Get Personal with the Right Segmentation Strategy

Mar 24, 2009 4:44 AM  By

Remember the movie Minority Report with Tom Cruise? When Cruise’s character walks into the department store, they identify him with a retinal scan which triggers a TV screen featuring products based on previous buying behavior and interests.

To watch this unfold on the big screen is fun and interesting. In real life, this is creepy.

We have not reached the point where personalization is an outright violation of our privacy—at least not yet. Many retailers are looking into adding personalization functionality to their Websites because of the much-hyped notion that in doing so you can create the ultimate customer experience.

Forrester Research corroborated this trend by revealing that 58% of shoppers now notice personalized product recommendations and 34% are influenced to purchase. With shoppers starting to feel comfortable with the concept, e-commerce executives continue to “buzz” about personalization as 42% of survey respondents plan to “implement” some personalization strategy on their Websites.

Despite consumer momentum, this technology is still not at the top of most marketers’ wish lists, because current technologies remain imprecise, especially for personalized page content. In light of this, retailers have changed direction from the grandiose strategy of one-to-one personalization to a more manageable goal of customer segmentation.

To truly understand how to segment customers for driving revenue, you need to use methods that yield richer customer insights aimed at building a tailored customer experience. Before you sit down to write a segmentation strategy, think about building a plan for implementing analytics and tracking as part of the data collection process.

Early in the process of building or modifying your ecommerce site, it’s important to A/B test all aspects of the online selling process with the goal of using this data to construct customer groups. Then use the data to create segments for marketing and merchandising purposes that span campaigns, products, promotions, pricing, and search.

You must think about how the numbers can be leveraged to drive the relevant offers and content out to customers. Creating the ideal customer experience is imperative so use the data in a way that creates relevancy without encroaching on privacy.

Doesn’t this all sound easy? Unfortunately, it’s not, and many segmentation efforts stall or get canceled because of its difficulty.

The principal reason for this lack of momentum is because people fail to follow the “Four Ts” of successful segmentation: Target, Team, Technology, and Timing. The “Four Ts” make up the basis for a solid segmentation or personalization plan.

Target
The concept of doing any type of segmentation—never mind personalization—starts with identifying your target segments. A segment is a subgroup of people sharing one or more characteristics or behaviors that cause them to have similar product desires or needs. A true segment is distinct, exhibits common needs and responds similarly to a common stimulus, in this case content, product or promotional offers.

The segmentation process consists of segment identification, segment characterization, segment evaluation and target segment selection. The utopian goal of a segment is to create an integrated blueprint of the customer’s demographic, geographic, attitudinal or behavioral data.

Next, you have to find common themes among customers that fulfill expectations, preferences, and satisfaction levels. Before you create segments, ask yourself the following questions:

  1. What is the makeup of my customer base and are there measurable patterns in behavior?
  2. Is my audience diverse enough that creating customer groups by demographics, psychographics and geography will deliver accurate results?
  3. Will your segmentation strategy provide you with differentiation from the competition that will resonate with this target (i.e. convenience, personal touch, value)?

If you can’t answer these fundamental questions with confidence, your business might not warrant a segmentation initiative.

Team
Setting up your organization for success is often overlooked when building a segmentation strategy. Depending on the number of segments you create, and the level of targeted merchandising and marketing you plan to deliver, you could be setting yourself up for an intensive human resource undertaking.

Unfortunately, most online retailers don’t accurately scope a segmentation or personalization initiative. The major reason is that in the person building the segmentation strategy often is not the person doing the implementation.

So what is the rule of thumb? Since any segmentation strategy starts with data, the team should include a marketing or database analyst.

Depending on the breadth of marketing and merchandising aimed at these targets, you most likely will want to add marketing and merchandising managers to the group. Depending on the size of the product catalog and the complexity of the promotions strategy coupled with the number of segments, you might also need several marketing and merchandising managers.

Just remember that adding staff adds cost to your site maintenance. Completing a detailed ROI on what you think can be generated in additional revenue is imperative before you invest in the extra manpower.

Technology
You can use many technologies to drive a segmentation or personalization strategy. All of the technologies that drive effective personalization strategies start with data collection.

Defining the data you want to collect and setting up the methods in order to collect are the most important aspect of any segmentation strategy. Once the data is collected it must be processed to trigger a beneficial shopper experience.

Finally, there must be an output vehicle for delivering the required customization. Below is a more detailed look at relevant technologies for driving data collection, processing and segmentation outputs.

Data collection inputs:

  • Data provided by shopper (address, home airport, product preferences)
  • Implicit information gathered through click stream or referral
  • Multichannel data collected through loyalty programs or the call center

Segmentation processing:

  • Rules engines
  • Algorithms
  • Neural networks
  • Collaborative filtering

Segmentation outputs:

  • Site versioning (different versions of the same site)
  • Dynamic content
  • Dynamic merchandising (cross-sells, upsells, and promotions)
  • Collaborative recommendations
  • Dynamic marketing (e-mails, banner ads)
  • Dynamic search sort orders

Timing
Timing, as always, is everything. The question often asked is when should you launch a segmentation or personalization strategy on your Website?

The answer is simple: when you have enough customer data to create accurate customer groups.

Segmentation or personalization is rarely the first phase in the life of a Website build out unless the online retailer has been gathering data through additional sales channels or a previous Website. That said, online businesses with frequent shoppers could introduce segmentation in as little as six months based on the level of data created.

For online retailers that have customers visit or shop once or twice a year, it could realistically be 18-24 months before you have collected enough meaningful data.

Segmentation, when used properly, can be a powerful tool for driving conversion, improving average order size and increasing customer retention. Keep it simple at first while you get to know your customers.

Once you have more customer data you can build a more convenient shopping experience that will encourage repeat customers. If you don’t exercise restraint, you risk creating a creepy customer experience that will lead to abandonment.

Scott Todaro (stodaro@demandware.com) is senior director of product strategy at Demandware, an e-commerce platform provider.