I only have a few hundred customers. Do I need Retrics?

Maybe not yet. With a few hundred customers and an hour a week, a spreadsheet plus Shopify's reports is a defensible stack. Retrics starts earning its fee when per-customer cadence stops fitting in your head.

Can I get my data back out?

Yes. Every customer table and report in Retrics exports to CSV, so your spreadsheet stays in the loop — as the place you check our work, or as your exit if you leave.

What does “holdout-scored” actually mean?

A holdout is a matched slice of customers we deliberately don't contact. Comparing them to the customers we did flag is the only honest way to measure lift — and it's why Retrics subtracts the ~2.7% of lapsed customers who come back on their own before claiming a single dollar.

Will I still understand the math?

That's the goal. Each of the 130+ metrics in Retrics is named and defined in the in-app catalog, linked to the report that computes it — the same auditability you built the spreadsheet for.

retrics

compare · spreadsheets

Versus the spreadsheet you already have.

It works for longer than most vendors admit — here's what it does brilliantly, and where the cells run out.

Explore the demo store

All comparisons

02Credit first

What the spreadsheet gets right.

A few hundred customers and an hour a week make it a defensible retention stack. We mean that.

It's free.

No subscription, no seat, no invoice to justify — a price no vendor argues with.

It's completely yours.

Any question you can phrase, it will answer. No roadmap sits between you and your data.

You trust the math.

You wrote every formula, so you can audit every number. That's the standard we hold our own metrics catalog to.

03The honest line

Where the cells run out.

Retention is a per-customer, per-day problem — four jobs that stop fitting in a sheet.

Reorder windows are per-customer

A store-wide average hides the woman who buys every 26 days. Retrics predicts each customer's own window and works the list, not the average.

Scores go stale between exports

A spreadsheet is a photograph — out of date the moment you save. Retrics re-scores every sync, as orders land.

health, return-probability, and reorder window recomputed as orders land

Cells don't draft the win-back

Knowing who lapsed still leaves the writing. Retrics drafts the win-back in your voice — nothing sends until you approve it.

Holdout math needs a holdout

About ~2.7% of lapsed customers come back with no outreach at all. Proving lift means a holdout — possible in a sheet; in practice, never done.

the share of lapsed customers who return with no outreach — always subtracted before we claim a dollar

130+

SQL-backed retention metrics

live in the product today — browse the catalog in-app

analytics report pages

orders, customers, cohorts, RFM, behavior, geography, forecast & more

every sync

each customer re-scored

health, return-probability, and reorder window recomputed as orders land

~2.7%

natural-return floor

the share of lapsed customers who return with no outreach — always subtracted before we claim a dollar

04Side by side

By the job to be done.

Several rows go to the spreadsheet. If those are the rows that matter now, keep it.

Retrics compared with a spreadsheet, job by job
	Retrics	Your spreadsheet
Per-customer reorder windowseach customer scored on her own cadence	●	a store-wide average, by hand
Re-scored as new orders land	●	stale the moment you save
Drafted win-backs, ready to approve	●	—
Holdout measurementsubtracting customers who'd have returned anyway	●	possible, almost never done
Cohort libraryretention curves, RFM, forecast and more, prebuilt	●	if you build and maintain it
Time cost	a brief you read with coffee	an afternoon, most weeks
Subscription cost	a flat monthly fee	free
Bends to any question you inventa spreadsheet has no roadmap	◐	●
Every formula auditable by youRetrics publishes each metric's definition in the in-app catalog	◐	●

05Questions

Fair questions, straight answers.

I only have a few hundred customers. Do I need Retrics?: Maybe not yet. With a few hundred customers and an hour a week, a spreadsheet plus Shopify's reports is a defensible stack. Retrics starts earning its fee when per-customer cadence stops fitting in your head.
Can I get my data back out?: Yes. Every customer table and report in Retrics exports to CSV, so your spreadsheet stays in the loop — as the place you check our work, or as your exit if you leave.
What does “holdout-scored” actually mean?: A holdout is a matched slice of customers we deliberately don't contact. Comparing them to the customers we did flag is the only honest way to measure lift — and it's why Retrics subtracts the ~2.7% of lapsed customers who come back on their own before claiming a single dollar.
Will I still understand the math?: That's the goal. Each of the 130+ metrics in Retrics is named and defined in the in-app catalog, linked to the report that computes it — the same auditability you built the spreadsheet for.

