DATA......, and now what?

Executive summary: Lately there is a lot of talks about DATA MANAGEMENT PLATFORMS (DMP), but what is it? what is for? how is it used? and most important what benefits do they play for Advertisers in their digital media campaigns? The article tries to explain this from a CMO standpoint.

Was attending a client-prospecting meeting (advertiser) some time ago in order to ask about their digital media strategy and how they work in RTB. Usually a friendly and sort of relaxed first encounter meeting.

From the advertisers side, a highly skilled and smart performance marketer attended to my questions.

The meeting soon turned from being relaxed to an advanced Q&A session based on tiny details; “give me concrete examples on how your DMP (Data Management Platform) can help me to utilize my offline DATA sources to improve performance on my online display campaigns!“ I was not prepared for this level of conversation since I did not really have all the information on the advertiser and their products. My emotions went sky-high and intelligence straight to the bottom.

I gently proposed a second meeting with a skilled platform engineer to answer specific items on the DMP.

This was the trigger to a small personal “crisis” in terms on how to approach companies with high expectations on DATA and how to explain it in a reasonable way and also to leverage their needs to what DATA can provide for them.

DATA!!....great, and now what’s for me as an Advertiser?

The quote; “if you cannot explain it simple then you really don’t know” suddenly got me very inspired :-)

How can I explain in a simple manner this “DATA jungle” that most marketers are dying to learn more about, and on top make them understand it?

First of all, it’s important that we accept that we are in the infancy of DATA management related to programmatic MEDIA buying, and that the whole thing also gets more complex because the industry is constantly changing due to technical and jurisdictional evolution.

To me there are actually only three basic things that need to be learned. 

  1. How DATA is collected
  2. How DATA is stored
  3. How DATA is applied in media buying

I will try to make it as simple as possible so that any person can understand (I assume that reader is in the marketing & advertising industry).

First of all, how is DATA collected?

There are many DATA sources where we can collect interesting user behavior information, but it’s important to leverage expectations and that we accept that today (May 2015) the most important and relevant DATA in relation to programmatic MEDIA buying is through Cookies that are stored in our browsers! (Other stuff is interesting but still limited in the programmatic MEDIA landscape).

How do we as a user know that a Cookie is activated in our browsers? By law, companies usually need to make sure that the users are aware that they are exposed with a cookie (either by the web or 3rd party technologies). It appears with a text on the top or bottom of a web page. It basically says that you are being cookie traced and that the owner of the content is allowed to use this data to improve your navigation etc.

There are programs such as ghostery.com where you easily can see what sort of DATA that is collected and through what Technologies.

There are other sources of DATA collection (non cookie based) such as mobile APPs but there is still a lot to be learned on the aspect of relating devise ID to cookies and make it useful for a DMP. There are companies specialized in unifying this information and will be a reality but as an advertiser it’s important that they understand that it’s not that simple and there is still some things that need to be figured out. (The mobile and APP penetration is so high that its really important that this piece of the puzzle gets solved as soon as possible)

For more info on Cross Devise Technologies:

Adform GO

Offline DATA is also a challenge to collect and make usable in a programmatic environment. A CRM has got a lot of information but if its not browser based then again it’s a challenge and will need engineers to pull it into something useful.

There are other DATA entry points to consider such as analytics and any type of tracking tools (Adservers, Attribution trackers, Analytics). These are all technologies that makes it possible to activate cookies for DATA collection (and enabled for a DMP) and is extremely useful.

Also Web based Emails or CMS are great sources to collect DATA.

Cool, so now we have collected the DATA in a “DATA warehouse” (DMP). It’s usually a massive DATA warehouse containing cookies exposed on millions of impressions where pixels are fired from left and right and that make absolutely NO sense at all :-)

So how do we store the DATA?

This is when the DATA scientist works out a plan with the Advertiser to set up a defined taxonomy system.

NOTE: when deciding on a DMP partner, make sure that they apply the right usage in terms of safety & privacy concerning DATA.

This classification in different groups are called Clusters.

Examples:

LEVEL 1 : Demographics
LEVEL 2: Age
LEVEL 3 :13-17

LEVEL 1 : INTENT
LEVEL 2: TRAVEL
LEVEL 3 : AIRLINES

LEVEL 1 : INTEREST
LEVEL 2 : HOME AND FAMILY
LEVEL 3 : PARENTING

This process of bringing user behavior into different clusters is the most important task in DATA storage.

There is a more advanced technique in DATA storage and that is to create “look-a-like” assumptions and to work on clever algorithm.

Example: A pixel is fired on a users navigating on a page that talks about safety. This cookie can be classified in a group as "NON-risk takers". Could potentially be brilliant information as an advertiser that needs to promote a product that assure 100% guarantee in safety. Who are concerned about safety? Yes, NON-risk takers!

