How strategic segmentation can take your personalization efforts to the next level

The process of dividing a broad market into smaller subsets is nothing new. In fact it’s been common practice ever since markets grew large enough for suppliers to produce different product models. And while there are many different ways to divide up an audience, the most commonly practiced have traditionally focused on geographic, demographic or psychographic (“values” or “lifestyle” based) divisions.

With the explosion in customer data, increasingly high customer expectations and the need to deliver targeted and personalized experiences, the role of segmentation has become ever more strategic as a means to provide best in class customer experiences. Current technologies make it possible to create segments based on any data point or dimension, on the one hand presenting a vast opportunity to really tailor the customer experience, and on the other hand resulting in something of a minefield environment, creating an almost limitless number of opportunities for businesses. In this blog we share some of our best tried and tested approaches to nailing segmentation strategy.

Starter for 10: different types of  segment:

Geographic segments can range from countries, through regions down to towns or postal codes.  The advent of technology, and more detailed customer information allows us to make distinctions using IP location, and to combine geographic data with other information, to produce segments like “lives in NYC but never purchased from our New York flagship store”.

Demographic segments include age, gender, and occupation. While this sort of segmentation can raise red flags when it comes to diversity and discrimination, for certain industries, the information is vital.  For example with insurance products, a company will need to be able to segment between homeowners and renters, and the types of property involved.  Again, different demographic information can be combined to create increasingly focused customer groups, using information like number of dependants, location, homeowner status and length of time at the property.

Psychographic segments are likely to be very industry, and even product specific. Health food retailers may be very interested in whether their customers are vegetarian, while energy suppliers may want to target different products, like renewably-sourced power, at those most concerned with climate change.

Behavioral segments - like VIP customers, first time buyers, or mobile shoppers - are some of the ones seeing the most change from technological advances.  Now customers can be targeted on products that they viewed but didn’t purchase, their behavior across a website, their preference for shopping using a mobile device.

Stepping it up: Best in class:

Segmentation does not need to be a new part in your team or business. If you have previously defined the most important personas that align with your team’s strategy, it is a great first step to create segments that reflect those personas.

Whatever type of segmentation you use to divide your customer base, a segmentation strategy will contain two different classes of segment.

Core segments are mutually exclusive, and usually describe strategically important, but relatively basic groups that are of interest for your team or company, often being the personas internally defined and commonly used when thinking about the market. They tend to be large, and have a limited number of dimensions in their definition. A couple of well-known and commonly used examples are the behavioral, “First time customers vs Repeat customers” or the different phases of a lifecycle journey.

Opportunistic segments are segments built for a specific use case. It can be a time-limited revenue opportunity or something related to the time of year eg Valentine's Day, The Grand National or Summer Sales season, a strategic initiative, a push into a new market or the launch of a new product line.

They are usually smaller than your core segments, and less reusable. They target a very specific subset of people, often using time conditioning as “... in the last month”, “... in the next day”, such as:

  • People who, in the last two weeks, viewed a flight departing from LHR
  • People who placed a bet on Cheltenham last year, and haven’t placed a bet for this year’s event
  • All people who purchased a winter coat and a scarf in the last month

When starting to use segmentation for personalization, it is very easy to immediately start building out a selection of opportunistic segments that are linked to specific use cases, or something you’ve had in your head for a while.

If you want your segment strategy to be a powerful tool and future proof, we advise starting at a high level, creating a number of core segments - potentially a translation of your main personas, and then layering your opportunistic segments on top.

Core segments, which are created with your fundamental customer groups in mind, are reusable. The structure of your core segments creates a consistent framework that can be used by your whole team and across the business, all through the year.

 

Screen Shot 2017-01-18 at 13.23.20.png

Having standard core segments that are available for your entire team, also helps provide your customers with a more consistent experience.

Another reason to start off with core segments, before thinking about opportunistic segments is customer centricity. Too often, there’s a temptation to create segments with a specific use case in mind, not the actual people. Creating core segments that are strategically important for your team will ensure the customer is at the center of the personalization and optimization process.

Opportunistic segments have the power to be very valuable, but they need to be envisioned and deployed in a different way. The value of these type of segments lies in the opportunities they address meaning they should be created based on key information and knowledge given there can be a vast number of opportunistic segments, and they often only stay valuable for a certain amount of time.

Screen Shot 2017-01-18 at 13.23.47.png

If we look at the examples above, there might be a certain reason why “People who, in the last two weeks, viewed a flight departing from LHR” pose a certain opportunity.

Often these opportunities are found by analysis of the data collected from the website and business knowledge. Whilst initially it might be a good idea to do that manually, scale and complexity make it impossible to continue finding specific opportunities in your market.

Advances in Machine Learning provide a great opportunity to create a sophisticated segmentation strategy and such functionality is becoming increasingly more widespread presenting a great to leverage the power to identify such opportunities. At Qubit, we released the first part of our Opportunity discovery tool in October of last year, which, based on the data collected from the site, discovers specific subsets of your online market (opportunistic segments) that are seeing a lower revenue than their “peer subsets”. And we’re excited that over the next couple of months we have big plans to continue expanding our Machine Learning capabilities.

In summary, segmenting your audience is a crucial step on the path to providing meaningful digital personalization for your customers - and most companies are already on this journey, with customer personas and target audiences in mind.

Advances in technology and data collection now mean that the only limits for segmentation are your own imagination. That having been said, we do recommend you identify and create core segments first and only then layer on your opportunistic segments -  that way you’ll get the breadth of coverage all the time, and the depth when you need it.

If you'd like to chat to us about your segmentation strategy, we'd be happy to help!