One of the most resonant themes that we saw coming out of the Strategy Fest was how important understanding your customer base actually is. As strategists, we can often get caught up with preconceived notions as to who our ideal buyer is, and it doesn’t always match up with the real customers that are coming through the door. As such, dedicated customer knowledge projects that aim to collect this sort of information and let that inform strategic thinking are extremely important.
But when you do these customer audits, you’ll likely find a lot of diversity throughout the customer base – that needs to be taken into account. It’s natural that you see certain patterns in the data and if you don’t understand how that should impact your operations, you can very easily paint everyone with the same brush.
This is the exact point that Tamara Grominsky made in her presentation, where she championed the power of good customer segmentation. And in this article, we’re going to delve into this particular concept and see how it can be leveraged to push your business forward.
Customer Segmentation is Function Agnostic
Before we start to discuss how to get the segmentation right, it’s worth emphasizing that this segmentation is important for everyone throughout the organization. It’s a common misconception that this is simply a marketing and strategic convention – because that’s where these discussions typically sit. But these groupings actually affect every single part of the organization and if your teams don’t understand the intricacies, they aren’t able to optimize processes and adapt their decisions accordingly.
“Confusion across teams is not only creating disjointed experiences, but it’s also creating friction among stakeholders.”
This is why this segmentation needs to come from the top. As strategists, we should be identifying these segments and then communicating them down through the lines of command so that everyone can pull in the same direction. Our goals and objectives should be built around them. Everything that we do as a company should be centered around these segmentation decisions so that we are meeting customers where they are and adapting our offering accordingly.
Now that we’ve made that point clear – let’s dive into what it means to segment your customer base effectively.
What is Customer Segmentation?
The basic concept refers to the act of splitting up a customer base into segments based on similarities and differences within and across different groups. Shared characteristics tend to group different types of customers together, and that can then lead to more personalized offerings for each group or segment.
The idea is that each customer segment can then be treated slightly differently, in ways that accentuate their unique characteristics. These small optimizations can make a big difference because the company can serve different interests, even within the same organizational framework.
It’s not just the marketing that can shift (as we alluded to earlier), but also things like the customer service, pricing, operational prioritization, product line, and more. As companies grow in size, these segments can even take on lives of their own, demanding their own teams, technology, and performance evaluations. In essence, they can become little companies all on their own.
All of this value, however, is predicated on the quality of your segmentation itself. You need a good understanding of your client base in order to do this well and you have to be able to differentiate between real, valuable differentiation and red herrings that actually don’t impact the buying decision. For example, geographic differentiation might not matter at all if your business offering is primarily online. In this case, you might be looking for something else that directs changes in user behavior.
Another key decision to make is how granular you want your customer segments to be. When you have too few customer segments you aren’t optimizing your potential adaptations, and when you have too many, you can drown yourself in too much complexity. The right answer for this is going to be different for each company, but it’s worth keeping a skeptical eye on this for yourself so that you don’t get carried away in either direction.
The framework below should help you get this right and often you’ll find that the segments arise from the data on their own – without you needing to pick a number in advance.
Speaking of, let’s now look at Tamara’s 3-step process for world-class customer segmentation.
The MAP Strategy
Tamara shared with us her acronym for creating leveraging the value of customer segmentation: MAP. Each letter refers to a step along the process and if you stick to these three, you’ll be in a good position to get this right.
Measure of Volume (M)
The first step is to analyze your data and identify the key segments that you want to create. This is where a lot of the work needs to be done because everything after this relies on your decisions here. As such, you’ll want to make sure that you’re tackling your data with the requisite care and nuanced understanding.
Go through whatever systems you have internally and pull as much customer data as you can. In these initial steps, you want to gather as many different variables as you can because you aren’t sure which ones are going to be valuable yet. You might think that some information is not helpful but until you go through the process you can’t be sure of that. A seemingly unimportant detail about a customer could potentially be combined with others that in totality form a useful data field.
Once you have all this data you need to go through it and clean it as best you can. Look for gaps and see if you can augment those with representative data. Reach out to other teams to access missing information. Anything you can do to complete the dataset will be valuable because it's going to make your segmentation much more precise and up to date. And even if it doesn’t directly affect the groupings, it’s still a valuable exercise for the company as a whole. These databases often trend towards chaos because of how chaotic day-to-day operations are, so any steps you can take to rectify this is going to be worth it.
As you’re going through this process, it’s worth noting down any potential weaknesses you identify in your data collection processes. Perhaps there is a weak link in the chain that is not receiving the focus it should be receiving. This data audit exercise helps to bring those to the surface so that you can improve things for future usage. As the business world shifts more and more towards data-driven decision making this becomes a crucial asset to look after. Note down the system improvements that can become lead dominos for improved data warehousing, processing, and analysis. Thus, the segmentation process is easier every time you do it because you’re improving your data as you go.
The final data set that you want to arrive at is one where you can slice and dice effectively. Take extra care to ensure that your data types are accurate, your labels are congruent, and you aren’t missing anything that should be. This part of the process is not fun and can be quite tedious, but it’s crucial for the end objective. The better the quality of your data at this point, the better your segmentation is going to be.
