It's Time to Move Beyond Segmentation

Written by Natasha Janic on June 14, 2019

In the marketing world, segments represent the nearly infinite ways that consumers can be grouped based on demographic, behavioral, transactional (the list can go on) data. Marketers often slice and dice data in creative ways to deliver personalized messaging to their customers. We can think of segments as filters you’re able to put on their customer data sets.

Segmentation Limits

Segmentation is based on historical data and only looks at customers who meet the designated criteria at a single, suspended moment in time, unable to account for anyone in the future who will meet those same criteria.

Segmentation Image

Marketers and data scientists have to re-run queries over and over and over in order to capture additional customers who meet the designated criteria for whatever campaign they’re running. Even then, this does not happen in real-time. That sounds like a lot of heavy lifting for less than optimal results, doesn’t it?

A More Dynamic Approach

Instead of thinking of customer groups as segments, start to think of them as more dynamic: what we like to call Audiences. You can start with the same criteria as you would for segments, but instead of putting that filter on your existing data set, an Audience listens for any time a customer meets these criteria once the offer goes live.

Audiences Image

Rather than retroactively assessing your historical customer data with segments, Audiences allow you to programmatically understand customer behavior as it’s happening and launch real-time activations. Thinking of your customers as Audiences allows for greater flexibility in your strategy, adjusting it as you go, enabling you to test different offers.

A Better Customer Experience

Have you ever gotten an offer for something you bought months ago? Or continued to get communication about something you’re no longer interested in? This is likely because that retailer is relying on segmentation to deliver what they believe to be personalized offers.

People change. Taking a snapshot of who they are and what they like at a single moment in time has an extremely short shelf life.

Approaching the customer experience dynamically allows you to evolve and grow with your customers, showing them that you understand and care about their needs, regardless of how their preferences change.

The key to a great customer experience is not only being able to group customers into Audiences in real-time, but it also includes acting on that data in real-time. Think about what you want to happen as soon as a customer meets the criteria for a specific Audience.

Segments vs. Audiences, a Comparison

The overall approach to creating impactful moments for your customers starts with a desired outcome, the Audiences that will likely help you reach that outcome, and the drivers that will motivate your customers to change their behavior.

For this example, let’s pretend you’re a coffee chain trying to put a stake in the ground in the QSR space, so you have introduced a line of lunch items including salads and sandwiches. These are priced anywhere from $7-$10 per item.

Desired Outcome: increase lunch item sales by 10% this quarter

Audience Criteria (we recommend having multiple Audiences, but for the sake of simplicity, we’ll choose one):

  • Visits in the afternoon less than once per week: focus on increasing their existing afternoon visits
  • Has purchased a food item: focus on customers who already view your brand as somewhere they would purchase food from
  • Spends $7+: focus on customers who will be more willing to come back for higher priced items than customers who typically only order a drip coffee in the morning

Offer: Earn 10 bonus points when you come back this afternoon and try a lunch item!

  • We chose this offer because this particular Audience is engaging in a lot of the behaviors needed to achieve your outcome, but need a little push to increase the frequency of those behaviors. This would likely need to be adjusted for Audiences where you’re trying to drive a brand new behavior, like purchasing a food item for the first time.

What this looks like with segmentation

Using the criteria above, you pull all the customers who meet that segment and run the offer for 2 weeks. Unfortunately, you only see a 3% lift in lunch item sales and need to rethink your strategy moving forward.

At this time, you have two paths forward:

  1. Run an additional campaign offering 25 bonus points on this same segment that is now stagnant (new customers have met the same criteria for this segment in the last 2 weeks).

  2. Create a new query that captures the new customers who meet the criteria. This may seem like a better option, but you’re now offering everyone 25 points, including the new customers in this segment who may have been responsive to the 10 point offer, costing your business more money.

What this looks like with Audiences

Using the same criteria that you used with segments, you can create an Audience that you call Lunchtime Tribe, but the difference is the “set it and forget it” aspect that Audiences gives you. Audiences programmatically analyze customer behavior to determine whether they meet your designated criteria. As soon as a customer does, they’ll receive the offer laid out above.

Let’s assume that you see the same outcome, only a marginal lift in sales. You can create a new offer that awards 25 bonus points with an added piece of criteria:

  • Has not redeemed Offer 1 after two weeks

By excluding anyone who redeemed the first offer, you’re saving your business money and hopefully changing behavior for the customers who weren’t enticed by only 10 bonus points, while continuing to capture new potential Lunchtime Tribe members.

With Audiences programmatically analyzing your data, you don’t have to do anything more than set the offer up. Rest assured that you’re capturing each and every customer, with no one falling through the virtual cracks and getting experiences that are irrelevant to them.

In Conclusion

Think of customers as Audiences that can change over time and still be marketed to with relevant content without heavy lifting on the marketer’s end. Segments can still be useful and included in the criteria of your Audience if they’re more static data points, like age range or zip code. Anything that can change more rapidly should be used as an Audience criterion.

Marketers have been trained to use segmentation as filters on their data, but it’s time to start thinking of grouping customers in a more dynamic, real-time manner in order to deliver the right message at the right time to drive your desired outcome.

Learn more about how Hatch can help you create dynamic Audiences.