Metrics glossary

Shopify metrics, defined and worked through.

20 metrics that Shopify operators run on, each with a plain definition, the formula, a worked example, and what to watch. The definitions follow how By the Numbers calculates these metrics.

Sales

How much the store sells, and what shapes each order.

Conversion rateCVR

Also called store conversion rate

Tells you how well the store turns visits into orders, separate from how much traffic it gets.

Definition

The share of online store sessions that end in an order, over the same period.

Formula

CVR=Sessions with an orderSessions× 100

What to watch

Sessions depend on who counts them. Google Analytics, a pixel, and Shopify each count visits differently, so compare rates from one source.
Point of sale orders have no session, so they stay out of the rate.
New traffic from a fresh channel often lowers the rate before revenue moves. Read it beside sessions and AOV.

25,600

Sessions

↓ 2.5%

640

Sessions with an order

Worked example

Sample store

Over a quarter, the sample store's online store logs these sessions.

Online store sessions: 25,600
Sessions that ended in an order: 640
Conversion rate: 2.5%

On the same traffic, half a point more would add 128 orders a quarter.

Average order valueAOV

Shows whether growth comes from bigger baskets or from more orders, and moves with pricing, bundles, and discounts.

Definition

Revenue divided by the number of orders placed in the window, leaving out cancelled, voided and expired orders; revenue is each order's total after discounts, including tax and shipping, before refunds.

Formula

AOV=RevenueOrders

What to watch

Revenue here is before refunds. Views that take refunds out first read lower in a month with returns, so compare figures on the same basis.
Where prices include tax, the tax is already inside the item prices and isn't added again.
Free orders, such as a $0 replacement, still count as orders and pull AOV down.

One order

Items$90.00(counts)
Discount code−$10.00(counts)
Shipping$6.00(counts)
Sales tax$6.40(counts)
Later refund−$20.00(does not count)

Counts toward AOV$92.40

Worked example

Sample store

The sample store took these orders over a quarter. A few were canceled before they shipped.

Orders placed: 649
Canceled orders: −9
Orders that count: 640
Revenue on those orders: $51,200
AOV: $80.00

Refunds made later don't change this figure. The refund rate tracks them instead.

Discount intensity

Also called discount rate

Shows how much of the store's sales lean on a code, an early sign of customers waiting for the next promotion.

Definition

Share of live orders in the period that used a discount code.

Formula

Discount intensity=Orders that used a discount codeOrders× 100

What to watch

Live orders leave out orders that were canceled, voided, or expired.
Automatic discounts need no code, so they don't count here. A store that runs promotions that way reads low.
The share of orders says nothing about depth. Check the dollars given away per code as well.

Used a codeFull price

Worked example

Sample store

Of the sample store's orders for the quarter, these used a code at checkout.

Orders: 640
Orders with a discount code: 224
Discount intensity: 35.0%

Refund rate

Shows how much of what the store sells comes back as money returned, and whether that share is growing.

Definition

Money refunded so far on orders placed in the window, as a share of those orders' revenue (each order's total after discounts, including tax and shipping); cancelled, voided and expired orders are left out.

Formula

Refund rate=Money refunded on the period's ordersRevenue from those orders× 100

What to watch

Counted by order date, a recent period reads low until its refunds come in. Judge a month once its return window has passed.
Some reports count refunds on the day they are paid instead, which gives a different figure. Compare rates built the same way.
Refund rate counts money. A return rate counts units sent back, so the two can move apart.

RefundedKept

Worked example

Sample store

The sample store checks refunds against the orders it took in the quarter, whenever each refund was paid.

Revenue from the quarter's orders: $51,200
Refunded on those orders so far: $2,560
Refund rate: 5.0%

Customers and retention

Who comes back, how often, and what they are worth.

Customer lifetime valueLTV

Also called CLV, lifetime value

Sets how much the store can afford to spend to win a customer, and which customers deserve the most attention.

Definition

What a customer is expected to spend with the store over the whole relationship. By the Numbers counts what they have spent so far, net of returns, plus a projection of what they are likely to spend from here.

