DMP IQ Test: How Intelligent Are Your Ads?

What if I told you that you could take your DMP and make your ads even better by integrating a Customer Data Platform? 

Let’s start by covering some of the basics and then let’s talk about how to make your ads even more targeted and more intelligent using all types of customer data.

What is a DMP?

If you’re new to the concept of a DMP, this article on gives a good overview.

A DMP is a marketing platform that is “most often used to manage cookie IDs… which are subsequently used to target specific users with online ads.”

In the fast-moving world of ad tech, definitions and functionalities are often fluid. But whatever a particular DMP might also do, they all have this essential function in common. A DMP stores cookie ID’s and communicates with programmatic advertising software to put a particular ad impression in front of a user carrying particular cookies elsewhere on the web.

A meditation on the ins and outs of cookies is best left for another time, and probably someone a little bit smarter than myself. In the context of a DMP and programmatic media, cookies allow an advertiser to serve ads relevant to the websites a user has previously visited and presumably products and content they are interested in. This is often called retargeting.

What I want to emphasize here is that while cookies are valuable and can help optimize digital media spend, if all you have on a particular user is a cookie ID, then you don’t know much about them at all. So unless you can prove otherwise, the users you have cataloged in your DMP are unknown users and fall into your anonymous Brand Audience.

Yeah… tell me something I don’t know!

Okay, so you’re a marketing pro and you don’t need to be taken back to school. But here’s why you should care about how CDPs and DMPs work together.

Umbel is a normalized index of customer data. Within the marketing organization, Umbel allows marketers to identify individuals across data sources and marketing systems by providing a universal key, an Umbel ID. By matching an Umbel ID to a user’s DMP ID, you unlock the full arsenal of customer traits and behaviors to inform your digital media spend and programmatic impressions. 

This is true whether you’re on the supply side or the demand side of the media buy. 

On-Site Display (Supply Side)

Advertising makes the publishing world go round.

Umbel’s media clients, like Vox, want to be as strategic as possible in serving visitors to their site with the right ad impressions. This serves the interest of their brand partners, who paid good money for the ad placement, as well as the experience of their visitors.

Using a DMP by itself, Umbel’s media clients have been able to provide a certain amount of precision in serving ads. If someone hits a publisher’s site after browsing the hottest new TV’s or mobile devices on a retail site, they are able to serve a relevant ad.

Using Intelligent Ads, media clients have been able to optimize the performance of these ads in real-time by including a visitor’s on-site behaviors into the calculus of serving a particular ad to a particular visitor in a particular place.

Where the moneyball really comes into play is bringing in data from your base of Registered Users to inform these impressions. If a user is logged in while browsing a publisher’s site, any other user level data points the publisher has can be included in the decision to serve a particular ad.

Let’s take social data as an example. If this logged-in user has a social brand affinity for LG and the DMP gets a signal that they were shopping for a particular LG product, an LG impression might be more relevant to this user than an anonymous visitor with just a cookie.

I’ll let the experts decide if that makes the impression worth more, in a monetary sense. Either way, in this very real scenario, a publisher can add some very real value to advertisers by using other data sources to round out the profiles of their visitors for ad targeting.

Off-Site Display (Demand Side)

Omni-channel is all the rage right now.

The basic concept behind omni-channel is that a brand wants to provide a continuous and seamless customer experience across marketing channels. From an integrated marketing perspective, fully realizing the omni-channel vision involves generating insights about the customer journey from disparate customer data sets, both online and offline.

Data sources that play a part in informing this unified customer journey include web analytics, social, beacons, transactions, DMP, Ad Server, ESP, CRM and more. In this context, matching an Umbel ID with DMP ID’s becomes extremely powerful.

Using a Customer Data Platform in conjunction with a DMP allows marketers to use digital impressions to target customer segments crafted from omni-channel behaviors and attributes. It will revolutionize the way you spend your digital media budget.

Lonely shopping cart seeking interested buyer

After an online shopping cart is abandoned, it’s common practice to send the customer a follow up email with an offer for that product. It works. But what about the customers who still don’t convert after the second touch via email?

By integrating a CDP with their DMP, marketing teams can put ad impressions for this offer in front of the holdouts. On top of that, they can exclude the customers who did convert from the email from the audience segments being served ad impressions for that product.

Putting an impression in front of someone who already bought that high dollar widget is just wasteful. Sell them something else! When it comes to making the most of your digital media spend, avoiding mistakes is just as important as capitalizing on successes.

Some cookies just taste better

Using intelligence from Registered Users as a starting point can do a lot to inform capturing unknown users from your Brand Audience with look-alike modeling. With all of your customer data at your disposal, the sky’s the limit for informing DMP targeting.

Using a Customer Data Platform, you can cross reference all of your data points for individual customers. Some customers spend more than others, and by using LTV metrics you can trace your high value customers back to their initial registration event.

For example, you might find that many of your big spenders have more in common than household income. Not only could you retrospectively see which ad impressions and offers converted these high rollers, you could even correlate which cookies they had on them when they converted.

How would you adjust your programmatic buying if you knew a customer who had previously taken certain actions on certain websites would spend 5x over their lifetime than DMP conversions from other cookies?

With a Customer Data Platform, your DMP can do more than retargeting. You’ve now entered the realm of pretargeting.

And if you missed my last two posts, learn about the Physics of Customer Lifetime Value (Part 1Part 2) where I introduced the idea of Umbel’s Atomic Audience Model.