True Cross-Device & Multi-Channel Measurement The Holy Grail of Online Advertising

To perform great marketing, ideally, we would have perfect data on the humans that interact with our marketing touch points. Unfortunately, when it comes to traditional offline marketing (TV Ads, a Billboard, etc.) we’re guestimating most of the time. Especially when we’re trying to actually attribute credits toward specific billboards and specific tv-ads at specific air times.

A lot of marketers assume that online marketing gives us perfect data about our campaigns and how potential customers interact with them. This is what we would like to be true. It’s also what most analytics and advertising technology vendors want us to believe.

It is not uncommon to see analysts performing advanced channel attribution analysis based on a standard web analytics dataset. They take all available data at face value and assume that the metric “Users” actually refers to “human users.” Potentially huge mistake, especially when the analysis goes up the chain and the context of the collected data gets lost along the way.

What we need is to have full insights in the entire customer journey that spans multiple devices and uses multiple marketing channels before it converts into a client. Why don’t we have that?

Current Limitations

It is important to understand the limitations of the current state of online marketing and measurement. These limitations are a combination of privacy laws protecting consumers and technical challenges in measurement technology.

Cookies

To remember what language preference your browser had, or what filters you like to use on a website, developers place small text files within your browser. These are called cookies. This way you don’t have to re-select all those settings every time you come back to the website.

Marketers also found out about these cookies. We use them to store information about you and your behavior. This way we can adjust our way of communicating with you through advertising and website personalisation.

The most important thing to understand about cookies is that they are always on the “client-device-browser” side.

  • Client side: meaning the information is stored on the consumers end.
  • Device: meaning the information is separated by device.
  • Browser: meaning the information is separated by browser.

Example: A person visits your website with his laptop, tablet, and smartphone. By default, you will see three (3) unique users in your analytics and advertising tools.

If this user uses the browsers Chrome and Safari randomly on all three devices, you might even see six (6) users.

If the user deletes their cookies (or has a privacy plugin), every new session creates a new unique user. Even if they use the same browser on the same device!

First Party User ID’s

If your website has a form of logging in available, we’re in luck. This action allows us to manually generate a user ID that is unique for the login. With a bit of extra effort, we can send this user ID towards our analytics and advertising tools. This allows these tools to use our User ID to identify “Users”, instead of unique cookies. We call this “first-party data.”

Using this “User ID” will enable us to tie together all browsers on all devices a user uses to log in to our website. This way our analytics tool can tie all devices together under the same user. You should always implement this and rely on your own user identification when possible.

There is one caveat, registering or logging in is something that usually takes place AFTER the users first purchase or signup. Since the bulk of your online marketing spend is most likely targeted at new user acquisition, this will not solve your problem of cross-device attribution of the customer journey before the first sale.

Retroactive User / Session Stitching

Let’s assume that every user that becomes a client needs to register with your company. And let’s assume that you can make all new clients log in with all their devices after they become a customer.

In this case, you would be able to tie all separated sessions with unique cookies together to this one user id. Assuming the user did not clear any cookies in the meantime. (And assuming you have a reliable analytics implementation.)

There is one problem, this requires a retroactive action to take place upon your data. Usually, it takes a few sessions before users register. Only upon registration, we are able to create a user id for this user. After that, we need to identify all their unique browser cookies and tie them together with the user id as soon as they log in. If we want to view a “cross-device” customer journey, we need to go back in our data and reattribute these sessions to this user.

This is something most analytics tools don’t support out of the box for a good reason. The result would be that all reports containing the “Users” metric would never be definitive, which makes it unsuitable for dashboarding and reporting. (“Hey! Why did our total amount of users for last month drop all of a sudden?!”)

When doing deep-dive analysis this technique works fine, especially if you’re analyzing historical data. But we’re not talking about doing deep-dive analysis after the fact. We’re talking about data-driven marketing!

To perform data-driven marketing, we need this information available at all times for every marketer.

A Technology Partner with Accurate First-Party Data

Right now, the only viable option is to ‘lease’ more deterministic first-party data. You need a marketing stack that can match cookies and device ids to users in near real-time.

It would have to be their first party data gathered through user logins. Additionally, it would have to be as fresh as possible and continuously updated as users change devices and browsers.

You’re looking for a company that can identify all consumer devices on a daily basis, no matter how many times we delete our cookies. A company that has a lot of widely used mobile apps to identify smartphones, tied together on the same login they use on their laptop and browsers for web services.

Additionally, this company needs to provide you with a marketing technology stack to access this data. This stack needs to enable you to perform marketing towards “humans” instead of “cookies.”

The technology to create “single users” from “multiple cookies” is not that complicated. Most marketing stacks have it. It’s the quality of the first-party data to help you identify a user’s different devices from that company which is key.

Most important, especially if you’re a European based company, you need to be able to use the data without actually infringing on the user’s privacy. This means that within your marketing stack, you should always be able to see the “actual human users” based on the first party data they have. But you should never be able to extract that data in any form or be able to match it to personally identifiable information from your CRM.

So now we’re just looking for that company combining both ;-)…

 

Rick Dronkers

I love data, marketing & technology and work with all 3 on a daily basis as the Head of Analytics at a dutch Online Marketing Agency called Maxlead. Views expressed online are my own unless stated otherwise.