• HomeHome
  • »
  • InsightsInsights
  • » Don’t Get Lost in the Data: The Easy Way to Drive Personalization
Insights | By Howard Tiersky

Don’t Get Lost in the Data: The Easy Way to Drive Personalization

In our last article on personalization, we talked about the three key tactics for personalization: identifying which customer you are interacting with, gathering or accessing relevant data about the customer and using the data to take action. 

Get FREE access to the first chapter of FROM`s
Wall Street Journal Best Selling Book


  • Learn the three patterns of all successful digital brands (including companies like Apple, Netflix and Uber).
  • Understand why many great new products fail, and the formula for building products that won’t.
  • Discover the key reasons companies resist change and how to overcome them.
Get FREE access to the first chapter of FROM's
Wall Street Journal Best Selling Book


  • Learn the three patterns of all successful digital brands (including companies like Apple, Netflix and Uber).
  • Understand why many great new products fail, and the formula for building products that won’t.
  • Discover the key reasons companies resist change and how to overcome them.

But there’s one specific question that I get from clients a lot: How do I determine which data I should use to drive personalization?

What’s great is that, in this era of “big data,” we very often have a wealth of data about the customer. But that means it’s also easy to get lost in the hundreds of fields and tables that may be available and to become paralyzed with uncertainty over which fields to actually use. At FROM, we’ve worked with many companies to define strategic, data-driven personalization models, utilizing a “bottom-up” approach to determine the most relevant data for driving personalization results. Based on this experience, I’ve got some good news and some bad news.

The good news is that the most effective personalization efforts focus on only 5-8 fields of data!

The bad news is that we often find that, once we take our clients through our detailed process for narrowing down those key high-value segmentation characteristics, they often only have half of the 8 fields they really need (even though they may have hundreds of fields of data!)

Don’t worry -- even if you don’t have all the data that you’d like for your customers, there are some clever techniques you can use to incent the customer to provide that data (which we’ll cover in a future article.) In this article, I’ll provide some tips on how to identify those key data elements.


The Sales Message Framework: For Identifying the Most Important Fields for Personalization

There are so many ways we can segment and target our customers: age, ethnicity, gender, past purchase behavior, zip code, etc. And if your goal is to sell something to your customer, then the objective of personalization is to tailor your sales message in a way that increases the effectiveness of your sale. So what types of fields help you do that? Generally, there are four components to a sales message. You can use those components to look for optimal fields.

The first component of a sales message is the problem your product or service will solve. Is it the hassle of having to do your taxes, and the uncertainty of whether or not you did them right? Or is it the pain of wearing high heels and having swollen ankles?

Connecting to the problem or needs of a target customer is the first step to effective marketing. However, it’s possible that your product can solve multiple problems in different scenarios. In this case, look for customer characteristics that will help you zero in on the problem they are most likely to have. Say you’re selling vacation packages. Families with small children face different challenges when taking a vacation than a couple looking for a romantic getaway, so a key component for this type of sale would be understanding the characteristics of the composition of the group of travelers.

The second component of a sales message is focused on features or benefits. Which are the characteristics of your solution that solve the customers’ problem.

Here again, different customers will resonate with different features. If you’re selling a minivan, you may have buyers whose number one concern is focused on safety features like airbags and anti-lock brakes. That buyer will resonate with different feature descriptions than a buyer whose priorities are focused on interior comfort or cargo space.

The third component of a sales message is focused on proof. You have asserted that your product has certain features and that those features solve the key problem of your customer. But is that true?

Customers look at testimonials, independent reviews, certifications and statistical data to verify that your claims are true. Personalizing these components makes them far more effective.

For example, if the potential customer of your accounting software owns a factory, they are going to be far more convinced by a customer testimonial of someone who manages car manufacturing plant than that of someone whose runs a day-care center.

Having a library of “proof” content that can be selectively presented based on appropriate segmentation characteristics can be very powerful. We’ve noticed that for consumer products, testimonials from customers whose home town is close to those of the prospect are viewed as being more relevant and truthful than those from customers located that are farther away.

The last component of a sales message is the offer. Via personalization, you can increase conversion by positioning different pricing or bundles to different types of customers based on key characteristics.

Disney World offers much lower pricing to Florida residents than residents of other states, so knowing that a prospective customer is a Florida resident allows them to automatically offer a better price. In other cases, customers who’ve made prior purchases, are senior citizens or are military veterans, may be entitled to different offers.

You can also choose to offer higher discounts to customers you have reason to believe require more of a discount in order to take action and avoid leaving money “on the table.”


Applying the Framework: The Experts Right Under Your Nose

Every marketing situation is different. The framework provided is useful, but still doesn’t tell you the precise fields to use for personalization in your company. So how do you figure out which features or benefits are most important to which customers, or for the other three aspects of the model? At FROM, we utilize an approach that combines leveraging the experts right under nose with A/B testing for validation.

Who are the experts? Your sales people! I worked on a re-conception of the sales sites for Cadillac, Saab and Chevy a few years ago. As part of that project, we spent time with their salespeople, trailing them and learning their tricks. Good salespeople have spent years figuring out what questions to ask prospective buyers and learning how those answers map to key sales messages. While the techniques we use in direct marketing may not be exactly the same as those used by salespeople, the more interactive and personalized our marketing becomes, the more they start to resemble a live sales situation.

In the case of this project, we gained a ton of insights from different salespeople, which we then turned into digital A/B tests. And, while not every idea translated, many of them did! We learned that car salespeople often ask what car the customer is driving now, which helps them understand a lot about that customer’s priorities and helps them tailor their message. Similarly, spending time with call center representatives can be helpful.

Where do I get the data?

As I mentioned at the start, the result of this analysis might be that you have an enormous amount of data (some of which you don’t necessarily need), but that you lack some key fields which should be your key personalization priorities. Based on the above example, you might wonder where that data’s supposed to come from? A salesperson can ask direct questions, but it’s not as easy as that for marketing. You can refer to my last article, where we discussed a variety of sources of data. In the next article, we’ll discuss some additional techniques you can use to get users to provide more data to help your personalization efforts.

If you have any comments or thoughts, please reach out to me admin@from.digital. Let me know any specific questions or challenges you’re facing around personalization, and I’ll try to cover them in future installments of this series.