Segmentating a customer database helps identifying groups of customers which, by having the same behaviours or characteristics, could be addressed the same way in terms of sales and marketing actions. In e-commerce, this is what can make the difference between being a good brand and being a customer-first great brand. Here are some tips on how to do a customer segmentation for an ecommerce site.
You should also read our central page on customer segmentation before reading further.
Customer segmentation for e-commerce: data comes first
All good customer segmentation starts with a good dataset. This sounds obvious, but this should really be the starting point and area of focus. First question would be: how do you define an active customer from a churned one? When do you set the limit to define what is a churned customer.
If you sell music instruments online, your customer might purchase several times a year equipement and consumables such as strings for the guitar. But what is you specialist in selling pianos online? How often do you buy a piano online? Would you consider that if a customer buys a your shop once every two years would be a statisfying return rate?
Use a cleaned and enriched dataset
Datasets must be cleaned. Have all the customer IDs you are about to use been deduplicated? Are you sure you don't have the same John Smith having two different emails but living at the same address? How can you make sure to deduplicate your customer database - and be certain you don't send emails to the same person with different messages?
Companies like Amazon or Google invest a lot of ressources making sure their customer database (or user database) is as exact a possible for obvious reasons.
Upload a CSV now and score your customer using the RFM (Recency, Frequency, Monetary) methodology
Getting the best enriched customer database should be a priority. While on brick and mortar companies, you might know every single person who comes often at your shop and spontaneously offer them some goodies or have a chat, this is not possible on e-commerce websites. Segmentating your database and knowing your customers is key to success.
To achieve customer excellence, good news, you could log every additional information possible: what are all the pages a customer saw before purchasing? Did he see your latest sales actions or not? Is he browsing product categories where he never purchased before? All this information could make a difference in the messaging when you will address by email your customers.
How to do a customer segmentation for e-commerce websites?
Behavioral vs demographic segmentation
Build a behavioral segmentation to categorize your customers, not based on their caracteristics but on how did they interact with you in the past. Behavioural segmentation is generally one of the most powerful one to drive sales. Getting demographic information on customer not only can be tricky and expensive to get but could lead collect sensitive information at GDPR times.
RFM segmentation for eCommerce
Recency - who are the customers who purchased recently? From your experience, are these customers likely to re-purchase soon - or should these customers not be addressed before at least six months? What is the customer lifecycle pattern of your online activity? How would you describe your customer's journey if you had to pitch it? A customer hasn't purchased since 6 months? By getting this information, you could push a message such as "Hey, we miss you... discover all the new products and our new collection. We offer you a $10 voucher if you purchase again with the next 7 days".
Frequency - How often comes a customer and purchases at your e-commerce website? Does this customer purchase once every year - or every quarter? It sound abstract but a high frequent purchaser might be the plumber going to its supplier every week to purchase all the materials he needs for his 1 or 2 current building sites. He will look for recognition. When he comes to your sites, you need to show him you know him quite well and get the right level of customer care.
Monetary - how much did these customers spent. Maybe the easiest customer segmentation criteria to apply for an ecommerce customer database. The only question should be: how do you define the value of a customer? You should withdraw some additional metrics, like the value of all the returns a customer made and the associated shipping cost for you. A customer with the habit of buying 1 article but shipping also 10 articles back don't have the same monetary value than a customer who bought 1 but never sent back any good.
Keep it simple
However, the combination of these customer segmentation criterias can be complex. Segmenting can become a complex exercise - always keep it simple and get a pragmatic approach when you choose and combine the customer segmentation criterias. Based on what you know of a customer - How would you address him to trigger an additional new purchase? The complexity of segmentation lies into the combination of all the parameters possible.
Datainsightout can help you do your eCommerce customer segmentation. You should have a look at our product page as we propose such as tool at a very affordable price. All you would need to use it will be to prepare a CSV of data and upload them on the platform.
What are the specific additional customer segments criterias to apply for an e-commerce website?
Long term segmentation helps driving sales and customer engagement. However, additionally to it, you should use a tactical customer segmentation for your online shoppers. Some profiles will be very interesting to drive short-term tactical sales actions. Remember: always go to the most efficient and simple tactics.
Boost your sales via tactical customer segmentation
Card abandoners. This is less a segment definition to use for long-term strategy, but definitely a good one to boost sales volumes. Sometimes, the customer stopped going further for obvious reasons: he might have been interrupted, he might want more time to think twice or he plans to finish the purchase later during the day. Card abandoners should be addressed, but you want to avoid cannabilizing sales which would have anyway happened without a voucher. Take a moment to think and define what is the right moment to target card abandoners. Should it be 3 hours after having added to the basket without a checkout associated? Or 2 days after? You should also flag the customers who already benefited from it - to avoid being tricked.
Hardcore coupon users are typically all your customers who purchased only using coupons or vouchers. These customers' purchases have been triggered by a price cut or a discount. This is even more if you see that a customer made several purchases and all of them with a coupon redeemed. You know what you need to do to generate a sale with this kind of profile. They might not be your most loyal customers, but if you have an interesting price reduction or a limited in time very good deal, then this segment of customer would be an ideal fit to target.
"Likely to" users are customers who, from what you know about where they live, what they purchased, might be interested into another similar product for the same category. While this can get very complex in terms of data analytics, easy things could be made. If you know your customer lives in a city near the sea, you could target them to showcase your latest umbrella beach collection in stock. If you know that a customer purchased a lawn-mower 6 months ago, it should be fair to assume he lives in a house and has a garden. You could invite him to discover your garden shed category, especially if you have interesting offers.
To wrap it up on e-commerce customer segmentation...
How to do a customer segmentation for eCommerce? It doesn't need to be complicated but it requires a solid and cleaned dataset for each of your customer. The more you log information about your customers and prospects going to your shop, the better.
You could then combine a classic behavioral customer segmentation with additional tactical segments. You could then define scenarios and trigger automated actions in your CRM system to make them even more engaged and avoid them to churn.