getting started with identity resolution

The era of mass marketing is long over. Marketers today are looking to understand their customers at the individual level and respond to them with relevant information or experiences.

But in a multi-device world where one customer can interact with a company multiple times across multiple channels within one single journey, understanding the identity of each individual is essential.

To address this topic, we recently hosted a webinar entitled “Understanding Customer Identity Across Channels,” presented by Carter Bradford, President of Precocity, and Matt Thompson, Senior Director of Partnerships at Evergage. In the webinar, Carter provided this definition of identity resolution:

Identity Resolution is the process of connecting all consumer identifiers, both online and offline, to a single individual.

Be sure to watch the full webinar replay to learn more about what identity resolution is, the challenges marketers face as they undertake it, and how to address those challenges.

In this blog post, I want to outline a few of the steps Carter laid out to get started with identity resolution.

1. Develop your overall customer data strategy

Before you get started, you need to figure out your strategy. Each company is different, so there’s no single strategy that Carter or Matt recommend.

You can start by getting relevant parties in a room and asking a lot of questions. What data sources do we have? What technologies do we currently have that collect and store customer data? What does our current view of identity look like? What identifiers do we have? What do our current customer data processes look like? Who is involved? What do we ultimately want to do with our customer data? How are our current technologies handling our data and what are the gaps between our current state and our ideal state?

As you ask and answer questions like these, you’ll have a better idea of what your needs are. Then you can start outlining a path forward to address those needs.

2. Identify a handful of key use cases

Once you’ve got your strategy laid out, start getting more specific. Outline several key areas in which you plan to use your view of individual identity so that you can start with your end goal in mind.

These use cases will likely be examples of personalization in action — the act of tailoring an experience or communication based on information a company has learned about an individual. You don’t just pull all of your data together for fun, you do it so that you can use it to provide a better experience for your customers.

By laying out the ways in which you plan to provide that better experience in advance, you can make sure you’re headed in the right direction.

The webinar offered some great examples of use cases you may want to consider, so be sure to check it out for some initial ideas.

3. Start with deterministic matching

As described in the webinar, when creating a single picture of a person’s identity, there will be a certain amount of stitching and merging profiles together as data is brought in from multiple sources and customers interact with you with or without identifying themselves. There are two main ways to stitch profiles: deterministic and probabilistic matching.

Carter recommends you start with deterministic matching, as he considers it table stakes for any identity resolution program. With deterministic matching, a system stitches profiles together based on clear, unique identifiers such as system ID or email address. For example, if a person visits a site multiple times but never identifies himself by making a purchase, registering for a webinar, signing up for a newsletter, etc., his profile remains anonymous. The first time he does provide his email address, the system recognizes all of his past sessions that took place with the same cookie, and stitches his anonymous profile together with his new identified profile.

Once you have deterministic matching under your belt, you can begin considering probabilistic matching, which has its pros and cons. The webinar offers more detail on both approaches.

4. Understand and weight identity signals

An identifier is a data point that allows you to recognize a specific person. You likely have a number of identifiers within your organization — some more valuable or more accurate than others. You may have cookies, name, loyalty ID, email address, physical address, etc.

You need to assess how you weight each of these identifiers and how you’ll tell your technology (often a CDP) how to react to each one.

For example, a phone number collected at the register may not always be accurate. In the webinar, Matt and Carter told the story about how they often hear that store associates type in 555-5555 when asked for a phone number just to keep the check-out line moving. If that’s the case for you, you don’t want to end up with hundreds of customers stored under one profile based on that one incorrect phone number.

In contrast, a physical address entered by a customer for an online order is likely to be quite accurate, since the customer wants to enter the correct address to ensure her package arrives to the right place. However, a physical address can be shared by multiple individuals, so it’s not always clear as to whether all data belongs to one person, or multiple people within the same household.

Each company has its own considerations, so think through your identifiers and how you will value each one going forward.

5. Carefully consider compliance

When dealing with identity, compliance is an area you can’t ignore. Before you get too far into this process, make sure you’re well-versed on the laws (e.g. GDPR) and general ethical considerations around using customer data.

There are a lot of different ways you can use customer data, but as Carter said in the webinar, “just because you can doesn’t mean you should.” Always keep the end customer experience in mind and don’t do anything with a customer’s data that doesn’t ultimately improve that person’s experience.

6. Define success criteria and key metrics, continuously evolve and refine

As with any business initiative, you need to carefully consider how you will measure success. Track these metrics over time to judge the effectiveness of your program and make changes to improve. For example, you can weight identifiers differently, address data quality concerns, tweak personalization campaigns, etc. and note the impact those changes have on your KPIs to improve your effectiveness.

7. Carefully consider tools

Finally, you need to be sure you’re using the right tools to make all of this happen. Keeping track of individual identity and responding to each person in a relevant way cannot be done at scale without technology.

Each company’s needs are unique, so each company needs a unique combination of tools to accomplish their goals. When looking at companies that will be managing your customer data, you want to look at those who will be long-term partners for you. So don’t take this step lightly.

Final Thoughts

In the webinar, Carter said that “Identity is the foundation of modern marketing.

This is because, as customers demand personalized experiences, we need to be sure that we know them well enough to speak to them as individual people. If you try to personalize an experience based on a flawed understanding of a person’s identity, you’ll have seriously missed the mark and delivered a very bad experience.