Formula

LTV=Spent so far, net of returns+Projected future spend

What to watch

LTV in revenue is not profit. Multiply it by gross margin before you set it against acquisition cost.
A customer with one order gives little to project from, so their LTV sits close to what they have spent.
Compare LTV by the month customers were acquired to see whether newer customers are worth more or less.

Lifetime value$305

$210

$95

Spent so farProjected

Worked example

Sample store

A customer of the sample store has placed 3 orders and returned one item.

Paid across 3 orders: $250
Refunded: −$40
Spent so far: $210
Projected future spend: +$95
LTV: $305

The projection comes from the customer's average order and how customers in their loyalty group usually keep buying.

Orders per customer

Also called purchase frequency

Shows how often customers buy in a period, the lever beside AOV in every revenue plan.

Definition

Live orders in the period divided by distinct customers with a live order in the period.

Formula

Orders per customer=OrdersCustomers who ordered

What to watch

A longer period raises it. A 90-day figure and a 12-month figure measure different things.
An average hides the spread. A few heavy buyers can lift it while most customers order once, so check the median too.
Subscription renewals raise it without a fresh decision to buy. Leave them out to see how often customers choose to reorder.

640

Orders

500

Customers

1.28

Orders each

Worked example

Sample store

Over the quarter, the sample store's orders came from these customers.

Orders: 640
Customers who ordered: 500
Orders per customer: 1.28

Repeat purchase rate

Also called repeat customer rate

Shows how many customers buy more than once in a period, the clearest early read on retention.

Definition

Of customers who placed at least one order in the window, the share who placed two or more in the same window; cancelled, voided and expired orders don't count.

Formula

Repeat purchase rate=Customers with 2+ orders in the periodCustomers with 1+ order in the period× 100

What to watch

The period sets the number. A 90-day rate reads far lower than a 12-month or lifetime rate, so don't compare them without saying so.
Repeat purchase rate and returning customer rate answer different questions. If you report both, define both.
To follow it by first-order month, cohort repurchase rates show the share who bought again within 30, 60, 90, or 180 days.

The 90 days

Ana
Two orders inside the period
Ben
One order inside; the other came before
Cy
One order
Dee
Three orders inside the period

Each dot is an order. 2 of these 4 customers count: a 50% repeat purchase rate.

Worked example

Sample store

The sample store looks at the customers who ordered in its latest 90 days.

Customers who ordered: 500
Ordered twice or more in the 90 days: 85
Repeat purchase rate: 17.0%

Returning customer rate

Shows how much of a period's demand comes from customers the store already had.

Definition

Of customers who placed an order in the window, the share whose first-ever order at the store came before the window; cancelled, voided and expired orders don't count.

Formula

Returning customer rate=Customers whose first order came before the periodCustomers who ordered in the period× 100

What to watch

A period that wins many new customers reads a lower rate, even if loyal customers buy as often as before.
The length of the period matters. A customer whose first two orders fall in the same month counts as new in a monthly view.
It says nothing about how often returning customers buy. Pair it with repeat purchase rate.

The 90 days

Ana
New: her first order is inside the period
Ben
Returning: first ordered before the period
Cy
New
Dee
Returning

The same 4 customers as the repeat purchase example. Again 2 count, but not the same 2.

Worked example

Sample store

Of the customers who ordered from the sample store in the 90 days, some had bought before.

Customers who ordered: 500
First order before the 90 days: 170
New in the 90 days: 330
Returning customer rate: 34.0%

Customer retention rate

Also called retention rate

Shows how much of last year's customer base the store kept, the base its repeat revenue rests on.

Definition

Of customers who ordered in the 12 complete calendar months before the latest 12, the share who ordered again in the latest 12 complete months; guest orders and cancelled, voided and expired orders don't count.

Formula

Retention rate=Last year's customers who ordered again this yearCustomers who ordered last year× 100

What to watch

Retention and churn are two sides of the same customers. 28% retained means 72% churned.
Base-year customers include that year's new buyers, so a fast-growing store reads lower retention.
A single yearly rate hides when customers drop off. Cohort retention by first-order month shows it.

