Marketing finance: ROI, ROAS, CAC, LTV and payback

Marketing finance is the set of numbers that links marketing spend to profit. It covers seven measures that every marketing leader is asked about sooner or later: marketing ROI, ROAS, CAC, LTV, the LTV:CAC ratio, CAC payback and MER, all resting on one quieter number, contribution margin. This guide is for founders, marketing heads and finance teams who need to agree on what the spend is doing. It gives the exact formulas, a worked example in rupees, the evidence on how budgets are set, and the mistakes that make good-looking dashboards lose money.

What do ROI, ROAS, CAC, LTV, payback and MER actually mean?

Each metric answers one question, and the trouble starts when one is used to answer another's. Here are the definitions I would put at the front of any marketing plan, so finance and marketing use the same words.

Metric	Question it answers	Formula
Contribution margin	What is left from a sale after the costs that rise with each sale?	Net revenue − variable costs (product cost, packaging, shipping, payment fees, returns)
Marketing ROI	Did the spend make more profit than it cost?	(Incremental contribution − marketing cost) ÷ marketing cost
ROAS	How much revenue came back per rupee of ads?	Attributed revenue ÷ ad spend
CAC (blended)	What does a new customer cost, all in?	Total acquisition marketing cost ÷ all new customers
CAC (paid)	What does a customer from paid media cost?	Paid media spend ÷ new customers attributed to paid media
LTV / CLV	How much profit does a customer bring over their life with you?	Average orders per customer × contribution margin per order (discounted for longer horizons)
LTV:CAC	Is a customer worth more than they cost?	LTV ÷ CAC
CAC payback	How long until a customer repays what they cost?	CAC ÷ monthly contribution per customer, or read from a cohort table
MER	Is total marketing growing total revenue efficiently?	Total revenue ÷ total marketing spend

A few precise points matter more than they look.

Contribution margin comes first. Gross margin only deducts product cost. Contribution margin also deducts shipping, packaging, payment gateway fees, and returns or RTO (return to origin, the cost of failed cash-on-delivery parcels, which is heavy in Indian e-commerce). Every other metric below should be built on it.

ROAS is a revenue ratio, not a profit ratio. Google's own help page for Target ROAS defines the target as conversion value divided by ad spend, with the example of 5 dollars in sales for every 1 dollar of ads equalling a 500% target. Nothing in that formula knows what your products cost.

Blended and paid CAC tell different stories. Paid CAC flatters paid media, because it ignores the customers who came through search, word of mouth or a retail shelf. Blended CAC includes them all, and also includes the salaries, agency fees and content costs that paid CAC leaves out. Report both, and say which one you mean.

LTV should be profit, not revenue. A revenue-based LTV makes every customer look three or four times more valuable than they are. For horizons beyond a year, discount future orders back to today's value at your cost of capital.

MER is the top-down check. Marketing efficiency ratio (some teams call it blended ROAS) divides all revenue by all marketing spend. It ignores attribution entirely, which is exactly why it is useful: it cannot be gamed by a platform claiming credit for a sale that would have happened anyway.

How do the numbers work in a real example?

The example below is a hypothetical skincare D2C brand in one month. Every number is made up for illustration; none comes from a real company.

The brand's month (hypothetical):

• Net revenue: ₹70,00,000 from 3,500 orders, an average order value (AOV) of ₹2,000
• New customers: 2,000. The other 1,500 orders came from repeat buyers
• Total marketing spend: ₹20,00,000, of which paid media is ₹12,00,000 and the rest is content, agency, influencer and team costs
• Ad platforms report ₹48,00,000 of attributed revenue and 1,600 attributed new customers

Contribution margin per order (hypothetical): AOV ₹2,000, minus product cost ₹700, shipping and packaging ₹200, payment fees ₹40 and a returns and RTO allowance of ₹160. That leaves ₹900 per order, a 45% contribution margin.

