Demystifying Event Match Quality
Before getting into Event Match Quality, let’s start with a simpler question: What is an event?
An event is an action — any action — that a user performed on your website. Common events are visiting a page, clicking a button, or submitting a form. These actions are then associated with an event type.
For example, a user may have clicked a button, but what does that mean? Did they Purchase your product? Did they get linked to another page to View Content? Maybe they signed up for your newsletter and became a Lead. These are all examples of Standard Events — event types that Facebook understands and supports out of the box.
Using event types is a good first step to delivering personalized ads, but it’s not enough. Facebook needs to know one more thing: who the user is. In this case, that means associating them with a Facebook account. This is a difficult task given that there are over two billion Facebook accounts.
The Old Way
Before CAPI, Facebook associated events with users through cookies stored in the user’s browser. You may have heard of these as “fbp” or “fbc”, which are the names of the two main cookies Facebook used. These cookies contained the exact ID of the user’s Facebook account. This approach made it very easy to link events to users, but it came with three major downsides.
First, it wasn’t meant to work with CAPI, which sends events from a server instead of a browser. This isn’t as big a problem as it sounds, because you can still send the cookies to the CAPI server and then to Facebook. Deviate Tracking handles this automatically, so you don’t have to worry about it.
Second, the cookie method only worked on users that were logged into Facebook. Not only that, but the user had to be logged in in the same browser. Given that many users have more than one browser, or only use Facebook in an app, this is a major issue.
Third, it’s easy to block. iOS 14.5 proved that, which is why Facebook decided to invent a better method.
Match Quality to the Rescue
It all starts with user data. As the name implies, user data is information about the user, such as their name, email, or gender. This user data allows Facebook to filter its billions of users and home in on a single person. Event Match Quality is a measure of how effectively Facebook was able to match the event with an account.
As a general rule of thumb, the more data you provide, the higher your match quality. You will never be able to get to 10/10 match quality because some of your users won’t have Facebook accounts. Match quality also varies sharply between event types.
PageView and ViewContent events have the worst match quality. This is because these events are often generated by one-off visitors who don’t provide any user data. Other events, such as Purchase, tend to happen after the user has provided large amounts of high-value data. Financial forms in particular are tracking goldmines.
Complicating matters is the fact that many users try to avoid being tracked. All major browsers offer an incognito mode, which disables third-party cookies. Users can also install a plethora of adblocker extensions. These block ads, tracking, and a slew of other content deemed undesirable by the user. Many browsers even come with built-in adblockers. It is estimated that around 30% of users have some form of content blocker installed.
To make things worse, the data in a person’s Facebook account is often different from the data they give you. “Robert” might register as “Bob” in Facebook, but provide his full legal name when buying from your store. He might use his personal email in Facebook, but his work email on your site. Phone numbers change and people move, and it’s easy to forget to update your information when this happens.
In this complex, shifting landscape, the task of identifying users can be nearly impossible. Help make it easier for Facebook by providing as much user data as possible, even if it’s tedious to set up in every single tag.
Using User Data
Deviate Tracking accepts all the user data that Facebook allows. You still need to collect the user data and feed it into Deviate Tracking. Listed below are the types of user data that we recommend providing, along with suggestions on how to get them.
The user’s email address.
This is one of the best pieces of data you can have to get a good match quality score. The reason is simple: Facebook requires its users to provide an email address when they sign up.
A good way to acquire the user’s email is to offer them a trial or a free article. You probably already have something along these lines, and all you need to do is hook it up in GTM.
Phone
The user’s phone number.
This is also a very valuable piece of data. Facebook does not require phone number from its users, but many people end up providing one as part of their account verification process.
The bad news is that users overwhelmingly dislike receiving phone calls. This makes them hesitant to provide their phone number to sites they’ve just visited. In most cases, you should ask for their email address instead. There are two major exceptions to this. The first is account verification — people are accustomed to getting codes over SMS. You can also offer phone-based support, which many users prefer because it lets them get a quick answer.
