New customer CAC vs blended CAC
New customer CAC is what you spend to win one first-time buyer. Blended CAC spreads the same spend over every customer who ordered, including repeat buyers, so it almost always looks cheaper than growth really is.
Blended CAC divides the same spend by all customers who ordered, new and returning.
What is new customer CAC?
Most dashboards show one CAC: total marketing spend divided by everyone who placed an order. That blended number mixes two very different things. Winning a first-time buyer is expensive. Bringing back someone who already loves the product is cheap, and often happens with little or no paid media at all.
New customer CAC counts only the first-timers. As your repeat base grows, blended CAC drifts down or stays flat even when each new customer is getting more expensive. nCAC is the number that tells you what growth actually costs.
How it's calculated
You need spend from the ad platforms and customer history from your store. The hard part isn't the division. It's deciding, reliably, who counts as new.
| Input | Source and range | What it gives us |
|---|---|---|
| spend | Meta, Google and other paid channels, the month | Numerator for both CACs |
| customers | Unique customers with a delivered order, the month | Denominator for blended CAC |
| first_order | Customer's full order history, matched on phone and email | Whether this month's order is their first ever |
| new_customers | Customers whose first delivered order falls in the month | Denominator for new customer CAC |
Count a customer as new only once, at the brand level, using their earliest delivered order. A customer who returns through a different channel is still a repeat buyer.
How to track it over time
nCAC is only comparable month to month if the definition of new stays fixed. Lock it down once, then trend the two CACs side by side.
- 1Fix "new"First delivered order ever, matched on phone and email, across all channels.
- 2Split the monthCount new and returning customers separately against the same spend.
- 3Wait for deliveryRecount once RTO settles, so cancelled COD orders don't inflate new customers.
- 4Trend bothWatch the gap between blended and nCAC. A widening gap means repeat buyers are hiding rising acquisition cost.
Worked examples
Example 1: Five months of a growing brand
An illustrative brand scales paid spend from ₹20,00,000 to ₹32,00,000 a month. Its repeat base grows fast, its new customer count barely moves.
| Month | Spend | New customers | Returning | Blended CAC | New customer CAC |
|---|---|---|---|---|---|
| Jun | ₹20,00,000 | 4,000 | 1,000 | ₹400 | ₹500 |
| Jul | ₹22,00,000 | 4,100 | 1,400 | ₹400 | ₹537 |
| Aug | ₹25,00,000 | 4,200 | 1,800 | ₹417 | ₹595 |
| Sep | ₹28,00,000 | 4,300 | 2,300 | ₹424 | ₹651 |
| Oct | ₹32,00,000 | 4,400 | 2,900 | ₹438 | ₹727 |
- Jun–JulBlended CAC holds at ₹400. Everything looks fine.
- Aug–SepSpend is up 40% on June but new customers are up only 7.5%. Returning customers are absorbing the growth.
- OctEach new customer now costs ₹727, up 45% since June. Blended CAC shows a 10% rise. The extra spend is mostly paying for people who would have come back anyway.
Example 2: The cheapest channel isn't the cheapest for growth
Two channels in the same illustrative month, with customers credited by BooleanMaths attribution rather than platform claims.
Using new customer CAC to decide
Read nCAC next to new customer share and what a new customer is worth (CM1 LTV over your payback window). These are starting rules to adapt to your margins.
Common mistakes
- Reporting blended CAC as "CAC"It improves as your repeat base grows, even when acquisition is getting worse.
- Counting "new" per channelA repeat buyer clicking a Google ad for the first time is not a new customer.
- Counting placed, not deliveredA COD order that comes back RTO didn't acquire anyone. Use delivered first orders.
- Matching on email onlyMany Indian shoppers check out by phone. Match on both, or repeat buyers show up as new.
- Using platform-claimed customersPlatforms over-claim. Channel nCAC should use orders your own attribution credits.
Frequently asked questions
Is new customer CAC the same as CPA?
No. CPA divides spend by orders, from anyone. nCAC divides spend by people buying from you for the first time. A brand with lots of repeat orders will have a CPA far below its nCAC.
Why is my nCAC higher than Meta's cost per purchase?
Meta's cost per purchase counts every purchase it claims, including repeat buyers and orders it over-attributes. nCAC counts only verified first-time buyers, so it's higher and closer to the truth.
Should influencer and agency fees be included?
If they're aimed at winning new customers, yes. Pick a rule and keep it fixed, otherwise month-to-month changes reflect accounting, not performance.
What is a good new customer CAC?
There's no universal number. A good nCAC is one a new customer pays back from contribution margin within a window your cash can carry. Compare it to CM1 LTV, not to another brand's CAC.
Related metrics
Know what each new customer really costs
BooleanMaths separates first-time and returning buyers using delivered orders and attributed channels, so new customer CAC is ready by channel and campaign.