Now the metrics:

1. ROAS = ₹48,00,000 ÷ ₹12,00,000 = 4.0x. The dashboard looks healthy.
2. Break-even ROAS = 1 ÷ 45% = 2.22x. Below this, ads lose money even if every attributed sale is real.
3. MER = ₹70,00,000 ÷ ₹20,00,000 = 3.5x.
4. Paid CAC = ₹12,00,000 ÷ 1,600 = ₹750.
5. Blended CAC = ₹20,00,000 ÷ 2,000 = ₹1,000.
6. LTV (if the average customer places 3 orders over 24 months) = 3 × ₹900 = ₹2,700 of contribution.
7. LTV:CAC (blended) = ₹2,700 ÷ ₹1,000 = 2.7:1.
8. CAC payback: the first order returns ₹900, short of the ₹1,000 CAC. If the second order typically lands in month five, cumulative contribution reaches ₹1,800 then, so payback is about five months. The simple formula, ₹1,000 ÷ (₹2,700 ÷ 24 months), gives about nine months, because it spreads the front-loaded first order evenly. For repeat-purchase brands the cohort reading is closer to the truth.

Then comes the test that changes the decision. Paid media ROI at face value is (₹48,00,000 × 45% − ₹12,00,000) ÷ ₹12,00,000 = +80%. Suppose a holdout test (showing no ads to a random group and comparing sales) finds that only half the attributed revenue was truly caused by the ads. Incremental contribution falls to ₹10,80,000, and ROI becomes (₹10,80,000 − ₹12,00,000) ÷ ₹12,00,000 = −10%. Same campaign, same 4x ROAS, opposite answer.

A 4x ROAS tells you the ads were present when people bought. It does not tell you the ads made them buy.

Why does marketing finance matter more now?

Because budgets are flat while the questions get harder. Gartner's 2026 CMO Spend Survey found marketing budgets rising only slightly, to 7.8% of company revenue in 2026 from 7.7% in 2025. The survey covered 401 CMOs and marketing leaders in North America, the UK and Europe, fielded January to March 2026, and found they were putting an average of 15.3% of marketing budgets into AI initiatives. That new line item has to be paid for from a budget that is not growing.

A year earlier, Gartner's 2025 CMO Spend Survey described budgets as having "flatlined" at 7.7%, the same as 2024, with 59% of CMOs saying they lacked the budget to deliver their strategy. That survey covered 402 leaders, fielded February to March 2025, mostly at companies with more than 1 billion dollars in annual revenue.

Two cautions before anyone in India quotes these figures. First, they describe large Western companies; a young Indian D2C brand or a regional chain will usually spend a much higher share of revenue to grow. Second, an average hides a wide range. Use the Gartner number as a sense check on your own, never as a target.

How are marketing budgets set?

Most budgets are set by one of two methods, then argued over in a third conversation: how much goes to brand and how much to performance. I would use the second method to build the number and the first to check it.

1. Percent of revenue

The simplest method: take last year's or next year's revenue and apply a percentage, often the industry average. Its strength is that finance can predict it. Its weakness is that it gets causality backwards: spend follows sales, so a brand in a bad year cuts marketing exactly when it may need it most, and a brand trying to grow fast is held to what it already earns.

2. Objective and task

Start from the goal and price the work. If the plan needs 24,000 new customers next year and your tested blended CAC is ₹1,000, acquisition needs about ₹2.4 crore. Add the brand work that keeps future CAC down, the retention programme, the tools and the team. The total is the budget. It is harder to build, but every rupee has a job and a metric, which is what a CFO wants to see. Its weakness is that it assumes CAC stays flat as you scale; it usually rises as you reach colder audiences, so build in a margin.

3. The brand and performance split

The best-known evidence here is from Les Binet and Peter Field's work for the IPA. Their 2013 report The Long and the Short of It, based on the IPA Databank of effectiveness award cases, found the most effective balance was about 60% brand building to 40% sales activation. Their 2018 follow-up, Effectiveness in Context, put it at 62:38, a point the IPA's own blog noted in 2023, adding that the "sweet spot" had stayed nearly constant.

The caveats matter as much as the headline:

• It is an average. Effectiveness in Context was written precisely to show how the rules change by category, business model, brand life stage and size. Mark Ritson, writing in Marketing Week in February 2023, called 60:40 a "crude average" built from many different categories and situations.
• B2B is different. Binet and Field's 2019 study for LinkedIn's B2B Institute found B2B businesses grow best with about 46% brand and 54% activation.
• The data is award entries. The cases are campaigns good enough to enter for effectiveness awards, not a random sample of all marketing, and the datasets predate today's Indian digital mix.

