Umbel won’t stop talking about data, and we’re not alone. Whether your thing is data visualization, big data, or AI, you’re not going to stop hearing about data any time soon. But don’t let all the noise overpower one clear fact: all data is not made equal.
For marketers, salespeople, and others who depend on audience data, the biggest distinction comes in first-party, second-party, and third-party data. Third-party data has come up in recent articles about the “death of X.” Sometimes that X is ad exchanges, sometimes it’s cookies, and then other times it’s the business of ad tech entirely. But a lot of marketers still don’t have a firm grasp of the different types of audience data other than third-party may not be as good as the others. So how about we unpack first-, second-, and third-party data?
First-party data: Cutting through the noise
Two-thirds (or more) of marketers believe that first-party data provides the best path to truly understanding their customers and overall better marketing performance. So, what is it? First-party data is information you own, collected directly from the customer by you.
Because it’s information that you own, you don’t have to pay anything beyond the cost of acquiring the data in the first place. Some sources of first-party data include:
Marketing automation (e.g., newsletter sign-ups or forms filled out)
Social networks (e.g., brand affinities, profile information)
Ticketing providers (e.g., past ticket purchases)
Mobile app and website usage (e.g., pages visited, time on site, referral sources)
Beacons (e.g., areas of a venue visited)
CRM data (e.g., content of phone conversations, sales engagements)
Why is first-party data a big deal? On their own, any of those sources are great data points. If you’re able to combine that data in an actionable way, though, collecting first-party data delivers amazing results for your brand and for your fan. You know nearly every time how someone is interacting with your brand: you can answer what channels they prefer, what they don’t, where they spend their time, and what of your messaging is working and what isn’t.
Let’s take a hockey fan, for example. Your first interaction could be they sign up for your team’s newsletter. You know their email, their first name, and then where on your website they went. From your ticketing provider, you know they bought tickets to an upcoming game. At that game, they engage with an unlock code that tells you they went to a concession stand on the third floor. From there, you get brands they’re interested in from social authentication including the visiting team, which is from their hometown. You know you can then target them for ticketing campaigns for future visits from that team, and what specific messaging they respond to.
What you have to keep in mind is that the responsibility is on you for both the quantity and the quality of that data. You have total control, which means you can improve how data is collected, and make sure that you bring all of your data together in a meaningful and actionable way.
A note about cookies
Cookies are small files stored on your computer or phone that tell you if a user has been on a particular website. When a user visits a website, cookies are downloaded onto your computer and then retrieved by the website for each new page you visit (although if the user does several things on one page, that can trigger several upload events) . You can then serve those users content based on where they’ve visited or keep them logged in.
Cookies can be either first-party, second-party or third-party data. A first-party example would be when you go look at tickets for a specific game on a team’s website, and then next time you visit the site, those same tickets are suggested for you. Or if you’re checking the stats of a specific player, that player is featured on the team’s homepage the next time you visit.
The first problem with cookies, whether they’re first-party or not, is actually proving that a cookie is you. Multiple people can use the same computer, and one person will use multiple devices without necessarily logging in across all of them. The other problem is one of time—cookies are finite. Someone can stay logged in and you’ll know everything they do on your site, but after that cookie expires or if they clear cookies from their computer and don’t log back on, you won’t know that it’s them. And that’s if they even log in at all.
Second-party data: Almost first-party data
Second-party data is actually first-party data. Wait, what? Second-party data is still audience data that’s been collected by a direct relationship—it just wasn’t collected by you. One example would be data you get from partnership with a business that’s offering some products or services that might be complementary to yours. Another example would be ticketing data that you have access to from a ticketing vendor, versus ticketing data from direct sales to your team or event. Each of you might have information that would help the other either get to know your existing audience better or add new relevant leads to your database.
Another example is keyword data from Google Adwords. When someone googles a keyword, they’re not giving that information to you. They’re giving it to Google, and you’ll get it because of your relationship with AdWords. (NOTE: you’re not getting every keyword from organic searches.)
Second-party data sounds good on paper, but to use it, you’ve got to dot your i’s and cross your t’s. If your company offers audience data up to another company, you’ve got to make that clear to your customers.
Third-party data: Not even close
Think of third-party data as the frozen meal of the data world: it’s pre-packaged, it’s available whenever to whomever, but it’s just not as good. Third-party data is audience data that you usually have to purchase, where neither you nor the vendor have any sort of relationship with the audience.
Compared to first- and second-party data, you can see a few problems with that. First, you can’t vouch for its reliability. You have no idea:
When it was collected
How it was collected
Who collected it
Examples of purchasable third-party data include:
Demographics (e.g., household income, number of children, address)
Psychographics (e.g., market research data)
Various data collected over time (e.g., website visits to different publishers working with ad platforms)
While you could collect brand affinities and interests directly instead of using market research data, other data points like household income might be harder to come by, which is when it’s useful to append your directly collected data with third-party data.
Third-party cookies aren’t always necessarily purchasable. A free example would be a Facebook like button on someone’s website could that could use a cookie from Facebook. Third-party cookies are also how ad providers drop tracking pixels and ads onto publisher web pages, so if you run display ads, chances are you use third-party cookies. The downside there is that third-party cookies are now easily blocked by the consumer.
Other third-party data you might have some familiarity with (even if you haven’t purchased it yourself): if you use Facebook’s ad platform and segment your audience by household income or some other factor, you’re using third-party data. Because reliability is an issue, you shouldn’t necessarily always depend on this type of data, but it can be useful for getting broad knowledge about your audience, and identifying narrower segments to target.
So what’s the right mix of all this data out there? Collect as much first-party data as you can. You can then unite it in a DMP, and then append with the third-party data you need yourself, so you can create segments to target across both your marketing (e.g., email) and advertising (e.g., AdWords, Facebook ads) platforms, without relying on different third-party data on every network and platform you want to use. Use third-party data when you have to—use first-party data every other opportunity you get.