Gender
The user’s gender. Must be either “f” or “m” (Facebook does not support nonbinary genders).
Gender is not very useful for match quality. It’s also generally seen as private information, and users will typically not want to provide it unless there’s a clear reason to do so.
An example of a case where it would be acceptable (or even expected) to ask for gender is a dating site or a social media site. In cases like that, you may as well provide it to Facebook even though it won’t have a huge impact.
Date of birth
The user’s date of birth.
This is a moderately useful piece of data. It’s generally considered private, but it’s also a standard field in many payment and financial forms. Other ways to get a user’s date of birth are through an account recovery security question, or as part of an optional account profile. You should not ask for it in the initial account registration, because users may perceive it as intrusive.
First name and last name
The user’s name. Note that these are two separate fields in Deviate Tracking.
Having both names is a strong matching signal, but either one alone is fairly weak. We recommend asking users for both their first and last name if you’re able to justify asking them for either. Financial forms almost always have a user’s name. Another easy way to get a user’s name is through an account profile, or in a sign-up form for a personalized service.
City, state, zip, and country
The user’s geographic address. Note that these are four separate fields in Deviate Tracking.
Combined, these four items are extremely valuable. We recommend asking users for all four if you’re already asking for any one of them. Common ways to acquire a user’s address are through financial forms, shipping forms, and account profiles. Another good way is by offering a map widget that lets users find the nearest location. This simultaneously provides good value to the user and allows you to record where they live.
External ID
An ID associated with the user by your website or a third-party system. Deviate Tracking automatically generates an external ID if you don’t provide one.
This is a weak matching signal, since Facebook typically won’t have access to the system that the external ID comes from.
Subscription ID
The ID of the user’s subscription.
This is a weak matching signal, since Facebook typically won’t have access to the system that the ID comes from.
Facebook login ID
The ID generated by Facebook when the user logs in through a Log in with Facebook button. This is also known as an App Scoped ID.
This is the most valuable piece of tracking data because the user has directly told you which account is theirs. You should always provide it to Facebook if you have it. The only way to acquire this piece of data is by offering a Log in with Facebook option during account creation. To maximize conversions, you should still offer a non-Facebook method. Many users consider social media logins a security risk and prefer to remain anonymous. You can still match them up in Facebook by providing other types of user data.
Click ID
An ID associated with the user’s click. Deviate Tracking automatically sends this parameter.
Click ID is a very valuable piece of tracking data because it’s only generated when a Facebook user clicks on an ad. You do not need to do anything to send this parameter.
Browser ID
The ID of the user’s browser. Deviate Tracking automatically sends this parameter.
This is a moderate-strength piece of tracking data.
User agent
The browser’s user agent. Deviate Tracking automatically sends this parameter
A user agent is a string that indicates the type and version of browser, such as Chrome 92 or Firefox 56. The user agent is a moderately strong tracking signal.
IP address
The user’s IP address.
A common misconception is that each user has a unique IP address. In fact, most IP addresses are shared by at least several users, and many IPs are shared by hundreds or thousands. Many internet service providers will also assign their customers a fresh IP once a month or whenever their modem is restarted. Nonetheless, the IP address is a strong matching signal, especially when combined with one or more other pieces of user data.
Facebook does not permit IP addresses to be sent in most countries and business verticals. If you still wish to enable it, there is a switch in the User Data Parameters section of the Deviate Tracking tag. You should monitor your Events Manager for several days after enabling this option. If you get a Potentially Violating Personal Data Sent to Facebook error, then you should disable IP tracking.
Final Notes
Match Quality might seem complicated, but it’s just a measure of how much data Facebook has on each user. The more it has, the better it can target your ads. The better your ads are targeted, the lower your costs and the higher your impact. In the end, that’s what this is all about: helping you maximize your return on investment.