and which aren’t—doesn’t change that dramatically.
Want to find out if multi-touch attribution is right for you?
Using Intentwise’s Amazon Marketing Cloud solution, we can run queries to determine whether or not utilizing a multi-touch attribution model makes sense for you.
If multi-touch attribution reveals that a certain ad—or series of ads—was unexpectedly more successful than it seemed to be under last-touch attribution, we can find that for you. That might be a sign that multi-touch attribution is worth pursuing.
