The direct impact of customer segmentation on ROI

Neeraj Agarwal
4 min readMar 6, 2021

Thousands if not millions of people visit your website every week adding to your customer base. However, a tiny chunk of that customer base takes an action that directly generates revenue for your company. You accept the fact that you will have to spend insatiably on marketing and branding without really hoping for tangible results. You learn to accept the risk and agree to shoot an arrow in the darkness with a slight hope of actually hitting a target.

The intangibility of the return of investments made in marketing does not mitigate the risk of losing huge amounts of money for nothing. Things should not be like that for your business in a world governed by the rules of physics, statistics, and psychology. We shall look at one of the many ways of shaping your marketing and customer engagement efforts and try to figure out how it impacts revenue.

Customer segmentation

The simple concept behind customer segmentation analysis is dividing your customer base into relatively small groups based on shared characteristics in order to form and apply better directed marketing efforts.

For instance, you can plot your customers on a graph based on their age and income. This sort of demographic based segmentation is a little crude, to be honest, but it gives you a foundation upon which you can work further to ensure better and more profitable segmentation.

Clustering algorithms can help you divide your customers based on age, gender, location, and purchase history, under certain labels. There are always some valuable outliers who do not fit into a customer segment, hence require a separate labeling.

Customer micro-segmentation

Micro-segments, as the name would suggest, represents much smaller groups of customers based on niche characteristics and psychological traits. If a segment represents married women between the age of 30 and 45 who are living in a metro city, a micro segment would add features like spending habits, value systems, favorite sports personality, alma-mater, you get the idea.

Creating customer micro-segments is harder in terms of technology implementation than traditional demographic segmentation, but it is worth the effort. 70% of customers prefer personalized advertisements. Personalized emails have a 25% more chance of being opened.

Any qualitative research focuses more on individual behavior and intentions behind actions rather than just physical information like location, age, and race.

Segmentation to strategies

Grouping the customers into segments gets half the job done. The other half entails targeting the right groups with the right message and positioning your business in a way that helps your most valuable customer segment relate to your product the most.

Who buys what from where under what circumstances? How often do they make a purchase? Where on the website do they spend the most amount of time? From where do they leave it? Who abandons their carts? Who clicks only on discounted items?

You need answers to these questions and form relationships between these answers and the characteristic features showcased by the corresponding customers. All of this functions like a chain and you end up having a marketing strategy with a somewhat tangible promise of return of investment.

The great gamble of assumptions

Well, you can ignore all of what you just read and continue with your business completely oblivious of your customer relationship metrics. You can act on assumptions and hold instinct over statistics. It has worked for enterprises before you and might just work for you as well. Remember, our brain is a machine that identifies associations and relationships quite effortlessly and we call it instinct.

There is however, a huge but involved in this approach. Your assumptions work for people whom you think you understand. But it is quite hard to update your understanding of people with time. Each person out there is exposed to different influences; value systems are always in a state of flux; some spend with a plan, some out of a whim.

The point is, the world around you is changing faster and in more ways than you can possibly process. Why put the time and devotion in such a gamble, if you can deploy some neural networks that can perform the associations faster, better, and more reliably than you.

Getting back to ROI

With the help of a robust customer segmentation tool you direct your marketing campaigns better; spend more wisely on your email campaigns; Lay out content that really speaks to your customers; save budget for a plan B; Focus on your most valuable set of customers. This gives your strategy an element of tangibility and yes you are more likely than ever to get a much better return on investments.

--

--

Neeraj Agarwal

Data Science | Big-Data | Product Engineering @ Algoscale