Identifying and Multiplying Your Best Customers with Customer Segmentation

What does it mean to segment customers?

Divide up customer base or market based on shared characteristics – the goal is to identify your company’s best customers and then build highly relevant products and experiences that will engage, retain and acquire them across the entire customer lifecycle. 

“Best” customers could mean one core segment, or multiple (depending upon your scale) – how many segments you focus on is going to depend on the scale and resources that your business has available. Many companies (especially younger, smaller ones) choose one core target segment to focus on, and then think about adjacent moves they can make as they start to win that market to propel growth.

Segments need to be based on qualitative and quantitative data – you can’t base segmentation on the opinions and beliefs of your C-suite or employees across the company.

What is the business value of customer segmentation?

Improving team momentum and alignment – focus is the key to a successful segmentation exercise. Sometimes people worry, “what if I go through this process and I make a bad choice?” Honestly, my answer is “you probably won’t”. At the end of the day, if you’ve taken all of the data points into consideration and you’ve gotten all of your team’s focus on one customer segment,you’ll gain momentum. You don’t want to ask 10 different employees who’s your target customer and get 10 completely different answers. 

Improving business performance (lower CAC, lower churn, higher LTV) – part of the process of customer segmentation is identifying high-performance segments. You want to identify the type of customers who usually perform better than the average customer. As you increase the ratio of your target customers, and decrease the ratio of your non-segmented customers, you’re actually able to increase business performance, drive down cost per acquisition, drive down customer churn, and increase ARPU or customer lifetime value. That’s the long-term impact that you’ll see if you’ve nailed that first alignment piece. 

More relevant and personalized experiences for customers and prospects – personalized experiences have completely become table stakes now. As a customer, you want to get your questions answered as quickly as possible, feel like you’re in the right place, and that there’s a sense of belonging. Segmentation provides that to you as a prospect because it allows you to decide quickly, “Is this for me and will I find this valuable?”

Segmentation factors in at the product level too – most of the time when I talk about segmentation, I think about it from a holistic business perspective, but it can be applied all the way down to the product level. For example, as a product manager or product marketer working on this product portfolio, what’s the best segment I can focus on out of our company’s main segments? So it really can be applied to both startups all the way to something like a scale up or a growth stage business. 

What are signs that a company should invest in defining its customer segments? (Or at what stage should a company think about this)?

If you ask 10 different people at your company who your target customer is and the answer isn’t the same – if you want to deliver customer value across all the functions of your organization, you need to be able to identify who the customer is.

Companies of all sizes benefit – you can apply it at different levels. The work just might look slightly different depending on what stage of growth you’re in.

Who runs the segmentation process, and who should be involved?

A product manager or product marketer runs segmentation – product marketing and product management need to be aligned, so that segmentation and product vision match up. I recommend product marketing take the lead if possible, because the team’s objective is to become an expert on both the customer and the market, and they’re market-facing. Product marketers already have the knowledge and skills to look both inward at your existing customers and outward at market potential.

A data team is a helpful resource, but shouldn’t lead – it’s valuable to have support from a data team in terms of gathering the right data sets and helping to analyze some of the data. Although your data team may be able to breeze through this exercise, you don’t want them to lead this project because they may lack the marketand customer context to put it all together. 

Other stakeholders and supporters will be users of the segmentation framework

• CEO/Founder(s) – if you’re really small, founders run the whole project
• Data team
• Marketing
• Sales
• Customer Success

What types of customer variables should companies use in segmenting customers?

Firmographics and Buyer Demographics – for B2B, things like industry, employee count, company revenue, and job title.

Geographic – it could range from what country they’re located in to a region to what neighborhood they’re in.

Psychographic – lifestyle choices, attitudes, personality traits. For B2B, this comes down to values–ultimately it’s still a human making that buying decision, and what they value is important. Some buyers might really value low prices and getting a deal, others might care a lot about service levels.

Behavioral – what are the different habits, purchase behaviors, and benefits that a customer would be seeking? Think about how they prefer to buy (self-service vs. sales-guided), how they prefer to consume information, etc.

Start with Demographic and Geographic, then layer in Psychographic and Behavioral – you’ll build your customer segments based on the first two categories. And then when you’re doing your persona work, go out and actually try to learn more about those segments to overlay psychographic and behavioral characteristics.

What are the steps in the research/analysis of defining customer segments?

Use the MAP model – I’ve developed a framework for building a segmentation strategy, MAP stands for measure volume, analyze performance, and prioritize potential.

Step 1: Measure Volume – the first step is to identify how many customers you have in different groups, and patterns between groups, to develop preliminary hypotheses about your most valuable segments. Start by understanding how many customers you have in different groups.

Build a big dataset – in Excel or Google Sheets. There’s no fancy tools that you need. Merging all the data together is the hardest part of this. You’ll want to gather customer data from current customers, trialers, and churned customers.

Slice and dice the data to get 3-5 segment hypotheses – look for patterns or affinities between groups of customers. You’ll want to try and find information for all 4 categories (firmographic, geographic, psychographic, and behavioral), plus a fifth category that I call account data (e.g. their subscription type). 

