What is Auto Advanced Matching and is it Affecting your Facebook Ads?
Lately we’ve been having a lot of conversations about CAPI (Facebook’s Conversions API) with clients and ad buyers alike, and realized that it has begun to shine a bit more light on Facebook’s Automatic Advanced Matching (AAM). As a result, we decided to dive a little deeper and share part of an actual conversation we had with a client. We think you will find it just as valuable as they did.
Question: If 15% of your Add to Cart events are receiving the following hashed customer information through your Advanced Matching setup: City, Email, First Name, Last Name, Phone, State, ZIP code … Does this mean that 15% are receiving all those data points but there could be a bigger percentage that is collecting one or two of those data points?
Our Answer: This is actually a percentage of a percentage. Facebook says this number is configured as follows:
The percentage is the average of the percentages of events having a match key, it represents the amount of events containing customer information related fields, it doesn’t mean that 15% of the events were matched to a Facebook user.
So, we gather that it is the percentage of all match keys (City, Email, First Name, Last Name, Phone, State, ZIP code, etc.) for that event (in this case, Add to Cart), divided by all visitors to the site who engaged in the event (ATC) to be represented as a matched percent.
Ideally, this number will ebb and flow. Currently with Shopify stores, the majority of this data is being matched at purchase, as that is where the user is filling out a checkout form with these match keys to complete the purchase.
Because of this, the purchase event should be where the overall match rate is highest. As for the Add to Cart event, we have a client who was at an 80% match rate when we put out the first CAPI blog, and a few weeks later is now at a 66% match rate. What we’re gathering is that this is the time where more consumers are adding products to their cart, but possibly not checking out and/or completing the checkout process, as they are browsing or holding off to see if they will be served a deal or planning to wait until BFCM for when the items are hopefully on sale.
Question: So, considering purchasers are at 100% here, is there not much to worry about? What would you do in this situation with lower Add to Cart match percentages, but 100% for purchase?
Our Answer: At this time, it is not something to worry too much about. We do want to keep an eye on it to understand how this fluctuates over time, but note that this is not taking into account the data that the Facebook pixel and server side data (on purchases) is passing. As ad blockers become more integrated and cookies become less valuable (and henceforth, the Facebook pixel becomes less effective) in 2021/2022, we will want to review and adjust based on those future learnings.
So, is match rate important?
Short answer: Yes.
Long answer: Yes, but not that important. The best way to think of AAM is similar to that of the additional fields used when creating a custom audience in Facebook. In that situation, we know that Facebook needs an email address to match to a user profile, but when a user does not use the email address provided on their profile, Facebook can default to data such as a phone number and name to make a likely match for an audience. AAM is the same for the Facebook pixel and while the pixel is still doing the heavy lifting, it can assist when, for example, a user has a cookie or ad blocker enabled using the match key fields you have turned on.
So how do we increase match rate?
One option would be to ask for more information earlier in the checkout process or by creating a scenario where customers have to login first. This could mean users need an account to log into, or even a loyalty play. Another way could be to use AAM’s twin, Manual Advanced Matching. This does the same thing as AAM, but is completed by custom coding these data calls into your site versus granting Facebook access to the collected data on the standard checkout form on your site.
That said, you also have to remember that match rate, even at the purchase level, can be lower, as many users may not complete all fields in a form if not required. They may also provide non-matching info. For example, non-matching email on your Facebook/Instagram profile as mentioned previously or a profile that uses a fake name.
All in all, AAM is a big support and the value of match rate will grow in time as we start to see the Facebook pixel’s abilities decline entering into a digital world without first- and third-party browser cookies.