Last year

This year

Ana
Ordered in both years
Ben
No order this year
Cy
Ordered in both years
Dee
Left out: no order last year

2 of last year's 3 customers ordered again this year. Dee first ordered this year, so she counts from next year.

Worked example

Sample store

The sample store compares its last two complete years of customers.

Customers who ordered in the base year: 4,000
Ordered again in the latest 12 months: 1,120
Retention rate: 28.0%

Customer churn rate

Also called churn rate, attrition rate

Shows how many customers the store lost over a year, the gap new customers have to fill before it grows.

Definition

Of customers who ordered in the 12 complete calendar months before the latest 12, the share who placed no order in the latest 12 complete months; guest orders and cancelled, voided and expired orders don't count.

Formula

Churn rate=Last year's customers with no order this yearCustomers who ordered last year× 100

What to watch

Shopping has no cancel button, so churn means no order for 12 months. That overstates it for furniture and understates it for coffee.
Many one-time buyers in the base year push churn up. Split it by order count to see whether repeat customers are leaving.
Act before the 12 months are up. Customers moving into the At risk and Dormant loyalty groups are the early signal.

No order this yearOrdered again

Worked example

Sample store

The same two years of the sample store's customers, counted the other way.

Customers who ordered in the base year: 4,000
No order in the latest 12 months: 2,880
Churn rate: 72.0%

Cohort

Also called acquisition cohort, cohort analysis

Lets you compare customers of the same age, so a strong month of new buyers can't hide weaker retention.

Definition

A group of customers who share a starting point, usually the month of their first order. Following each group over time shows how many come back, and when.

Formula

Cohort retention, month N=Cohort customers who ordered in month NCustomers in the cohort× 100

What to watch

Read down a column to compare cohorts at the same age. Read along a row to see one cohort fade.
Recent cohorts have fewer months behind them, so leave their empty cells out of any average.
A cohort won on a big promotion often comes back less. Check what brought each month's customers in.

Mo 1

Mo 2

Mo 3

Mo 4

Jan

14%

Feb

12%

Mar

16%

10%

Apr

13%

Each row is the customers whose first order fell in that month. Each cell is the share of them who ordered again in that later month.

Worked example

Sample store

The sample store's January cohort is every customer whose first order came that month.

Customers in the January cohort: 400
Ordered again in February (month 1): 56
Month 1 retention: 14.0%

RFM analysisRFM

Also called RFM matrix, loyalty matrix

Turns a customer list into groups you can act on: who to thank, who to win back, and who is slipping away.

Definition

RFM stands for recency, frequency, and monetary value: how recently a customer bought, how often, and how much. The By the Numbers loyalty matrix uses recency and frequency. It sorts customers into 12 cells by last order and order count, which roll up into 7 groups.

Formula

Loyalty cell=Days since last order×Lifetime orders

What to watch

This matrix has no monetary axis. Two customers with the same last order and order count share a cell, whatever they spent.
Cutoffs should follow the buying cycle. Recent means weeks for a coffee store and a year for a furniture store, so each store sets its own.
Each group can sync as a segment to Klaviyo, Mailchimp, Attentive, Meta, Google Ads, TikTok, and Reddit Ads.

Days since last order

366+

181–365

31–180

Last 30

4+4+ orders

2–32–3 orders

11 order

Dormant

At risk

Loyal

Best

Promising

Defecting

Recent

Columns show how long ago a customer last ordered. Rows show how many orders they have placed: 12 cells in 7 groups. The cutoffs are an example; each store sets its own.

Worked example

Sample store

With the example cutoffs, here is where one of the sample store's customers lands.

Last order: 20 days ago
Recency column: Last 30 days
Lifetime orders: 5
Frequency row: 4+ orders
Group: Best

The same customer, silent for 200 days, would move to At risk.

Marketing

What ad spend returns, and what each new customer costs.

Return on ad spendROAS

Shows what each ad dollar brings back in revenue, the first check on whether a campaign is worth its budget.

