Identity resolution can be a confusing concept to get your head around, with so many related terms such as identity graphs, probabilistic matching, deterministic matching and unique identifiers to understand. In this blog, we’ll look to simplify what identity resolution is and how marketing teams can use it to enable more accurate marketing and personalization strategies.
Why is ID resolution important in marketing?
It’s pretty certain that, at some point, you’ll have been to a party or social event where you have met lots of different and interesting people and remembered some really clearly, perhaps because of a shared interest or inspiring conversation, while others failed to make an impression. After all, no one has a 100% perfect memory, right?
Fast forward to the next time you attend one of these functions or events: what happens when you bump into the person you forgot? You're aware you know them from somewhere… you just can’t place them. Inverting the situation, what happens if you are the person that has been forgotten? How do you feel if you remember someone perfectly and they immediately address you with the wrong name and ask after your kids, by name, except they’re not your kids, you don’t even have kids?
This scenario, and others like it, are what your customers are subjected to when you cannot personalize their experience of your brand. From inaccurate product recommendations, to using the wrong name when suggesting they sign into the app or website, the impact of these personalized experiences, positive or negative, directly correlates with the customer’s perception of your overall customer experience.
In our hyper-connected world, the demand for personalized experiences has risen exponentially, as demonstrated by a Gartner survey that found 62% of customers will punish a brand for irrelevant or inaccurate communications. This breaks down into 48% of customers saying they will unsubscribe from brand communications, while 14% pledge not to do business with that brand again. The latter figure may not sound all that high, but what would your bottom line look like if the top 14% of your customers stopped buying from you? Identity resolution helps to build a Single Customer View and a memory of all purchases and interactions to prevent that from happening!
Identity resolution definition
How are personalized experiences related to identity resolution? And how can brands avoid these types of customer ‘punishment’? Forrester defines identify resolution as:
“The process of integrating identifiers across available touchpoints and devices with behaviour, transaction, and contextual information into a cohesive and addressable consumer profile for marketing analysis, orchestration, and delivery.”
So, identity resolution is all about that accurate customer profile, including how it is created and how it underpins your specific use cases, which often include better personalization and an enhanced brand experience.
The challenges of ID resolution
How can identities be resolved? Well, if you only had a few data points and a couple of customers, you could quite easily manage this in a spreadsheet tool. You could take all of your data – email addresses from the ESP, phone numbers and addresses from your CRM system, mobile details from surveys and device IDs from your website – and use your knowledge of the data and pattern recognition skills to establish links between the data points. After all, the aim is simply to try and link multiple data points, in order to better understand the device or person that they relate to.
An example might be if you link a list of customers with the items that they have purchased over time. Your customers come from the CRM, whilst the purchases come from an eCommerce platform. Once linked, that is once the identities have been resolved, you can then understand what subsequent products you could promote to each customer, based on their purchase history.
But, how does this approach scale up when you’re dealing with hundreds of thousands, or even millions of customers, thousands of products, and potentially hundreds of millions, or even billions of data points? Many businesses are looking to answer that exact question, and technology can help to provide the answers, either in the form of identity graphs, CDPs or other data management solutions.
What advantages can identity resolution deliver?
ID resolution can deliver substantial benefits and open up many possibilities for marketers when done ‘correctly’, i.e. in a reliable, trustworthy, compliant and accurate manner. These can include:
- Improved database management, providing foundational accuracy for all customer-centric business decisions.
- Removal of some of the risks around poor data management
- Improved data compliance.
- The linking of customers’ online and offline activities, allowing you to better understand the full customer journey.
- Removal of continual manual efforts within the business to link siloed databases.
- Enhanced targeting/segmentation, which in turn informs more accurate, timely and relevant personalization.
Types of identity resolution
There are two main types of ID Resolution, known as probabilistic and deterministic. Probabilistic matching uses probability in the form of algorithms to link various data points together. The process has room for error, since it looks for a commonality within the pool of data points and then, using a matching threshold percentage, collects together records it thinks may belong together. The match is not 100% definite and therefore can be problematic for multichannel marketing, but for anonymous personalization these tactics can be very handy indeed. In contrast, deterministic matching is focused on accurately determining a 100% match between two or more data points, especially PII, without resorting to probability calculations and the like.
Probabilistic matching is great if you want to give yourself reach, but you will have to be comfortable with the lower level of accuracy the approach provides. So, when the benefits of reach outweigh the need for precision, probabilistic ID resolution is probably the best way to go. Speaking broadly, this is why advertising and other device-level optimizations tend to leverage more probabilistically matched data.
Deterministic matching, then, is going to be best for you when you are looking to obtain maximum accuracy in your matching and have access to the right data points to enable you to leverage this approach in engagement campaigns. CDPs and many other solutions that use deterministic matching typically rely on personal data to aid the matching process.
This can be as simple or as complex as your business needs it to be. For example, your ESP (Email Service provider) will commonly perform identity resolution by using an email address as a unique identifier to join two records together. A CDP like BlueVenn, meanwhile, may also use email, but if the email address is not supplied, or a customer provides two email addresses, then the BlueVenn CDP will take a series of data points such as the phone number, postal code or ZIP, mobile phone number and device ID, and use algorithms to match an individual across all those data points.
In summary, probabilistic matching provides scale, particularly when data is anonymous on digital channels, for advertising and acquisition strategies, while deterministic matching adds accuracy by ensuring the very best match possible between two or more records, for personalized marketing to your customers for cross-sell, up-sell, retention and other first party data tactics. There’s no need to stick to using just one type of matching, though. In the context of identity resolution, the two types can work together to meet the many requirements of your business.
So, how can identity resolution enable true omnichannel marketing?
This 30-minute webinar will help you to better understand how identity resolution can enable you to improve personalization and the customer experience, as part of your omnichannel marketing strategy:
To learn more about how a CDP can help you to perform ID resolution, register for the next live demo of the BlueVenn Customer Data Platform to see identity resolution and omnichannel marketing in action!