You want to walk away from this stage with at least 3-5 different potential customer segments or segment hypotheses.

Step 2: Analyze Performance – this is still the meaty step where we’re in the spreadsheet and we want to either validate or invalidate those hypotheses that you have around these potentially good segments. After this analysis, you want to have 1-3 high-performing segments that you want to pursue. 

Compare segments based on performance metrics – Once you’ve identified a handful of potential customer segments you have a lot of, you’ll want to validate that they are high performing. For example, do these customers convert better than average, or have an above average high retention rate. 

The variables that matter most depend upon your business model – if you’re a non-subscription business that depends upon repeat purchase, the number of times someone is purchasing from you, and lifetime value may be more important. If you’re a self-serve business, trial to paid conversion rates is going to be incredibly important. If you’re an enterprise B2B, close rate is going to be more important. 

Build cohort analyses – the dataset spreadsheet will tell you how much money you got from a segment over the course of 12 months, but it won’t tell you what that segment’s conversation rate was or what their survivability was like. So you’ll want to build a cohort analysis for things like survivability over time or lifetime value over time.

Model the potential financial impact of focusing on specific customer segments – for example, if I were able to increase the amount of customers I have in this customer segment by 10%, and they were performing at this conversion rate (compared to the average customer who performs at some lower conversion rate), how much would I be able to increase the overall conversion rate of my business? You can start to see the impact and the ROI from focusing on really specific customer segments.

Step 3: Prioritize Potential  – the problem we want to solve here is, can we win this in the market?

Is the market big enough? – how large is this market? Even if you’ve found the best performing customers, it does you no good if there are only 10 more of them. You want to make sure that the target market you’re going after is going to be able to sustain your own growth ambitions as a business. Rule out if some segments are too small for you to focus on. This will also help you to hone in on a segment if you have limited resources and can only afford to focus on one

How fast the market is growing? – how many new businesses or potential customers would we expect to enter into this market in the next few years? That could be from new startups coming into existence, existing companies adopting this product for the first time, or existing companies switching from one solution to another. Start asking yourself, what percentage has a solution today and what percentage will be looking for a solution.

What’s the customer acquisition cost – understand how much it’s costing you to acquire these customers today and whether that’s affordable. What channels are you acquiring them on? Is that acquisition strategy scalable? You don’t want to invest in a segment that you can’t scale meaningfully, so you want to assess that and then understand the impact to your strategy. 

What does the competitive landscape look like – I say this one with a grain of salt because you never want to change your strategy based on what your competitors are doing. It’s more about understanding if your segments are a “blue sky opportunity” or if they’re competitive. This will shape your marketing strategy as well as your modeling. 

Use MAP to pick the top segments – knock out segments that are too small or too costly. Take everything into account to choose the most attractive markets (if you can’t go after every segment).

Add color with qualitative research – perform prospect/customer interviews & surveys to inform positioning, go-to-market, messaging, and pricing you have for each segment.

What are the steps in rolling out segmentation to start treating segments differently in sales and marketing?

Bring key stakeholders along for the journey – you don’t want them to be shocked at the very end, and you want to start to understand from them, what do you need to be successful? They’re probably going to need some new positioning and understanding of how does this product, our existing product, work for this customer segment?

Audit what the customer experience looks like now for this segment/these segment(s) – to determine where along the customer journey you need to optimize. Are you doing a good job at delivering relevant and personalized experiences for the segment today or is it completely generic? You kind of want to put yourself in the in the mind of that customer segment and start the entire journey.

Examples of segmentation rollout in marketing

• Test value props with PPC (pay per click) campaigns – validate the value proposition that resonates with new segments or potential segments. 
• Buying journeys on the website – if your customers land on the homepage, do they know where to go? If they go to your drop down menu, is there somewhere that directs them where to go?
• Use segment-specific language – even if your value propositions are almost identical for each segment, you should word the proposition in a way that resonates with that particular segment.

Examples of segmentation rollout in sales

• Pitch deck for each segment – if you have multiple, the reps need to understand what the value propositions are for each of those segments. 
• Different battle cards – reps should be able to identify early on a call what segment a prospect falls into. Then they’ll go to that specific battle card. 
• Different pricing models – based on the segment the prospect falls under and what your business would be offering. 
• Sales enablement and training should incorporate segmentation – when you’re doing role playing, try to make it as personalized as possible within the lens of your new customer segments so that the sales reps are getting that experience.

Examples of segmentation rollout in Customer Success

• Capture information through in-app onboarding to inform messaging – present different product experiences based on the information the customer provides. For example, if your two main verticals are customer success teams and marketing teams, but I sign up and say I’m on a marketing team, the communication sent to me should acknowledge that I’m on a marketing team instead of it being generic. 
• Begin meeting expected service experiences – if your customer segment expects that they’d be able to self-serve through chat, but you only offer phone support or email support at the moment, then you’ll want to meet with your customer success team to discuss how you can provide a support model that better fits with the customer segment.