Again, I point out an important thing. There are very smart guys working on this, but things are still in the development phase. Nevertheless, the one that first starts to store the DATA correctly and to get things organized will be the one that obtains the greatest benefits. In other words, the laggers always get to eat the last piece….

Great, now we have collected DATA, its stored so now what?

The DATA collected in a DMP for an advertiser can be used in 3 ways.

1) Improve the navigation experience for your users.

For instance, why should a Spanish user entering an airline company be addressed in English? It could be that some Spanish users prefer to read the content in English but the majority does not. Also when registered for an airline companies login page its actually nice that your name appears or that you get offered with cheep prices on routes that you usually fly. There are many examples on how to improve things based on user DATA.

2) Advertisement Campaigns

Targeting your 1st party DATA!

With the cookie information stored we can retarget the users with messages that is related to the last navigation in the page. It could be a person that searched an airline ticket with the destination to Rome. From a media buying perspective this is brilliant. We can expose the same user with the same destination but with some additional information. We can also exclude these users because they might already have paid the ticket. Why waist money, exposing this user again?

NOTE:

  • The users being exposed on an ad-campaign and then targeted in a second campaign is referred as Re-Targeting.
  • The users that is already considered ours (1st party), that we target in campaigns are called 1st party Targeting.

In campaigns we can target not only specific users but also groups of thousands of people. Lets say that we have a cluster of Male users interested in sports. Would it not be fantastic to run a campaign during a few days directed to Male users if there is a very important sports event in a mayor city? Fly to Berlin to see the Championship game and experience the atmosphere!! This could improve substantially chances to sell airline tickets.

Target 3rd party DATA via a DSP

The official DSPs have access to sophisticated targeting options based on DATA that is used for PROSPECTING campaigns (users that are not yet customers). They usually follow the standard taxonomy systems and come from the supply side of advertising (publishers). Lets say that an airline company is offering a special business class offering. It’s easy to program the campaign to users that matches the criteria “business” or “high standard” demand through channel, or contextual targeting options. This has not too much to do with an Advertisers DMP but worth mentioning.

The 3rd party external DATA can actually be imported to an advertiser DMP in order to do direct targeting. This could be done by 3rd party DMP players or striking a deal directly with a publisher to import DATA (2nd party DATA). An example of this would be if an airline company strikes a partnership deal with a tourist guide for luxury traveling. These cookies in a DMP would be classified as Business class targets that would be identified when purchasing media through a DSP. 

NOTE: the most sophisticated DSPs have partnerships with 3rd party DATA providers and can be purchased directly from the buying platforms. 

Cookie extension though sophisticated testing’s and assumptions can also be a useful technique to drive relevant traffic to an advertisers site. There might be the assumption that the people that searched an airline ticket with the option of an infant of 0-2 years has a higher likelihood to rent a family car at its destination. It could be interesting to allocate these users and expose special advertising campaigns for these users.


All in all when using a DMP in a smart way we can get collateral benefits such as: 

  • Minimize budget waste
  • Discover New audience segments
  • Improve response rates & brand recognition
  • Cross devise targeting to avoid duplications to same users


3) DATA as an information center.

Usually in relations to campaigns the attribution tracking has the best view on what is the real benefit from different advertising channels and the analytics to understand what is going on related to the site in terms of user behavior (have a look at the Article, Paradox, Performance OR Branding). A DMP and the evolution of clusters based on taxonomy also give marketers information of great importance. On the dashboards one can see how clusters evolve around consumer INTEREST and INTENT for specific products. 

One important thing to mention is how valid PROSPECTING campaigns are. If we do not expose new users to the advertisers site, DATA will remain same and insignificant. Therefore when running a campaign through programmatic media, if its not attached with a DMP the info will not be as rich. Its like running like crazy to find people on the street make them entering your store and then never ask them what they need.

Therefore programmatic advertising campaigns applying DATA have two main purposes. To gain RESULTS & LEARNING.

So back to my challenge of explaining what is a DMP and how does it help an advertiser in a simple way, I will try to turn it into one sentence (or two):

“A DMP is a platform where an advertiser store DATA on user behaviour (could be from his own users, (1st party) or from external sources (3rd party)). This DATA can be used to improve navigation in your site, extend targeting options in your advertising campaigns (plus many other campaign benefits) and will provide you with valuable DATA information related to user behavior”

The more DATA you have, the better it is classified the more you learn and the more you learn the better your campaign RESULTS will be.

I actually still hope for a more simplified explanation but I tried my best :-) if you are interested I will be more than happy to offer you a DEMO of a DMP platform that I highly recommend (both for publishers and advertisers). I could also arrange a meeting with a DATA specialist that could figure out the best way to work your  DATA sources in order to make your Digital advertising campaigns better than your competitor.  

andreas@folq.biz, follow me @folQ_DigitalAds,

 

 

 

 

Social

View Andreas Akesson's profile on LinkedIn