Customer segmentation is plural. You have multiple approaches when you want to build a customer segmentation. But remember: the most important is to use criteria which you think are important to drive your sales and engage your customers. You should always use a combination of the following criteria. But first, why don't you read our central post on customer segmentation?
Attitude is key to segment
In a behavioral segmentation, you just focus on the intents and the past interaction of a customer.
RFM segmentation is based on the recency, frequency and monetary characteristics of each customer. It goes in the past, and measure how often did they purchased, how much they spent and when was the latest time they shopped. It's based on the attitude of the customers.
Large / Medium / Small segmentation
A very simplistic version of it would be to segment your customer database into the following 3 buckets, just based on their spendings:
- Small (bottom:80% in numbers, but 20% in value) - this is the long tail of your customers.
- Medium (top: 15%) - you need to make them grow and upgrade them to the Large/elite segment.
- Large (top 5% but maybe contribution to over 40% of your total turnover) - the elite customers, those you want to keep and retain at almost any cost and they generally make.
Behavioral segmentation focusing on customer engagement thresolds
One very - which would be also a consequence of the behavioral segmention - would be to have a proper focus on the active or churned status of each customer. In the middle, you can have all flavors:
- Highly engaged customers
- Engaged customers
- Adopting customers
- New joiners
The "golden path" to customer engagement
This information has tremendous value. You might want to define a golden path to engagement. For instance, companies such as Facebook might want a customer to perform the following actions within their first 30 days post activation:
Upload a CSV now and score your customer using the RFM (Recency, Frequency, Monetary) methodology
- Upload a profile picture
- Add at least 10 friends
- 20 likes and 3 comments
- Come 1 day out of 2
This golden path will entirely be grounded on behaviors and quantitative interactions you want your customers to do. For an eCommerce website, that could mean: "I want 70% of my customer database to visit at least 3 pages on my website within a 6 months timeframe". Once the golden path to engagement is defined, you need to align your marketing activities.
All relies on behavioral segmentation as you can see.
Imagination is the limit: add as many dimensions as needed
You can add all the criteria and metrics you might find relevant for your case. You run an online shop and want to do some segmentation? You should then to the behavioral segmentation the following metrics in front of each customer:
- Average duration time on site
- Last login date
- Number of pages viewed on average per session
- number of sessions per month
- Inbound clicks from email campaigns
If this information helps you scoring your customers and target them better to understand their needs, then this is helpful.
You need time to build a behavior-driven segmentation
Behavioral customer segmentation need time to be built. You need to log enough data over a long period of time. At least one year. You should not lose any time and log all the data you need!
Don't rely only on purchases - add as much as possible information: how many times a customer spoke to a sales person offline (physically in-shop experience)? Any frequentation statistics around the website usage, social network information and so on.
Caracteristics of the person
This customer segmentation is based on who the customer is. Purchases and interests won't be the same whether the customer is a man or a woman, if the customer is a millenial or a boomer.
What is their education? Did they went to college, have a bachelor/master degree or not? Education can also be a good predictor of the disposable income of a customer, educated people tending more to get a higher income than the rest of the population. This information can be used further as a scoring dimension.
Life cycle stage and personal situation can be interesting to get: is the customer married / divorced / with children or senior ? Based on this information, your marketing speech and products you want to sell and cross-sell might be totally different. The tone of voice you will use in your communications might also be different. This is actually quite an important and interesting information if you can get it.
Protected information - handle with care!
This information must be however protected. You should pay attention when using it and check with your compliance department what you can store and how long. This is depending on the country / region where you live.
You might want to store this PII (Personally Identifiable Information) is seperate databases and grant access to only a handlfull of trusted people in your organisation. GDPR makes it clear that personal data should be protected at any price.
Is a customer rich or not?
The disposable income information can be interesting to log as well. How to know the average income of a customer? You already know their address. By enriching their living location with external databases, you might assume their socio-professionnal category. Some part of a city or street are known to be richer than others. You shouldn't log with precision their actual true income - this information is only available to the IRS - but you could establish an income scoring. Building a score can be a good methodology to assess the purchasing power of an household.
You might ask them during the checkout to fill-up a field form with their professionnal occupation. However, this is declarative and might not be the most reliable source of information. It also depends on the nature of your business: If you sell clothes online, your customers will also find it weird to be asked this question. However if you sell online insurance online, this is a pretty legitimate question to ask your prospect and customers.
Tell me where you live...
Some of the information is relatively straightforward:
- ZIP Code
But also all the contextual information: does the customer live in a house or in a flat? You might also want to know more about the location; information such as the public transportation around them (if you sell car tyres), their average time to commute, do they live next to lakes, mountains or beaches (if you sell outdoor sport equipement).
You might also want to capture if where the customer has many bars and restaurants at proximity (E.g. for a food delivery app).
Psychographic customer segmentation
Psychographic might be the most abstract and less data-driven type of customer segmentation.
Psychographic segmentation is based on very subjective information such as :
- Values and beliefs
- Attitude and psychological mindset
- Tribes and belongings
- Life motivations and goals
- What do they fear or want?
- Social status and social perception
These psychological characteristics can vary over time and days. They are not easy to collect and to assess. You might not be able to use them to drive short-term sales actions. However, when you build a personas you might want to conduct a serie of customer interviews. In these interviews with your customers, you will be able to assess this information. Psychological customer segmentation can help building products and know better who are your customers in the long run.