Now that you have your data, you can start to analyze it. You’re looking for large clusters that share specific attributes. Here’s where the ‘measure of volume’ part comes in, you want to find groupings that are actually statistically significant. It doesn’t help if your groupings only have a few customers in them. You want to identify combinations of attributes that arrive at relatively large groups at first. For example, perhaps you find a combination of attributes that sorts the customers neatly into three different clusters. These become your segments and you can break those down further if you wish and if you find enough differentiation to do so.
The act of creating these clusters can be accomplished in two ways.
- The first one is where you use trial and error, playing with filters until you hit upon what you’re looking for. This is less precise perhaps, but it doesn’t require much statistical expertise and is often the most efficient way to do things if your customer dataset is not that large. For many small and medium businesses, this can work great. The only thing that you have to keep in mind here is that you’re likely to be skewed towards those attributes that you personally think are true differentiators. This may be the case, but it is also very possible that there are other combinations that you underestimate. Take your time and try as many different options as you can- you might just be surprised at what you discover.
- The second way is to use the power of statistical analysis and machine learning to identify these clusters through algorithms. There has been a lot of advancement in this field in recent years and if you have some of these skills in your team, then this is the way to go. It will give you the most precise clustering possible and it can identify combinations of attributes that you might never have come to on your own. For large data sets, this is often the only way to segment customers efficiently. We won’t go into detail here because the mathematics behind these algorithms is a science all on its own. But safe to say, machine learning is perfectly placed to perform this sort of clustering in a way that’s never been possible before.
Now that you have your segments, it’s time to plan how you are going to respond.
Analyzed Performance (A)
The next step is to take your segments and assess their performance. To do this, you want to group your customers into their segments and then calculate the past order data for these groups in as much detail as you can. Pull out as many metrics as you can to get a sense as to how these segments have performed in the past.
You should then look to compare these metrics against one another to evaluate each segment on its merits. You’re looking for potential insights or opportunities that can be leveraged to improve overall business performance. For example, you might find that one segment lags behind the others, which points to the fact that you might have room for improvement when it comes to targeting and servicing that particular cluster. Or you might find that one segment is performing incredibly well, and you might be able to scale that grouping or increase margins in that area to take advantage of that strength.
There is a myriad of ways to analyze this performance, but the key is to take those insights and build them into your strategic decision-making going forward.
That brings us to the last step of the process.
Prioritize Potential (P)
With these insights at your fingertips, you now need to put them into action. It’s only through action that this whole exercise can actually make a difference in your organization. So, in this step, you want to prioritize those areas where you see potential.
You only have limited resources and you want to make sure that any strategic decisions that you’re going to make are worth the efforts. We’ve seen many companies be overwhelmed with this segmentation process and as a result, they become paralyzed by the information. You have to fight against that but breaking those action plans down into smaller steps that you can tackle one by one.
The prioritization enables you to create a roadmap where you can slowly make the necessary improvements in a systematic and scalable way. By breaking it down into components, you make it that much easier to execute. And as an organization, it gives you a game plan to work from.
This alignment is important because, as we mentioned above, this needs to be an enterprise-wide exercise. It’s imperative that everyone understands the priorities, and knows how to execute them in their specific functions. The way you do this is through carefully considered communication that must spread throughout the organization. These lines of communication need to share the insights that were discovered in the segmentation process and create a single source of truth that everyone can pull from, regardless of their position.
Tamara mentioned that some organizations have done this through what they call ‘Voice of the Customer’ meetings which we think is a really cool way to do it. In these meetings, you try to embody your customers as best you can, as if they were there with you. By going through this thought experiment, you can encourage team members to really empathize with the people that they are serving and keep them front of mind as they go about their day.
Whatever it takes, do your best to democratize decision-making, but keep it aligned with the segments you’ve identified and evaluated. When you do this well, your company can fine-tune its marketing, operations, and long-term strategy to better serve a diverse customer base. It’s magical when it works.
In summary, customer segmentation is not just a nice-to-have. It is a crucial concept that any business can use to reorganize itself in a way that can serve different customers in a personalized way. When done well, customers feel like you have a much better understanding of who they are and what their needs are, and your back-end infrastructure can be set up to make everything as smooth as possible.
Good segmentation requires a solid grasp of your company data, objective clustering, and nuanced performance evaluation. And once you have your clusters and the associated insights, that needs to translate into action. It needs to move the needle and inform everything else that the company does.
The only way you can achieve that execution is through a company-wide effort that understands the segments and is incentivized to ponder those differentiated characteristics as they make decisions. When all this comes together, you have a powerful engine for business growth, and you can completely transform how things work behind the scenes.
“You can start to see that instead of all of those teams focusing on slightly different customer segments. When you have these artifacts and these rhythms of business that really keep everyone connected, you're all going to start to focus and obsess about that one customer segment that matters most to your business, and what this is going to do is it's going to drive focus across teams and build momentum that you can't even imagine.”
This way of thinking has given us a lot of food for thought here at Cascade, and we hope that it inspires you as well to re-examine how you think about your customer base and bring the entire process into the modern, data-driven world that we’re all trying to cultivate.