Welcome to our blog series on decoding identity resolution. This is a nine part blog that offers an attempt at a friendly, comprehensive view of how to think about the concept of identity resolution as well as how to interpret the way it is represented in marketing and sales materials by different companies across the tech landscape. The other articles in the series can be found here:
Part Two: The Value of Identity Resolution
Introduction
What is all the fuss about identity resolution? Why is it so important to a business?
This entry focuses on the business value of high quality identity resolution. Let’s spend some time digging into why a “good enough” solution isn’t actually good enough.
The costs of bad identity
Let’s say your company spends $10M a year sending emails to customers.
The lists of customers you target are built on the data management strategy you employ. Somewhere along the way a particular process is creating lists of people to target.
Looking at Amperity’s wide range of customer data sets, an average duplication rate we see in a reasonably well managed customer data set is about 7%. This means that 7% of customers have more than one record assigned to them.
Without accurately deduplicating, around 7% of the time marketing is sending two emails when they only need to send one.
7% of $10M is $700k. Most marketing activation providers charge by volume — so almost a million dollars a year is wasted, and that’s just the tip of the iceberg.
The opportunity cost of inaccurate data
The impact of marketing campaigns is directly limited by the quality of the data used to deploy them. This is a fancy way of repeating the industry adage, “garbage in, garbage out.”
High-quality identity resolution and data management practices can make sure you're maximizing open rates and clicks by providing the most relevant marketing to customers.
For example:
Jane buys from your company regularly.
Jane gets married and decides to change her last name.
Next time Jane buys, an overly simple identity resolution strategy leads to her getting an entirely new profile in your ecommerce platform.
Jane continues to buy, but your brand sees two Janes. One that churned, and the other that is new.
This is an incredibly common scenario. When you market to Jane you should see one high engagement customer, not two. You might waste money trying to win back the “churned” Jane, providing irrelevant marketing. You might not recommend relevant products to the new profile.
The most profitable customers often produce the most data for a business. By not identifying them correctly, brands miss out on some of the biggest revenue opportunities. More worryingly, the poor experiences brands provide because of duplication may actually cause customers to churn in the future.
The limitations of incomplete data
Another common problem crops up when tools are specialized in a specific type of data, instead of being able to handle all the different types of data a customer can produce.
If you choose a platform that’s heavily biased toward online experiences, you run the risk of retaining unintentional duplicates at a drastically higher rate. These platforms are often incompatible with other channels of data from legacy systems like point of sale data or loyalty data because everything is “online first”. The same is true in the opposite direction, if you choose a platform that has a bias toward loyalty data or in-store data.
An enterprise-grade data management platform should be able to store and make use of data from any channel. If not, it’s too specialized to be transformative for a business, reinforcing silos instead of knocking them down.
Summarizing
In short:
Most marketing activation channels charge by volume so having bad identity resolution wastes money in the form of duplicate marketing.
Overly simple identity resolution makes it impossible to accurately understand who the most valuable customers are, leading to bad customer experiences, incorrect analytics, and inaccurate personalization.
The marketing technology landscape is cluttered with SaaS offerings that overly specialize in a subsection of data. Adopt a strategy that can handle all your data, otherwise, it will just reinforce silos and not truly solve the problem.
A great identity resolution strategy takes messy data and turns it into value — forget about “garbage in, garbage out;” it should be “garbage in, gold out.”
Next up
Next we cover PII identity resolution specifically.