What do you do when you know third-party cookies are going away? You double down on your first-party data and build the customer experiences your customers deserve. Or that’s the theory. How you do that is the larger question.
Katrina Wong, VP Marketing for Segment, a business unit of Twilio, joined me on the Content Matters podcast, and we had a good discussion around the demise of third-party cookies and what brands need to do to move forward.
Wong talked about how privacy laws have transformed the way companies look at how they use third-party data and how they think about identity resolution. She argued that companies are tackling these questions with technology - specifically Customer Data Platforms (CDPs). The challenge is very much about getting a handle on first-party data and activating a single view of the customer. Wong also posits that you that can create relevant messages and personalized experiences without third-party data.
The many types of customer data
Picking up on that idea, we talked through a brief overview of the different types of customer data:
- Third-party data is collected by data providers and then sold to companies. Therefore, there is no direct party relationship.
- Second-party data is collected by a brand's partners and then shared with the brand. This is the partner's first-party data that they either sell to the brand or have some other sharing relationship.
- First-party data is collected directly by the brand through interactions with customers when the brand provides something of value. Wong said this also could include insights from analytics and user behavior.
- Zero-party data is a term that was coined by Forrester Research. Customers intentionally provide this data through actions such as setting preferences. The difference between zero and first-party data is that zero-party data solicits direct interaction with the audience.
Historically, companies have struggled to collect and unify first-party data and signals, which is why, suggested Wong, third-party data is so widely used. A CDP allows companies to unify the customer record by bringing together more of the first-party data. And it's not just collecting and unifying data. It's also cleaning, controlling, and merging it into a unified profile the company can use to activate customer experience in real-time:
All of those touchpoints can tie together what a unique user is doing on your particular website, and therefore you can then message to them in a very tailored way. That's definitely the new way of thinking about how you would create hyper-personalized customer experiences. And you can do so in a better way than using third-party data, which isn't very accurate at all.
You could try to unify your customer data without a CDP - companies are doing it through custom development- , but it's hard to do. There are too many data silos across a company that you need to bring together. Reconciling them manually, especially at scale and in real-time, is almost impossible, concluded Wong.
Building blocks and the API economy
A CDP does something else. It helps companies clean up their tech stack. The API economy has changed the way companies build and integrate their technologies. According to Wong, it has democratized everything for marketers, giving them access to business logic that was previously locked in applications that never did exactly what they needed to do:
I recently heard this from one of our customers, and it resonated with me. There are a lot of marketing tools and technologies out there, but we really don't have the data. When you give marketers building blocks and these APIs where developers naturally build for any marketing use case, it's the perfect marriage.
Modular development isn't really new. Service-Oriented Architecture (SOA) has been around for many years. But an SOA approach is very much about all custom development. Building with APIs is about modular development, but not from scratch.
Wong referred to the ‘build or die’ mantra of Twilio CEO Jeff Lawson. This says companies need to build custom solutions to differentiate, but they do it with building blocks as opposed to building from scratch. It's been an evolution - developing from scratch to buying packaged software that can't keep up with the pace of change and what customers are doing. The answer is a hybrid model, where you are not building from scratch but working with building blocks, tailoring to specific use cases, said Wong, noting that Twilio, and Segment were both founded on this philosophy.
With a philosophy that emphasizes building blocks and APIs, developers are not surprisingly vital to the company's story. At the same time, Segment also provides products tailored to marketers that go beyond what developers are doing. The recent launch of Journeys is an example where a marketer can build a custom journey in Segment, leveraging the data in the CDP. Wong explained:
So marketers have access to a lot of technology. What they don't have direct access to is data and good data. Data that has been cleaned and unified into this single view of the customer. And what's powerful about connecting a CDP with actual tangible tools that marketers can use, such as building a customer journey, is the next era - is being able to connect the two.
Without that great data, accurate data, compliant data, you're not going to be able to deliver the right message at the right time across the right channel for your customers.
Good data is foundational to great customer experience
This means we can now argue that a CDP is foundational technology. But not all CDPs are the same, so how does a company know what CDP will work best for them? Wong said the holy grail is combining the data infrastructure and architecture and having clean, compliant data that marketers have direct access to so they can run campaigns directly from the CDP. As a result, a CDP can help streamline the marketing stack, she said.
One key component of any CDP is going to be predictive analytics. We haven't seen a lot of success stories yet, which Wong attributes to user still being in the early stages of Machine Learning (ML) and Artificial Intelligence (AI) - she compared it to being in the second or third inning of a baseball game. However, she said Segment has seen some customers doing innovative things like stitching together experiences across channels.
As a case in point, she cited a major shoe company that is offering a true omnichannel experience that is hyper-personalized:
For example, if you see a pair of shoes on a Netflix show, you can now go on the mobile app of this brand, with a huge community, and basically send a message in real-time to a brand ambassador, and actually chat about the new release of the shoe.
Predictive is a journey, Wong said, and it takes a lot of data to get it right.
Not every company has a CDP, but adoption and implementation is growing, especially among enterprise companies. As to the direction of travel, from Segment's point of view, the first step is going heavy with the data, architecture, and infrastructure and then giving marketers the tools to create journeys and orchestrate the customer experience. The next step is executing it all within a single platform and not having to move the data into another platform, Wong concluded:
So that we get the end-to-end experience and the real signals feeding back to your customer profile in your CDP.
Twilio Segment has been adding many new capabilities to help both marketers and developers leverage the data managed in the CDP. Along with the new Journeys tool mentioned above, the company has also announced a set of CDP building blocks in a new Developer Toolkit.
Are CDPs a critical component of the tech stack? I think so. Pulling together all that customer data is complicated and time-consuming, and when you need to build the right experiences at the right time, you need technology and tools to make that happen fast.
Third-party cookies and third-party data will eventually go away. Some brands and tech companies are already moving away from them and finding success. For most, it's just a matter of time.