I would treat 60:40 as a starting hypothesis for an established consumer brand, then move the split based on your own incrementality tests. There is a fuller treatment in brand vs performance: keeping CAC honest.

When should you track each number?

Different numbers move at different speeds, so review them on different clocks:

• Weekly: ROAS against break-even ROAS, paid CAC and spend pacing. These steer bids and creative.
• Monthly: blended CAC, MER and contribution margin after marketing. These tell you whether the whole engine is working.
• Quarterly: cohort LTV, CAC payback and LTV:CAC, because cohorts need time to repeat. Run at least one incrementality test a quarter on your biggest channel.
• Yearly: the budget method and the brand and performance split, reset with finance before the year starts.

Where does marketing finance go wrong?

It goes wrong when a ratio that is easy to see replaces a profit that is hard to measure. Three mistakes cause most of the damage.

Treating ROAS as profit

A 3x ROAS sounds good until you learn the contribution margin is 25%, which makes break-even ROAS 4x. Discount-led categories, cash-on-delivery-heavy regions and products with high return rates in India all push break-even ROAS up. Always publish the break-even line next to the ROAS line.

Trusting attribution as if it were cause

Platforms credit themselves for sales that happen near an ad, and each one counts the same sale. Field experiments show how far that drifts. Thomas Blake, Chris Nosko and Steven Tadelis ran large experiments at eBay, published in Econometrica in 2015, and found that brand-keyword ads had no measurable short-term benefit and that returns from paid search were a fraction of non-experimental estimates. Brett Gordon and colleagues compared 15 US advertising experiments at Facebook with standard observational methods in a 2019 Marketing Science paper and found the observational methods often failed to reproduce the true effects, even with rich data about users.

Ignoring incrementality

The fix for both is to ask what would have happened without the spend. Holdout tests, geographic tests (switching spend off in matched cities) and the lift studies the big platforms offer all answer that question better than a last-click report. Brand search, retargeting and loyal-customer audiences usually look best in attribution and worst in incrementality tests, so test those first. For D2C specifically, see D2C customer acquisition cost in India.

A fourth, quieter mistake is cutting brand spend because it does not show up in ROAS, then watching CAC rise two quarters later and blaming the platforms.

Who owns these numbers: the CMO or the CFO?

Both, with a written split. The CMO owns the plan, the tests, the channel metrics and the case for each rupee. The CFO owns the definitions that sit underneath: what counts as variable cost, how returns and RTO are allocated, the discount rate for LTV and the cash the business can tie up in payback. Disagreements are rarely about marketing; they are about two teams using two versions of contribution margin.

In practice I would agree five things with finance before the year starts:

1. One contribution margin formula, owned by finance.
2. Which CAC is the headline (I would make it blended) and what costs go into it.
3. The break-even ROAS for each major product line.
4. Which incrementality tests will be run, and that their results override platform reports.
5. A brand budget floor that is not cut in a bad month without a joint decision.

Smaller brands without a CFO should give one person, often the founder, the job of keeping these definitions fixed. The danger is not a wrong formula; it is a formula that quietly changes every month.

My view

When I owned a ₹1.2 crore monthly performance budget across Google, Meta and affiliate networks, the useful conversations with finance were never about whether ROAS was high. They were about whether we agreed what a customer was worth. My view is that most marketing finance arguments are definition arguments in disguise, and the fastest way to end them is to put contribution margin at the centre and blended CAC on the front page.

The trade-off is real. Incrementality tests cost money, because you deliberately withhold ads from some people, and they slow decisions down. Contribution-based metrics also make marketing look worse on paper than platform ROAS does, which is uncomfortable in a board meeting. A small brand with one channel and thin data may get more from a simple MER trend than from an elaborate attribution model. But a business that only ever looks at ROAS is measuring its ads' presence, not their effect, and will eventually overspend on the customers it would have won anyway.

Start with one sheet this week

Build a single sheet with your contribution margin per order, break-even ROAS, blended CAC, paid CAC and MER for the last three months, and send it to your finance lead with one question: do we agree on these definitions? Once the answer is yes, pick your largest channel and plan its first holdout test for next quarter.