How to increase repeat sales: tuning up RFM analytics
The analysis itself based on three characteristics:
- Recency - how long ago the last purchase was made
- Frequency - how often customer purchases
- Monetary - customers average bill value
Referring to RFM-analysis settings here everything depends on what recency, frequency or monetary is concerned to be high or low in your business. You set necessary value to each of the segments and send widget to re-count. RFM-analysis collects data all over the base for the whole time and doesn’t depend on the time limits set in your global analytics.
After the recount is made you will have the data of current importance. If you tune up widget for the first time the process will take several minutes.
More about how it works and how to set up the activities in detail.
All the sales volume in retailCRM RFM-analysis module is taken for 1. Customers who made purchases in your shop are sorted in descending order of the sum of those purchases: the most active are at the top, the least active - at the bottom. Further, the system sums them one after another.
For the Monetary calculation process to be clear let’s take total sales volume of the N shop for 100 euro. And the customer base will consist of 4 clients who purchased for 60, 20, 15 and 5 euro.
1) 60 / 100 = 0.6
2) (60 + 20) / 100 = 0.8
3) (60 + 20 + 15) / 100 = 0.95
4) (60 + 20 + 15 + 5) / 100 = 1
We found out than N shop would need to set the limits about 0.7 and 0.98 for the first customer to get to the 1st segment, 2nd & 3rd customers to get to the second segment, and 4th - to the third segment. There are lots of customers in real shop, that’s why these borders are closer to the beginning. The defaults in retailCRM are 0.2 and 0.35.
Here you set breakpoints according to the frequency of your customer’s orders. To get these numbers enter retailCRM Customer section and sort by number of orders.
For example, customer A has the biggest amount of orders equals 10, and customer B has the smallest equals 1, and there are also customers C, D and E, who have 8, 6 and 3 orders accordingly.
In this case the number of orders in the limits between «Frequent» and «Mid» equals 7, and between «Mid» and «Rare» - 4. Under these circumstances customers will be analyzed as following:
Number of orders more than 7 - Frequent
Number of orders less than 7, but more or equals 4 - Mid frequency
Number of orders less than 4 - Not frequent
The remoteness index of the first purchase is unique in each business. That’s why set the period of time from the moment of the last purchase, which will be proper for each of the metrics (mid-remote/recent).
How to use it in practice?
Let’s have a look on specific examples and take floristic business for that purpose.
Let’s examine the following situation:
You have X customers, who buy flowers once in 2-3 months for private holidays. Their average bill values are 50-80 euro. Under these circumstances they are in that point:
If some percent of them quit buying with same periodicity these customers move here:
What can be done?
A distribution with favourable offer limited in time (expiring discounts, season offers, etc.)
Situation № 2
A VIP-client who buys bunches of flowers for more than 120 euro 1-2 times per month is here:
If he decreases the number of orders he makes he moves here:
What to do?
Trace this event and make a notification for manager which can figure out the situation and return the customer.
For different reasons N customer shifted a position down (from high importance to mid, from mid to low, etc.).
Solution: Trace and react with sending a personal offer (e.g. second item is cheaper for N% or third - as a gift).
In fact, you can tune up the reaction for every customer enter/exit from one RFM segment to another. And use it for distribution or making tasks for managers. Or entrust this work to our team, we’ll make it on a turnkey basis.