Keep the spreadsheet. Add the engine.

Every table in Retrics exports to CSV, so the sheet stays as the place you check our work.

Start free 14-day trial

Explore the demo store

No credit card · read-only Shopify scopes · live in ~10 minutes

retrics

compare · spreadsheets

Versus the spreadsheet you already have.

It works for longer than most vendors admit — here's what it does brilliantly, and where the cells run out.

Explore the demo store

All comparisons

02Credit first

What the spreadsheet gets right.

A few hundred customers and an hour a week make it a defensible retention stack. We mean that.

It's free.

No subscription, no seat, no invoice to justify — a price no vendor argues with.

It's completely yours.

Any question you can phrase, it will answer. No roadmap sits between you and your data.

You trust the math.

You wrote every formula, so you can audit every number. That's the standard we hold our own metrics catalog to.

03The honest line

Where the cells run out.

Retention is a per-customer, per-day problem — four jobs that stop fitting in a sheet.

Reorder windows are per-customer

A store-wide average hides the woman who buys every 26 days. Retrics predicts each customer's own window and works the list, not the average.

Scores go stale between exports

A spreadsheet is a photograph — out of date the moment you save. Retrics re-scores every sync, as orders land.

health, return-probability, and reorder window recomputed as orders land

Cells don't draft the win-back

Knowing who lapsed still leaves the writing. Retrics drafts the win-back in your voice — nothing sends until you approve it.

Holdout math needs a holdout

About ~2.7% of lapsed customers come back with no outreach at all. Proving lift means a holdout — possible in a sheet; in practice, never done.

the share of lapsed customers who return with no outreach — always subtracted before we claim a dollar

130+

SQL-backed retention metrics

live in the product today — browse the catalog in-app

analytics report pages

orders, customers, cohorts, RFM, behavior, geography, forecast & more

every sync

each customer re-scored

health, return-probability, and reorder window recomputed as orders land

~2.7%

natural-return floor

the share of lapsed customers who return with no outreach — always subtracted before we claim a dollar

04Side by side

By the job to be done.

Several rows go to the spreadsheet. If those are the rows that matter now, keep it.

Retrics compared with a spreadsheet, job by job
	Retrics	Your spreadsheet
Per-customer reorder windowseach customer scored on her own cadence	●	a store-wide average, by hand
Re-scored as new orders land	●	stale the moment you save
Drafted win-backs, ready to approve	●	—
Holdout measurementsubtracting customers who'd have returned anyway	●	possible, almost never done
Cohort libraryretention curves, RFM, forecast and more, prebuilt	●	if you build and maintain it
Time cost	a brief you read with coffee	an afternoon, most weeks
Subscription cost	a flat monthly fee	free
Bends to any question you inventa spreadsheet has no roadmap	◐	●
Every formula auditable by youRetrics publishes each metric's definition in the in-app catalog	◐	●

05Questions

Fair questions, straight answers.

I only have a few hundred customers. Do I need Retrics?: Maybe not yet. With a few hundred customers and an hour a week, a spreadsheet plus Shopify's reports is a defensible stack. Retrics starts earning its fee when per-customer cadence stops fitting in your head.
Can I get my data back out?: Yes. Every customer table and report in Retrics exports to CSV, so your spreadsheet stays in the loop — as the place you check our work, or as your exit if you leave.
What does “holdout-scored” actually mean?: A holdout is a matched slice of customers we deliberately don't contact. Comparing them to the customers we did flag is the only honest way to measure lift — and it's why Retrics subtracts the ~2.7% of lapsed customers who come back on their own before claiming a single dollar.
Will I still understand the math?: That's the goal. Each of the 130+ metrics in Retrics is named and defined in the in-app catalog, linked to the report that computes it — the same auditability you built the spreadsheet for.

Keep the spreadsheet. Add the engine.

Every table in Retrics exports to CSV, so the sheet stays as the place you check our work.

Start free 14-day trial

Explore the demo store

No credit card · read-only Shopify scopes · live in ~10 minutes