Definition

Revenue from orders with a paid UTM medium or source, divided by ad spend from connected ad platforms.

Formula

ROAS=Revenue from orders with a paid UTM tagAd spend

What to watch

Ad platforms report ROAS from the conversions they claim, and their totals can add up to more orders than the store took.
Break-even ROAS is 1 divided by gross margin. Judge each channel against that line, not against 1.
Clicks without UTM tags fall out of this figure. Tag ads consistently, or paid revenue reads low.

Ad spend$13,200
Revenue from paid-tagged orders$29,040

Each ad dollar returned $2.20 in revenue.

Worked example

Sample store

The sample store ran ads on two platforms for the quarter and tags each ad with UTM parameters.

Ad spend, connected platforms: $13,200
Revenue from paid-tagged orders: $29,040
ROAS: 2.20

At a 57% gross margin, break-even ROAS is 1.75, so these ads clear product costs.

Marketing efficiency ratioMER

Also called blended ROAS

Shows whether the marketing budget as a whole pays, with no attribution to argue over.

Definition

Revenue divided by marketing spend over the same period. Every sale counts against every dollar, so it shows the return on the whole budget rather than one channel.

Formula

MER=Net revenueAd spend, including costs counted as ad spend

What to watch

MER can't say which channel worked. Pair it with ROAS by platform.
Revenue includes customers who would have bought anyway, so MER flatters a store with a large returning base.
Fees, creators, and agencies are marketing costs too. Count them in the spend, or MER reads high.

Marketing spend$15,200
Net revenue$48,640

Worked example

Sample store

The sample store adds an agency fee to its platform spend, and uses revenue after refunds.

Revenue after refunds: $48,640
Ad spend, connected platforms: $13,200
Agency fee counted as ad spend: +$2,000
MER: 3.20

Read the other way, marketing costs 31.3% of revenue.

Customer acquisition costCAC

Also called blended CAC

Shows what the store pays, on average, to win one new customer.

Definition

Ad spend from connected ad platforms divided by new customers in the period.

Formula

CAC=Ad spendNew customers

What to watch

Blended CAC divides all ad spend by all new customers, including those who found the store on their own. Paid-only CAC reads higher.
Returning buyers don't count, however much ad spend reached them.
Decide whether agency fees and creative belong in the spend, then keep the choice fixed.

$13,200

Ad spend

330

New customers

$40

Each

Worked example

Sample store

New customers are those whose first order with the sample store came this quarter.

Ad spend, connected platforms: $13,200
New customers: 330
CAC: $40.00

LTV to CAC ratioLTV:CAC

Shows whether new customers are worth what they cost to win, across acquisition and retention at once.

Definition

Customer lifetime value divided by customer acquisition cost: the dollars a new customer brings in for each dollar spent to win them.

Formula

LTV:CAC=Lifetime valueCustomer acquisition cost

What to watch

Use gross-profit LTV to judge spend. A revenue ratio overstates it by whatever your margin is not.
3:1 is a common rule of thumb, not a law. How long you can fund the gap until a customer pays back matters as much.
Match the LTV to the customers the spend acquired, from the same period.

CAC$40.00
LTV in gross profit$174.61
LTV in revenue$305

The revenue ratio looks almost twice as good. Gross profit is what pays the ad bill.

Worked example

Sample store

The sample store's new customers have an LTV of $305, at a 57% gross margin.

LTV in revenue: $305
LTV in gross profit: $174.61
CAC: ÷$40.00
Ratio on revenue: 7.6 : 1
Ratio on gross profit: 4.4 : 1

CAC payback period

Also called payback period

Shows how long the store waits to earn back what it paid for a customer, which sets how much cash growth ties up.

Definition

The time it takes the gross profit from a new customer's orders to cover what it cost to acquire them.

Formula

Payback=First month where cumulative gross profit per customer≥CAC

What to watch

Revenue payback looks faster than profit payback. Use gross profit when you plan cash.
Blended CAC hides channels that pay back slowly. Work it out by campaign or platform where you can.
A long payback is affordable only if customers stay. Read it beside retention.

