Recency vs Frequency vs Monetary Value: Which Customer Signal Matters Most?
A customer bought yesterday.
Another bought eight times but not for three months.
Another bought once and spent $1,500.
Which one matters most?
That depends on the question.
This is why recency, frequency and monetary value should not be treated as interchangeable.
Each tells you something different.
Recency tells you relationship warmth
Days since last purchase.
Useful for:
- lapse,
- purchase readiness,
- deterioration.
But recent does not mean valuable.
A new one-time buyer has perfect recency.
Frequency tells you relationship habit
Number of purchases.
This can be powerful because repeated behaviour demonstrates that the customer chose the brand more than once.
But high frequency at very low value may produce less contribution than you assume.
Monetary value tells you economic scale
How much has the customer spent?
Useful.
But one large purchase can create high monetary value without loyalty.
Shopify combines the dimensions for exactly this reason
Shopify’s recency, frequency and monetary-value groups use recency separately and then combine frequency and monetary scores to place customers into relationship groups such as Loyal, At Risk and Previously Loyal. (help.shopify.com)
No single dimension is enough.
Which matters most for reactivation?
Often:
historical frequency + current recency decline
is particularly interesting.
Why?
The customer proved a habit.
Then the habit broke.
That gives you a clear change to investigate.
Which matters most for VIP strategy?
Monetary value and frequency usually matter more.
But recency tells you whether the VIP is still active.
Which matters most for a new customer?
Recency tells you very little.
They all purchased recently.
You need:
- first product,
- order value,
- cohort behaviour,
- predicted future value.
Example
Customer A
Yesterday
1 order
$30
Customer B
90 days
10 orders
$1,000
Customer C
20 days
1 order
$1,200
For onboarding:
A.
For recovery:
B.
For high-potential monitoring:
C.
Same data.
Different decisions.
The framework is useful because the tradeoffs are useful
Do not obsess over producing one perfect score.
Sometimes the disagreement between dimensions is the insight.
High value + low recency:
recover.
High frequency + low spend:
perhaps loyalty or bundle expansion.
High recency + low frequency:
develop.
That is the point.
Add the missing fourth and fifth dimensions
For Revenue Recapture, I would add:
Product logic
and:
Reachability
Because a customer can score beautifully on historical behaviour and still have no sensible recovery path.
The principle
Do not ask:
Which metric wins?
Ask:
Which signal answers the business question we are dealing with right now?
That is a more useful way to use customer data.
Related reading
How to Prioritize Lapsed Customers