Advantages and Limitations of Implicit Data
In my previous post, I asked if your brand is making the most of its data by delivering implicit personalization. Implicit personalization can be more intimate and relevant than explicit personalization, which relies on external or previously gathered data to tailor a site to demographics like age, industry, and location. Implicit personalization draws on users’ moment-to-moment movements and choices to adapt to their most pressing desires and up-to-the-minute goals.
But exactly what types of data should you be looking at to create the implicit experience? Where and how can you gather meaningful, real-time data? And how do you know if it’s truly valuable to your brand and users?
The Particulars of Personalization
Implicit personalization is still based on formulas, making it a form of prescriptive personalization. For example, if a visitor clicks on a product they’ve never looked at before, your site responds with a sidebar of similar products for comparison. Right now, these formulas are the primary logic of implicit personalization, and marketers must anticipate possible user behaviors in order to construct appropriate responses for a multitude of scenarios in advance. Although the process can be demanding, the conversion results can be rewarding.
Because implicit personalization is about responding to a user’s particular behavior while navigating a particular channel, the data that drives it must also be very particular. You have to observe and crunch the data in real time with a human logic that makes the user feel supported. It takes several types of data working in concert to achieve sophisticated implicit personalization, but beware: not all types of data are created equal.
The Benefits and Risks of Three Data Sources
- Profile Data
Individual names, contact information, account numbers, IP addresses, past purchases, and behavior history all populate an individual’s profile data. This type of data does play a role in implicit personalization, as it influences how you respond to a user in real time. Whether or not this data is good (i.e., truly relevant and useful) is sometimes difficult to tell. Let’s say I shopped for a birthday gift for my classical-music-loving uncle on Amazon last week; it’s no longer relevant to me if Amazon recommends a collection of Bach’s violin concertos this week.
To evaluate the usefulness of profile data ask yourself, does this information put me in touch with what the visitor wants and needs today?
- User Segments
User segments are groupings of individuals who share certain demographic characteristics, like age, gender, location, industry, and more. Marketers speculate and predict what each group of similar people will respond to. The idea is that an optimized experience for young professional women should look different than one designed for retired male hobbyists.
Think of segmentation as the backbone of implicit personalization: we need it to support the bones and movement (or structure and processes) of content targeting, recommendations, and to broadly accommodate different visitor needs and goals.
We use segments to make the overwhelming and complex task of personalization manageable. This type of generalized prediction can provide a good skeleton for personalization, but it’s only a skeleton.
The danger of segmentation is that you can end up grouping superficially similar people who actually have very divergent views, tastes, and desires. You have to use deeper data and more dynamic analytics and automation to provide truly intimate, flesh and blood personalization.
- User Intent Data
People come to your site, app, or service with goal in mind. Their goal may be as vague as wanting to learn more about your brand, or as specific as needing specs and pricing on your latest B2B product upgrade. Some schools of thought say that personalization is less about the person and more about their purpose. In other words, understanding who is browsing your site is less important than understanding why. Lane Cochran of iPerceptions says it well:
“Sites demonstrate they know a visitor’s location, what the visitor did on their last visit or even what they’ve done elsewhere on the web. Personalization can then seem less about the visitor’s needs and more about the person themselves.This creates user discomfort without necessarily helping them accomplish what they came to do. Part of personalization’s slow growth and looming privacy concern is a consequence of this focus on the visitor, and not on the visit itself.”
Both the who and the why are undoubtedly valuable insights for the digital marketer, but knowing user intent can bring powerful precision to your personalization efforts.
To know what each and every user wants, you’d need to ask them. This works in person, when a prospect calls or walks into a store, but it’s more complicated in digital. Intent can be gleaned from a visitor’s real-time data, as you look at their browsing history, search queries, click-throughs, and more. Because you’re gathering and responding to the data while the person is on your site or app, it mimics an in-store or on the phone interaction.
The danger, as always, is that you can misjudge the visitor’s goals. Without asking directly and getting an explicit answer, you are inferring what the user is after, and responding with content and recommendations that you hope are truly relevant.
What about the Multichannel Experience?
The outcome of implicit personalization ranges from personalized recommendations and offers, to “a dynamic, personalized site experience in which . . . the actual site content displayed changes from user-to-user.” But just having the implicit data is not enough to create valuable, connected customer experiences. Companies today are looking to use implicit data to perform adaptive personalization—a continually evolving, dynamically automated personalization strategy that supports meaningful interactions across multiple channels and touchpoints.
We’ve looked at three useful types of data that drive implicit personalization, and where they fall short. Now we need to think more deeply about the contexts and journeys of your users. Instead of treating each channel as separate areas of interaction with distinct user goals, we must learn to see them as interconnected pieces of a complex but unified user journey. This more holistic, integrated perspective will help position you for a leap into adaptive personalization.