Leveraging second party data in AAM
15 Jan 2017 » AAM
In an Adobe Audience Manager implementation, the first and most important data source is the data you already own. Then, when no more juice can be squeezed from first party data, we switch to purchasing third party data. Finally, in some cases, we go beyond and look for second party data. Today, I will focus on this last resort, which can be more interesting than what it initially looks like.
Definitions of 1st, 2nd and 3rd party data
The first time you read or hear about these data sources names, you wonder what they really are. At least, this happened to me when I started consulting for AAM. Let me try to explain each of them:
First party data. This is data you own. You do not need to request permission to use it outside of your company. Typically, there are two types of first party data:
- Behavioural data. This is the data you capture from your web analytics or data layer.
- Offline data. Any data you have in your databases, like CRM data or loyalty card information.
- Third party data. Some companies specialise in gathering (anonymised) data from various sources and package it. For a fee, anybody can use it and combine it with first party data. There are usually some limitations on how you can use that data, specified in the contract. Adobe offers a marketplace to exchange this data.
- Second party data. You will have noticed that I have changed the typical order. The reason is that it second party data is less common and more difficult to explain. The idea is to get data from a partner, with which you have a special relationship: a partner, an exclusivity contract…
Both second and third party data are data sources you do not own, but you are granted the rights to use them. The main difference between the two is that, second party data is usually only shared with you, whereas third party data is offered to the market and anybody can purchase it, like your competitors. In other words, you can make sure that any second party data you use is not shared with anybody else, but you have no control over third party data. This difference is very important if you want to differentiate from your competition.
Second party data use cases
OK, we have some nice definitions. The next question would be, now what? I must admit that 2nd party data is not for everybody. For example, banks will be very careful with the data they ingest and will probably not want data they cannot control. But there are many other cases when it can be really useful and, in some cases, the only solution to use a DMP.
Mutual benefit: data exchange
This is the most typical case. The idea is that two companies reach an agreement and each party shares a certain amount of data in exchange of data from the other party, so the cost is zero. Generally speaking, these two companies do not compete with each other. With this extra data, marketers can better tailor the messaging to certain audiences. The most recent example I have seen is a mobile operator and a mobile phones manufacturer. These are two non-competing companies and both can benefit from this data exchange.
One side benefits
This is the case when one company sends data to another, but there is no counterpart to this data flow. I would assume there is some kind of monetary compensation, but from the DMP perspective, this is not important nor necessary.
One example I worked for was a cable operator and a number of comparison websites. The cable operator was paying these comparison websites to show its offers and include them in the comparison results. At some point, it was also requested to add an AAM pixel to some pages in the comparison websites, to get additional behavioural information. This information can then be used later, when the user lands on the cable operator’s website, to tailor the messaging knowing what the user has been on the comparison website.
Main data source
There is another use case for companies, which do not have good first party data. In this case, they need to look for other data sources. I find this the most interesting use case of the three.
The best example I have worked with is a CPG. It would be naive to believe that consumers of breakfast cereals, chocolate bars or hair shampoo will go to the manufacturer’s website and open an account, let alone to comment about the products. These companies always sell through supermarkets and other channels. Therefore, CPG websites receive very low traffic and of poor quality. If they want to create audiences, they need to find other means to be able to create audiences.
The solution entails sponsoring a related website and getting behavioural data from the sponsored website. Obviously, this website needs to be properly chosen, knowing that the visitors are of quality. These are some examples I know of:
- A global CPG sponsored a section in a media website targeting Millennials.
- Another CPG signed a partnership with a popular website for recipes and cooking.
- A retail bank sponsored a credit card.
- A content producer signed an agreement with an MVPD to get the users’ behavioural data from the MVPD.
In all these cases, part of the deal was to be able to add AAM code to some parts of the website of the sponsored website, on behalf of the sponsoring company. With this information, the sponsoring company can create audiences and target them.