Good companies care about their customers. But to really understand their wants and needs, brands should know who their customers are. Sounds simple, right?
No such luck. The key first step in understanding customers is the process of identity resolution, and while any number of vendors boast they’ve solved this process it’s pretty clear most don’t provide any richer data than a Census report circa 1950. And that’s left marketers more confused than ever.
What is Identity Resolution?
To get marketing right, you need to get your data in good shape to feed your tools and campaigns, and the first step in getting the data right is identity resolution.
Identity resolution entails matching and bringing together all the bits of data associated with a given customer to be able to say “these all relate to this person; this is what we know about this person.”
And why is this important? Because it affects the quality of all marketing and outreach. Who you market to and how you connect with them is all dependent on the quality of the data input.
Getting identity resolution right powers personalization and revenue growth while saving you on marketing costs by not sending misplaced communications to the wrong person. It’s the critical first step to more informed marketing and experiences.
But this isn’t simple to pull off.
Most companies interact with their customers or collect information about them digitally. And that transformation between people in the real world and all their digital records and profiles across all the different tools that brands use don’t always line up very cleanly because:
- the data sources don’t integrate properly
- the information is out of date
- the data was entered incorrectly in the first place
How identity resolution is changing
In the 1990s, companies struggled to collect data, operating off any digital data they could find (like email lists and public records off the web). Then came the tedious task of trying to match the customer-supplied data against that.
In the 2000s and 2010s, the explosion of ecommerce and new channels suddenly made it much easier to collect customer data, but that resulted in the data being an even messier tangle. First-party data (consent-based data given directly to brands by consumers) was difficult to unify because the data came in from different tools and in different formats that weren’t designed to link up, and the tools available couldn’t account for the complexity. At the same time, third-party identifiers collected from vendors were able to fill the gaps and give a functional picture of the customer.
Today, the deterioration of the third-party cookies, mobile ad IDs (maids), and location data has triggered a complete reboot of digital marketing and a renewed focus on solving identity resolution. The focus is now squarely on first-party data since it’s more accurate, more economical, and privacy compliant. But for brands to work with their own data they need to make sense of the data mess through identity resolution and data unification so that they can put the data to work.
Customer data is burying us all
Customers provide businesses with a wealth of information, but their Personal Identifiable Information (PII) often doesn’t add up to a complete or accurate picture because it’s scattered across systems that don’t talk to each other.
Data is messy. And customers with the messiest data are usually those who’ve been customers the longest or who interact the most, and therefore are the most valuable. This is a key consideration in a tighter budget environment where retention efforts are getting more focus than acquisition.
Understanding your customer identities can completely change how you market (segment by tiers, categories) and where you market (channels) to provide more meaningful and personalized experiences that customers expect (and earn their loyalty).
Avoiding the void (in-house builds)
Some data analytics companies tell you it’s possible (even easy) to design, build, and manage your own identity resolution solutions. Here’s the thing: That’s not exactly true.
Common pitfalls of doing it yourself:
- The typical methods don’t work. Deterministic-only approaches can erode data quality. Rules-based approaches to matching data cannot account for less-than-perfect matches and can lead to duplicate records and incorrect merges. Both methods have a hard time adjusting to changes and updates in customer data so the profile isn’t up-to-date.
- Jobs are expensive to run. Running matches through multiple systems can quickly rack up costs and subscription overage penalties.
- The effort isn’t worth it. Some companies spend months trying to do what could be done in hours.
(And these are just a few – there are other common problems too.)
Sophisticated solutions that help
An ideal identity resolution technology should empower you to get identities right before activating customer data, so you can:
- Lower your marketing costs
- Increase your marketing efficiency ratio
- Increase your ROI by using high quality data
The bottom line
If you really want to know your customers and deliver the kinds of experiences they appreciate (and expect), forgo the Census data. And forget about unrealistic homegrown builds that’ll have your data scientists and CFO cursing in the halls.
In a nutshell:
Going it alone with identity resolution will only give you heartburn. Building a solution in-house requires massive investments in infrastructure, storage, SAAS platforms, and data scientists, but even with the substantial resources, 87% of efforts fail.
AI and machine learning technology that includes deterministic matching at scale will help you establish an unrivaled customer foundation that is accurate and error-free.
The right technology will also facilitate the creation of detailed profiles that are accessible across your organization and informed by historical, transactional, and behavioral data to paint a picture of every customer’s complete experience.