New customer CAC vs blended CAC
New customer CAC is what you spend to win one first-time buyer. Blended CAC spreads the same spend over every customer who ordered, including repeat buyers, so it almost always looks cheaper than growth really is.
Blended CAC divides the same spend by all customers who ordered, new and returning.
What is new customer CAC?
Most dashboards show one CAC: total marketing spend divided by everyone who placed an order. That blended number mixes two very different things. Winning a first-time buyer is expensive. Bringing back someone who already loves the product is cheap, and often happens with little or no paid media at all.
New customer CAC counts only the first-timers. As your repeat base grows, blended CAC drifts down or stays flat even when each new customer is getting more expensive. nCAC is the number that tells you what growth actually costs.
How it's calculated
You need spend from the ad platforms and customer history from your store. The hard part isn't the division. It's deciding, reliably, who counts as new.
| Input | Source and range | What it gives us |
|---|---|---|
| spend | Meta, Google and other paid channels, the month | Numerator for both CACs |
| customers | Unique customers with a delivered order, the month | Denominator for blended CAC |
| first_order | Customer's full order history, matched on phone and email | Whether this month's order is their first ever |
| new_customers | Customers whose first delivered order falls in the month | Denominator for new customer CAC |
Count a customer as new only once, at the brand level, using their earliest delivered order. A customer who returns through a different channel is still a repeat buyer.
How to track it over time
nCAC is only comparable month to month if the definition of new stays fixed. Lock it down once, then trend the two CACs side by side.
- 1Fix "new"First delivered order ever, matched on phone and email, across all channels.
- 2Split the monthCount new and returning customers separately against the same spend.
- 3Wait for deliveryRecount once RTO settles, so cancelled COD orders don't inflate new customers.
- 4Trend bothWatch the gap between blended and nCAC. A widening gap means repeat buyers are hiding rising acquisition cost.
Worked examples
Example 1: Five months of a growing brand
An illustrative brand scales paid spend from ₹20,00,000 to ₹32,00,000 a month. Its repeat base grows fast, its new customer count barely moves.
| Month | Spend | New customers | Returning | Blended CAC | New customer CAC |
|---|---|---|---|---|---|
| Jun | ₹20,00,000 | 4,000 | 1,000 | ₹400 | ₹500 |
| Jul | ₹22,00,000 | 4,100 | 1,400 | ₹400 | ₹537 |
| Aug | ₹25,00,000 | 4,200 | 1,800 | ₹417 | ₹595 |
| Sep | ₹28,00,000 | 4,300 | 2,300 | ₹424 | ₹651 |
| Oct | ₹32,00,000 | 4,400 | 2,900 | ₹438 | ₹727 |
- Jun–JulBlended CAC holds at ₹400. Everything looks fine.
- Aug–SepSpend is up 40% on June but new customers are up only 7.5%. Returning customers are absorbing the growth.
- OctEach new customer now costs ₹727, up 45% since June. Blended CAC shows a 10% rise. The extra spend is mostly paying for people who would have come back anyway.
Example 2: The cheapest channel isn't the cheapest for growth
Two channels in the same illustrative month, with customers credited by BooleanMaths attribution rather than platform claims.
Using new customer CAC to decide
Read nCAC next to new customer share and what a new customer is worth (CM1 LTV over your payback window). These are starting rules to adapt to your margins.
Common mistakes
- Reporting blended CAC as "CAC"It improves as your repeat base grows, even when acquisition is getting worse.
- Counting "new" per channelA repeat buyer clicking a Google ad for the first time is not a new customer.
- Counting placed, not deliveredA COD order that comes back RTO didn't acquire anyone. Use delivered first orders.
- Matching on email onlyMany Indian shoppers check out by phone. Match on both, or repeat buyers show up as new.
- Using platform-claimed customersPlatforms over-claim. Channel nCAC should use orders your own attribution credits.
Frequently asked questions
Is new customer CAC the same as CPA?
No. CPA divides spend by orders, from anyone. nCAC divides spend by people buying from you for the first time. A brand with lots of repeat orders will have a CPA far below its nCAC.
Why is my nCAC higher than Meta's cost per purchase?
Meta's cost per purchase counts every purchase it claims, including repeat buyers and orders it over-attributes. nCAC counts only verified first-time buyers, so it's higher and closer to the truth.
Should influencer and agency fees be included?
If they're aimed at winning new customers, yes. Pick a rule and keep it fixed, otherwise month-to-month changes reflect accounting, not performance.
What is a good new customer CAC?
There's no universal number. A good nCAC is one a new customer pays back from contribution margin within a window your cash can carry. Compare it to CM1 LTV, not to another brand's CAC.
Related metrics
Know what each new customer really costs
BooleanMaths separates first-time and returning buyers using delivered orders and attributed channels, so new customer CAC is ready by channel and campaign.