Examples of segmentation rollout for Product

• Build roadmap around segments – be clear that this feature or this product is being built for this segment to drive this impact. That makes the product team feel empowered because they know exactly who they’re trying to solve the problem for and then they can find the best solution to that problem. It also  makes it so much easier for everyone else across the organization to know why the product team is investing in that, who it’s for, and where it will impact.

How long does the process take (the research, and the roll-out to start acting on segmentation in sales and marketing?

Research phase length varies based on resources – as quickly as one quarter if you have one product manager or product marketer who can commit a full-time effort to this project. It could take you 6 months though if it’s something you’re half investing in or if it’s a side-desk project. 

Roll-out is continuous – Once you have your customer segments, you should always be rolling out strategies that are focused on those customer segments. But look for quick wins early, especially if you’re the product marketer. If this is the work that you’re bringing forward, you want to actually start to see those ratios going in the right direction as your customer base becomes weighted more toward the target segment(s). 

How does segmentation impact pricing and packaging?

Pair segmentation with a willingness to pay study – you have a couple of options here. You could do a conjoint analysis, which is quite complex. Or you can do a Van Westendorp, which is a much simpler model that involves surveying buyers about the points at which a price is suspiciously cheap, a good deal, getting expensive, and prohibitively expensive. If I do the Van Westendorp, I also pair that with a MaxDiff kind of feature preference study as well.

Examine your value metrics and feature signals – are you pricing your product based on feature access or is based on a usage metric (or both)? Is that a metric that your customers see as being meaningful for their business?

You might create one price book with different packages for different segments, or you might create multiple price books – if one price book, you’ll want your different packages to resonate with particular customer segments; it’s going to be clear who that segment is, and pricing is going to reflect that segment’s willingness to pay. Or, a great example of the multi-price book approach is businesses that support both companies and agencies, and have completely different price books for each.

Pricing should be iterative – this work that you’re going to do with your customer segments will give you a really great signal, but only once it’s out in the market and live with these customer segments will you actually be able to see how it performs and what pieces of it need to be iterated on. So it’s just part of that evolving, go-to-market strategy for each of your customer segments.

What tools and resources make segmentation easier (software tools, research tools, internal documents you create, etc.)?

Excel – for the data model slicing and dicing.

A tool that supplements your customer data (e.g. Clearbit) – you can place a lot more information into your dataset with a tool like this because, as long as you have something like the customer’s email address, then you can supplement things like, what vertical are they in, how large is their business, how much revenue does their business run, especially in B2B. That’s what Clearbit specializes in. 

Internal docs

• New positioning docs – frameworks for every customer segment is critical to making sure that all the go-to-market teams are aligned. 
• A segment profile one-pager – a quantitative buyer profile including metrics and key pieces of performance data from each of your segments. Also includes the customer’s “job to be done” that they want along with what the pains and gains are. 

How does your segmentation evolve?

Continuously develop a deeper understanding – I love to use the analogy of the game of Risk because most people have played it. It’s about world domination. The segmentation business is kind of about market domination. In Risk, the number one thing you want to do in order to win is win one part of the board and fortify your bases there so that you can start to get more resources. Most people will choose Australia as the first place that they want to own, but you’re never going to say “now I own Australia, I’m going to go after North America” because those are on completely two different ends of the map. You’re going to say, “what are the territories adjacent to Australia that I can start to make one move out, and then another move out.” 

Evolve segmentation to enter adjacent segments once you “own” a segment of the market – customer segmentation is truly the same way as the strategy in Risk. You want to start to think that once you own one segment of the market, what are the adjacent segments of the market you could own which wouldn’t require a complete shift in your go-to-market strategy? 

What are the most important pieces to get right?

Get buy-in from the beginning – make sure that the business is onboard with conducting a customer segmentation. If the C-suite leaders or different stakeholders aren’t willing to change their business strategy around this customer segment, it’s going to be wasted and there’s going to be a lot of frustration.

Take a data-backed approach – leave your ego on the table because this isn’t about opinion or different people’s biases. Those may drive some initial hypotheses that you have going into the exercise, but you have to be willing to let the data speak to you and show you what patterns are going to emerge. 

Talk to customers – you can do the entire process with just the data, but there’s probably something along the way that you’ve missed. Why does this customer segment convert better than the other segments or why do they stick around longer: getting those nuggets of gold out of the customers are going to be key to being able to roll out your go-to-market plan effectively. 

What are the common pitfalls?

Don’t just turn up the volume – don’t stop after step 1, “Measure Volume”. You haven’t validated that this segment actually performs well yet and you haven’t validated that you can win more with this segment then your average customer yet. 

Don’t wait until the end to bring in your stakeholders – you want to make sure they understand how they’re going to be able to use this and how it’s going to help them reach their goals. 

Don’t get too stressed about the dataset – especially in the beginning, any data is better than no data. Some companies may have 100 columns of data while some may have 5, and that’s okay. Just starting with some opinion about your customer segment, you’re going to be better off than the average business who doesn’t even think about this.

Don’t have the data team lead the project – they don’t always have the market context, product context, or customer context to put it all together. Making sure that you actually have the right team on this with the right experience level is really important too.