Every business can benefit from segmentation. Segmentation is a way that allows you to split your customers and prospects into different groups. These groups are called segments. You can segment your list in various ways for various purposes. From email marketing, promotions to even website design you can use segments to improve sales, produce better ROI for your marketing campaigns and have better understand of your ideal target market. There are many different ways you can segment and different segmentation techniques you can utilize. Here are some of the most common segmentation models which are easy to implement:

  • Geographic – You can look at your customer or prospect list and see where customers are coming from. You can divide your customer list by city, state, if they live by the coast or inland, in big city or a rural communicate, what country or region they reside in.

You can use geographies to uncover new regions that are missing from your target market, offer different products to different type of customers, improve shipping rates for certain areas and so on.

  • Demographic – demographic segmentation looks into elements such as age, gender, family composition, household income, race, education and so on. For B2B marketers this data includes: company size, revenue, industry, contact’s title and so on. Typically, this data is available from third party data providers that you can append.

Segmenting based on different demographics can help identify your target market and help structure your marketing communication and targeting better.

  • Psychographic – Psychographic information looks into person’s traits that include lifestyle, interests, attitudes, political affiliations and personality traits. This is often hard data to come by. You may have this information if you ever ran a survey, have a niche product that you are selling or have acquired this data from a third-party source. One source that can be utilized is Facebook Ads. You may be able to run ads to different group and using the results establish your ideal target market. While it is not a perfect science it can help directionally.
  • Products/services purchased– Looking at what product or service a customer has purchased can be a key to your segmentation. This is especially important for companies that are selling different categories. For example, if you are an e-Commerce shop that sell women’s clothing as well as men’s, teenagers and children clothing putting a customer’s purchase can help identify if they are a family with young children, a woman, man or a couple. Therefore, you can segment by only sending women’s clothing promotions to those that have purchased women’s clothing and ignore those customers that historically only bought men’s clothes.
  • Lifetime Value – You can segment your customers based on their value. That’s where the lifetime spent per customer comes into play. By identifying high value and low value customers you are able to run loyalty programs to your best customers, promotions to low value group. Typically, you can combine lifetime value along with other attributes such as demographical attributes, products purchased and so on.
  • Behavioral – The behavior of a customer can be a key to identifying if they are a good candidate for your latest marketing campaign, how likely they are to purchase in the future and so on. Some ways you can look at behavior includes: website visits, email opens/clicks, how often they purchase and so on.
  • Price Sensitivity – For certain companies, price sensitive customers should be identified. This can help with your pricing strategy, what promotions to send to what group of customers etc… You can identify the price sensitive group by seeing their response to promotions such as 50% off, and using average spent.
  • Source of Customer – For marketers it is often key to understand the value of a customer based on where they have started their customer journey. By looking at the channel/source of the customer you are able to not only look at the value that each channel brings, but also establish ways to upsell to customers using channels that they prefer. For example, if a customer has called in instead of filling out an online form they are more likely to respond to a telemarketing campaign as oppose to an email campaign.
  • Current Customer Stage – For companies that have multiple customer stages you maybe, able to identify the customers based on their journey with you and your brand. This is typically used in B2B Marketing or B2C marketing for large purchases such as house, car, furniture and so on. Using funnel analysis, you can identify prospects or customers and where they fit. For example, you may identify prospects that have just made an inquiry, those at the pricing stage, demo stage and final paperwork stage, customer stage and upselling stage.
  • Other ways – While the above are the most commonalty used segments there are other ways to segment. You can segment on any piece of data that you have available, all you need is an attribute and a metric. Some example of unique segments: restaurants using allergies to identify special required customers, Netflix identifying your movie preference based on your historical viewed movies, Amazon using product views to identify similar products and cosmetic company segmenting their customers based on colors that their customers purchased.

Now that you have seen all the different ways to segment, you may wonder how to go about and start your segmentation.

Start your Segmentation Model:

  1. Start with identifying your objective. Do you want to segment to improve your email marketing, re-design your website, expand into a new region, establish a target market or looking for customers to upsell to? Once you have an objective you can go ahead and use some of the above segments to get to that goal.
  2. Do you have the data needed? If you are missing data for your segmentation such as demographic information, look into third party data providers to append that data. If you are missing psychographic information you may not be able to use that segmentation.
  3. Prepare your data. Make sure your data is usable. It should be categorized/standardized so you can easily use it. For example, if you have age information based on date of birth you first have to extract the year of birth and then group it by age groups such as 30-40, 40-60, 60+ as an example. This is probably the most time consuming step of your segmentation analysis but without it you cannot move to the analysis.
  4. Establish your metric. What will you use the distinguish the groups? You can use customer count, average spent, lifetime value of the customer, number of open emails, website visits and so on. You often only have one metric with one or more attributes.
  5. You can use two different approaches for your analysis. A simple approach which looks at one attribute at a time where you would just split your customers into a 2 to 5 groups. This approach works best for companies with a specific objective and/or where data is limited.

The second approach is to do cluster or predictive model analysis. This approach works best if you are looking to combine multiple attributes such as age, gender, lifestyle and products. It typically requires statistical software such as SAS, SPSS or machine learning to run the analysis on, but more importantly it requires a large amount of data to run. The advanced approach is often performed by marketing analytics companies as it does take statistical knowledge, software and time, something that most companies do not have in-house.

Conclusion:

Once your segmentation model or models are established, it is now time to put it into practice. Building a segmentation model is often the beginning of a longer journey. Begin by adding the segments into your CRM for a quick reference. Next is time to adjust your marketing to match your new segments you may want to change your communication style based on segments, adjust your product placement online, run different promotions based on customer’s value and so on. When done right, you should see an impact on your bottom line within a few months or earlier.

Menu