Every customer looks the same
Reports carry revenue and order counts, but not who will buy again and who is already gone. Advertising pushes just as hard on both.
Which customers still have money left in them
Valkiria predicts lifetime value per buyer, shows how each cohort keeps coming back month after month, and splits the base into seven RFM segments. All of it runs on the orders from your connected store, with no manual exports.
Why budget keeps chasing customers who are never coming back
Nobody measures retention
You know how many customers arrived this month. You do not know how many of March's buyers still order — and that is where the real economics live.
Churn gets noticed too late
A regular buyer stops ordering and it stays invisible until the monthly revenue dips. By then the moment to win them back has passed.
Three steps to a customer base you can read
Connect the store
Shopify connects over OAuth, WooCommerce over store keys. Orders and customers sync into Valkiria and become the basis for every calculation.
Read the predicted LTV
The module works out average order value, purchase frequency and expected lifespan per customer, then sorts the base into value segments.
Work through cohorts and RFM
The heatmap shows how each cohort returns, and RFM says who to keep, who to win back, and who is not worth chasing any more.
What you get about your customer base
Predicted lifetime value per customer
Average order value, purchase frequency and expected lifespan collapse into one number you can hold against acquisition cost.
Total predicted revenue
How much money the current base is expected to bring, and how that sum spreads across the value segments.
Cohorts by month of first order
A retention heatmap: what share of each cohort comes back in the first, second and tenth period after their first purchase.
Seven RFM segments
Champions, loyal, potential, new, at risk, hibernating and lost — each with its share of the base.
A numeric portrait of each segment
Every segment carries its average recency in days, its average purchase frequency and its average spend.
An early churn signal
A customer with no order for more than 90 days is flagged at risk automatically — before the monthly report notices anything.
In numbers
- 7
- RFM segments the customer base is split into
- 3
- views of the same customers: LTV, cohorts and RFM
- 0
- changes made to your store — the module only reads orders
Questions about customer analytics
Where does the customer data come from?
From your connected online store: Shopify authorises over OAuth, WooCommerce over store keys. Valkiria syncs orders and customer records, and every calculation runs on those. Until a source is connected the module shows an honest empty state with a link to connect one.
How exactly is lifetime value calculated?
For each customer it takes the average order value, the purchase frequency over the lifetime so far, and a forecast of the remaining lifespan; the product of the three is the predicted LTV. The base is then split by percentile, and anyone with no order for over 90 days moves into the at-risk segment.
What does the at-risk segment mean?
It is a customer who used to buy but has placed no order in the last 90 days. They are pulled out into their own segment regardless of how much they spent before — which is exactly where the money you can still recover tends to sit.
How is RFM different from the LTV segments?
LTV segments answer how much a customer will bring. RFM answers what state they are in right now: recency, frequency and monetary value are scored from 1 to 5, and those scores roll up into seven readable groups, from champions to lost.
How deep does the cohort analysis go?
Cohorts are grouped by the month of the first order, and each one shows how many customers returned in every following period. Retention is measured against the cohort's size in period zero, so rows stay comparable no matter how large an intake was.
See who comes back and who already left
The module only reads the orders of your connected store and changes nothing inside it. Connecting takes a few minutes.