data for personalization
Stock photo

If you want to personalize the experience on your website, mobile app, or any other channel so it is relevant to each individual, the first thing you need is the right data. It should be obvious why that is: because you can’t create a personalized experience if you don’t know anything about the person.

But what data do you need to create a complete picture of an individual? In the spirit of Thanksgiving, I want to answer that question for you, relating each type of data to a Thanksgiving dish. Because just like every individual dish on Thanksgiving can come together to create a delicious meal, different types of data can come together to create a clear picture of each of your customers or prospects. So loosen up that belt, fasten your bib, and let’s dive right into this festive analogy!

1. Turkey: Behavioral Data

It goes without saying that the centerpiece of the Thanksgiving dinner is the turkey. You simply can’t have a traditional Thanksgiving without one. And just like turkey, behavioral data is the centerpiece of all the data you need for effective personalization. A person’s behavior in the channel they are engaging in tells you a lot about their interests, what they are looking for, and who they are. You can’t deliver an effective personalized experience without the ability to collect behavioral data.

But you don’t want just any turkey at Thanksgiving. You want juicy and flavorful turkey — not dry and bland. The same goes for behavioral data — just any behavioral data isn’t going to cut it. Site-wide behavioral data (like number of times visited or time since last visit) or page-level data (like how many times each pages was viewed) will not do the job alone. To deliver truly individualized personalized experiences, you need deep behavioral data which includes engagement information such as time spent on a page, mouse movement, scrolling, hovering, inactivity, etc..

2. Gravy: Contextual Data

Turkey goes hand-in-hand with gravy. You wouldn’t want one without the other. The same goes for behavioral data and context. Without context, behavioral data isn’t as useful.

With behavioral data alone, you can identify that a visitor was very engaged with one blog post over another, or one product over another, but you don’t know what that means until you can put it in context. What is the topic or category of the blog post? What is the price and the brand of the product? With that information, you can piece together what a person’s favorite topics are, which price range he prefers, what brands and styles are his favorite, what he is looking for on the site right now, and much more. With this information, you can deliver experiences that take all this information into account.

3. Stuffing: Cross-Channel Data

Just like with turkey and gravy, it’s not Thanksgiving without the stuffing. It’s the Thanksgiving equivalent to cross-channel data, because after making sure you have deep behavioral data and contextual data, cross-channel data is the next most-important data source you need.

This may seem obvious, but if a person starts to engage with your company on a laptop but switches to a mobile device, her interests and who she is as a person do not fundamentally change. And particularly if she is a regular customer, there is no reason to treat her differently across channels. Yet marketers continue to struggle to recognize a person across channels and then use omnichannel data to power experiences regardless of where that experience takes place.

4. Cranberry Sauce: Campaign Engagement Data

Rounding out the key pieces of a traditional Thanksgiving dinner is the cranberry sauce, just as campaign engagement data rounds out the critical components of the data you need for personalization.

Campaign engagement refers to the actions an individual has taken in response to any of your campaigns across channels. It includes behaviors such as: personalized experience views and clickthroughs, email opens and/or clicks, push notification dismissals or clickthroughs, correlation of the above to device, time of day, or other variables. Campaign engagement data ensures that you are able to incorporate learnings from your campaigns to avoid leveraging the same tactics over and over again if they aren’t working for an individual. It allows you to recommend more of a brand if a person regularly clicks through product recommendations for that brand, or avoid sending specific messages via email if a person has already viewed them in an email.

I’ve heard people grumble about the uselessness of cranberry sauce in the past. Similarly, campaign engagement data is often overlooked by marketers planning personalization campaigns. But it plays an important role in your understanding of each person, just as cranberry sauce plays an important role in your Thanksgiving meal.

5. Mashed Potatoes: Web Attributes

After you secure your turkey, gravy, stuffing and cranberry sauce, you can move on to other side dishes — and the mashed potatoes are always my next stop.

While not as mission-critical as some of the other data types I’ve mentioned so far, web attributes can be hugely beneficial to personalization. Web attributes are detected the second someone lands on your site. For example, geolocation, source (such as search, email, social, paid ad, referring site, etc.), industry, company, company size (revenue or employee count), browser type, device type and time of day are all web attributes.

There are many opportunities to leverage web attributes in your personalized experiences. Source data can help you be as relevant as possible to first-time visitors. Firmographic data is critical for ABM efforts. Geolocation data is important for informing visitors of events in their area they might be interested in, targeting visitors with products or content relevant to their regions or weather conditions, and much more.

6. Family or Cultural Traditions: Company-Specific/Database Attributes

You probably started this list saying to yourself, “she had better include my family’s favorite side dish or I will be outraged.” Of course, every family puts their own spin on the Thanksgiving meal. Some include cultural dishes, others have included different dishes over the years as their traditions evolved. The good news is that this ties in perfectly to personalization data as well.

There will always be data points that are unique to your own business that you can use for personalization by tying in other data sources, such as a CRM. Attributes such as a visitor’s loyalty program status, whether that visitor is an existing customer or not, what financial products that customer uses, etc. can be used to personalize a person’s experience. What those attributes are will depend entirely on the needs of your business, just like your Thanksgiving meal uniquely meets the needs of your family.

7. Dinner Rolls/Cornbread: Third-Party Data

Concluding the savory portion of the Thanksgiving meal, we have the bread on the side — such as dinner rolls or cornbread. Bread isn’t as critical to the meal as some of the other dishes I’ve listed so far, but it has a role to play nonetheless. Just like third-party data. Third-party data (data collected by outside sources across the broader internet) can be used for personalization as well — but it’s not always relevant to every situation.

Third-party data can tell you a person’s demographics, can inform you of specific buying signals (such as whether the person is in the market for a new car/home/technology solution), or let you know about any self-defined attributes the person has listed somewhere else on the internet (such as relationship status, career information, etc.).

In my own opinion, you could live without the bread on the side of the meal, but it can definitely add to your overall experience. The same goes for third-party data.

8. Dessert: Survey Responses

Finally, survey responses are like the dessert at the end of the meal. You don’t need dessert to satisfy your hunger, just as you don’t need to incorporate survey responses when there is so much that you can learn about a person from the other data types I’ve outlined so far. But you may opt to ask a survey question or two to learn something from your visitors that you can’t infer from other data sources quickly — such as how a visitor feels about a subject — as the final piece of your picture of each individual.

Wrap Up

All of this data is meaningless unless you can bring it into one central place, decipher what it means about each individual, and act on it in real time — much like how all these delicious dishes would be pointless if they didn’t come together in your belly and fill you with joy (am I taking this analogy too far?). For that, you need a complete personalization platform.

Have a great Thanksgiving, everyone!