Frankly - if someone runs an e-shop today and doesn't address customer loyalty, they are throwing money out the window for nothing. Not because he's a bad marketer, but because he's playing without one of his strongest cards. They say the most expensive customer is the new one. And yet, most e-tailers are pumping their budgets right into acquisition instead of holding on to the most valuable thing they already have - their own customers. Want to know the key to higher sales and lower costs? It's called RAFT.
No, we are not floating on a wild river, but we are floating on Recency, Frequency, Amount and Total data. This foursome helps you segment your customers by behavior and get the most out of each segment.
What is RAFT segmentation?
RAFT is a clever method of segmenting customers based on how much (and how often) they actually do business with you. Every letter has a role to play:
Recency - when did the customer last make a purchase? The sooner the better. A fresh customer is like a warm roll - ready for the next purchase.
Frequency - how often does he shop with you? A regular customer brings you money repeatedly and practically for free.
Amount - how much does he spend? One big purchase is more than ten purchases for a few pennies.
Total Value - the sum of all his purchases over 3 months is at least X times the average order value of the e-shop.
Typical customer groups according to RAFT
Because the "average customer" is a marketing myth. You have the champion who spent 10x more than everyone else, and then the silent ghosts who ordered once and disappeared. Interacting with them the same way is like packaging all customers on the same catchphrase. It doesn't work.
This is where RAFT makes sense. Instead of one bunch of anonymous contacts, you have clearly defined groups:
Segment
Behaviour
Recommendations
Champions
Purchased recently, often and for a lot.
VIP program, early access, ambassador program.
Loyal customers
Regular, loyal, spend steadily.
Retention emails, loyalty bonuses, stable care.
Customers with potential
They shop occasionally, they have moderate spending, there is a growing interest.
Special offers, motivational emails, upsells.
New customers
First purchase, low spend, shortly after registration.
Welcome series, onboarding, first discount, education.
Customers at risk
They used to buy more, but they're stopping.
Reactivation campaigns, personalized reminders.
Lost customers
Not bought for a long time, very low activity.
re-engagement campaign (discount coupon, personalized reminder), analysis of the reason for leaving.
Example: who is a "Champion" according to RAFT?
This customer is not just active. He's the key. And here's how you'll know:
Recency: last purchase made 7-14 days ago
Frequency: made 5+ purchases in the last 3 months
Amount (AOV): its average order is at least 2 times higher than that of a regular customer
Total Value (TV): the sum of all his purchases over 3 months is at least 5 times higher than the AOV
For example, if the average order with you is 1 000 CZK, then champion:
Spend an average of CZK 2,000 per order
And a total of 10 000 CZK in 3 months
Now ask yourself the question: What are you doing for this customer? If the answer is "we send him a newsletter like everyone else", you're wasting gold.
Practical use of RAFT segmentation
When you have RAFT segments, it opens up a whole new world of possibilities:
Targeted retention campaigns: e.g. "Buy again this month and get double points" - only for those with high A but declining R.
Behavior-based email automation: e.g. "Your points balance has dropped below 500, use them before they are forfeited!" - for those with high T but low F.
Discount optimization: don't give coupons to everyone. You give them to those with high Amount but low Recency - and make them come back.
Why segment? Because the average customer does not exist
Two people bought you a thousand bucks worth yesterday. But one of them is a loyal VIP who has already spent 50,000 CZK, and the other is a random passer-by. Will you reward them equally? I hope not!
Retention campaigns without segmentation are like throwing flyers out of a helicopter.
Segmentation, on the other hand, allows you to shoot sharp - right on target.
How much can loyalty marketing earn you?
The data from Ellita shows clearly:
+30% of orders for clients with an active points system
+26% conversions after deploying VIP levels
12% of revenue comes from point rewards
Now put that into your numbers. If you have an e-shop with a turnover of 2 million. CZK a month, then 240 000 CZK of that can be just because you rewarded people for their loyalty.
How to do it: practical implementation of RFAT
Collect data - without data there is no segmentation. Connect your e-shop with Ellity and get: last purchase date, order frequency, total spend, engagement score.
Calculate the score - for example, using the formula:
S = wR⋅Rn + wF⋅Fn + wA⋅An + wT⋅Tn
Each metric is normalized (e.g. Recency 0-1, where 1 = purchase yesterday). You can adjust the weights (w) according to the business - for example, frequency will be more important for fashion, amount will be more important for electronics.
Divide your customers into segments - and set up automated campaigns for each. This is where emailing tools linked to Ellity (Ecomail, Smartemailing, etc.) come in handy.
Test, optimize, iterate. RFAT is not a fixed model - tweaking weights and thresholds will give you a more accurate hit.
Loyalty is no accident. It's a system.
Customer loyalty is not about hearts and emotions. It's hard data + smart rewards + the right timing.
If you don't build customer loyalty today, you pay for advertising over and over again and end up overpaying new customers who will run away anyway.
You want it different?
Start by understanding who your champions are. And make them ambassadors. The rest... Ellity will automate.
Start an effective loyalty system for Shoptet today that will work for you forever.
.svg){kind=small}