Month 0, cumulative$35
Month 1, cumulative$47
Month 2, cumulative$62
Month 3, cumulative$76
Month 4, cumulative$92

The bar turns green once cumulative profit passes the $90 CAC.

Worked example

Sample store

Customers from one of the sample store's paid social campaigns cost $90 each.

Gross profit, month 0: $35 (total $35)
Gross profit, month 1: $12 (total $47)
Gross profit, month 2: $15 (total $62)
Gross profit, month 3: $14 (total $76)
Gross profit, month 4: $16 (total $92)
Pays back in: month 4

Month 0 is the month of the first order.

Free toolLTV to CAC ratio calculator →

Profit

What is left after product costs, ads, and fulfillment.

Cost of goods soldCOGS

The cost of the products behind each sale, and the starting point for gross margin and contribution.

Definition

What the products a store sold cost it: each line's quantity times its unit cost. By the Numbers uses each variant's cost per item from Shopify.

Formula

COGS=Units sold×Cost per item

What to watch

A product with no cost set counts as zero cost. Fill in costs across the catalog before you trust any margin.
The cost used is the variant's current cost, so changing a cost restates past periods too.
COGS covers products only. Shipping, payment fees, and packaging come off later, in contribution margin.

Mugs sold

$6.00

Cost per item

$12.00

COGS for the line

Worked example

Sample store

One order from the sample store, using the cost per item set on each variant in Shopify.

2 mugs at $6.00: $12.00
1 grinder at $38.00: +$38.00
1 sticker with no cost set: +$0.00
COGS: $50.00

The sticker counts as free because no cost is set, which flatters the margin.

Gross margin

Also called gross profit margin

Shows how much of each sales dollar is left once the products are paid for, and caps what the store can spend to sell.

Definition

Revenue minus cost of goods, as a share of revenue, on orders placed in the window.

Formula

Gross margin=Revenue − COGSRevenue× 100

What to watch

Products with no unit cost count as free, which overstates margin. Check cost coverage first.
Reports that take refunds off revenue, such as a P&L, read lower in a month with returns, since returned goods keep their cost.
Gross margin stops at product costs. Contribution margin takes off ad spend, shipping, and fees as well.

Revenue on the quarter's orders$51,200
Cost of goods sold−$21,888
Gross profit$29,312

Worked example

Sample store

The sample store's orders for the quarter, before any refunds.

Revenue on the quarter's orders: $51,200
Cost of goods sold: −$21,888
Gross profit: $29,312
Gross margin: 57.3%

Contribution margin

Also called contribution profit

Shows what each sale leaves to cover fixed costs and profit once the costs of making that sale are paid.

Definition

Revenue left after the costs that rise with each sale: cost of goods, ad spend, shipping, and payment fees. That remainder is contribution profit; the margin is its share of revenue after refunds.

Formula

Contribution margin=Contribution profitRevenue after refunds× 100

What to watch

Shipping and payment fees rarely reach Shopify. Enter them as estimates, or contribution reads high.
A campaign can show a strong ROAS and a negative contribution once product costs come off.
Read it per order and per channel. A store-wide figure can hide a channel that loses money on every sale.

Revenue after refunds$48,640
Cost of goods sold−$21,888
Ad spend−$15,200
Shipping and fees−$4,016
Contribution profit$7,536

Worked example

Sample store

The sample store's quarter, starting from revenue after refunds.

Revenue after refunds: $48,640
Cost of goods sold: −$21,888
Ad spend, with the agency fee: −$15,200
Shipping and payment fees: −$4,016
Contribution profit: $7,536
Contribution margin: 15.5%

Rent, salaries, and software are fixed costs, so they come out after this line.

Want to run the numbers on your own store first? The free Shopify calculators take your figures and need no signup.

Put your store's numbers behind the definitions.

By the Numbers builds 70+ reports from your Shopify orders and connected ad platforms.

Install free on Shopify Talk to a Shopify data expert →

60-second install · 14-day free trial · No credit card required