5 things about web personalization and customer data platforms

Many customer data platforms got their start with a focus on web data unification and deliverability, one way or another. There are both a data collection side and a personalization side to the integration of CDPs and websites. With a unified profile and real-time decisioning capabilities right there connected to the website, the days of a static web are over.

Here are five things I thought about when considering how customer data platforms are contributing to web personalization.

Thing #1: Customer data platforms are disrupting A/B testing a bit

Overlap is, at the least, a yearly point of interest for technology buyers. Many CDPs have built-in web personalization capabilities. From the buyer perspective, here are two systems that "do the same thing" -- why pay for both?

It makes sense that overlap brings this kind of cost/benefit analysis, but customer data platforms don't replace A/B and MVT vendors. The best, or someday we might say "the true" customer data platforms integrate with multiple A/B and MVT vendors, however.

But there is overlap. This overlap in the area of A/B testing (not usually MVT) doesn't happen only with customer data platforms, it happens with email service providers and content management systems as well as other types of marketing technology.

flickr/horsager

There is a big big big difference, though! CDP play nice as a data source to enhance segmentation, 1:1 messaging, and measurement capabilities. Customer data platforms have the advantage over these other categories thanks to access to lots of data, to being the source of a unified customer profile. Whether tests are run through the CDP, or in conjunction with the CDP, audiences can be selected better, tests can be more relevant, and results can be measured against preconceived and automatically discovered segments.

At the same time, the CDP might cut into some testing/activity that would traditionally have happened within a testing platform, replacing some tests altogether.

Often, there is no testing and optimization program in place at all. In these situations, such a program can be initiated on the back of a customer data platform. But not just any customer data platform! Which leads us to:

Thing #2: Capabilities, again, are quite discrepant

When talking "5 things about email and customer data platforms" and "5 things about data collection and customer data platforms," there was a spectrum of capabilities each time. This area of web personalization is no different. Some vendors that qualify as customer data platforms do not have any user interface for testing/optimization/personalization. Data can be used, but that is the extent of it. Other vendors have full testing suites bolted onto the CDP -- of varying levels of sophistication, of course.

Some customer data platforms have plug-and-play personalization capabilities with little to no development effort required, if your web team is cool with that. Good news for the devs: all of those same customer data platforms offer more dev-friendly methods of personalization in addition to these UI-based personalization capabilities.

The capabilities here are so discrepant, I could see a single organization finding a need for two different customer data platforms. Perhaps this will converge over time so that decisions don't have to be made on a platform that favors execution versus a CDP that focuses more on customer records, or analytics.

Thing #3: Individualized content makes a difference for both known and anonymous users

flickr/horsager

On one hand, web personalization is just scratching the surface of what a customer data platform can do. For a lot of businesses, though, this connection of online and offline data and the ability to tailor communications on the web and in email to be relevant based on behaviors is still entirely unexplored. For this reason alone, customer data platforms will continue to take off as organizations scramble to find reciprocal value for their customers.

Vendors are all over these use cases in case studies, with claims of 70% lift for this and 5x the results for that. And they are generally true, though sometimes rosy interpretations. I look forward to continued posts on Web Analytics World that go into further depth around some of the activations enabled by Tealium's Audience Stream / Universal Data Hub. The email use cases mentioned are representative of typical initial CDP+email use cases, but that doesn't make them any less valuable! And, it's nice to read the specifics, like the fact that propensity modeling (proprietary as it might be!) is in place to ask the "right" customers for a recommendation.

The point is that the individualization of content makes a difference, though it does take some forethought.

Thing #4: Make the best possible use of personalization capabilities

While it can be easy to put on the blinders and get some things done, web personalization should not be a marketing-only thing. New web products should be built from the ground up to make use of a customer data platform, and all groups owning some piece of the web pie should be made aware of the capabilities. There are avenues for CDP use throughout the organization, and other data points that can be combined to round out profiles.

Highlighting support articles for products owned, or prompting customers to confirm/correct questionable data are fantastic use cases for a customer data platform and web personalization that might fall outside the typical acquisition lens personalization tends to take on.

Thing #5: Understand which areas are next-most important to web personalization

Use cases for customer data are related. They are all based on the same unified profiles! What data can be shared? What data should be shared as-is versus transformed in some way for the delivery system? Understanding the areas data will be used is key to making sound decisions during implementation steps.

For example, the same behaviors used to segment customers based on their interests are likely to be useful across channels. The data needed from one channel to the next might differ, however. Let's say I land on a website for the first time and read a pile of baseball related content. On the website, and in email if I give it, content might be tailored to include more baseball products and articles based on that information alone. In advertising, it might not be enough to know that "Todd likes baseball". Instead, it might be better to know that "Todd has a baseball engagement score of 15 over the last 30 days" from a bid optimization and timeliness perspective.

Years of hype

Customer data platforms were created out of a need recognized by marketers, and early successes with the technology are now driving extreme growth in an entirely new category of marketing technology.

Marketing automation and CRM have created years of hype around "the same thing" (relevant communications by way of customer data). In the case of website personalization, it simply has not been as easy to integrate customer data into your website as it is today.

Leave a comment