Marketing pro forma: the assumptions, the three scenarios and the return it projects

A marketing pro forma is a projected financial statement for your marketing, built on stated assumptions. It turns spend into expected leads, customers, revenue and payback so that a finance lead can approve or challenge it. Build it with an assumptions block, a monthly projection and three scenarios, and every line traces back to a number somebody chose.

What a pro forma is, and how it differs from a budget

A budget lists what you plan to spend. A pro forma shows what you expect that spend to return. The first is a list of costs; the second is a forecast of outcomes, and its value is that it makes your assumptions visible. If you need the cost side, the marketing budget format guide covers the columns and categories. This page picks up where that one stops.

The assumptions block

Put every input in one place at the top, so anybody can change one cell and watch the projection move. The inputs for a typical lead generation programme:

• Monthly spend per channel.
• Cost per click or cost per thousand views, per channel.
• Landing page conversion rate, from visit to lead.
• Lead to customer rate, from your sales records.
• Average first order value and gross margin.
• Repeat purchase rate, if repeat revenue is part of the case.

Source each input. Use last year's actuals where they exist. Where they do not, use a small test result, and label it as a test result so the reader knows how firm it is.

The projection, line by line

Each month the model runs the same chain:

1. Spend divided by cost per click gives visits.
2. Visits times conversion rate gives leads.
3. Leads times close rate gives new customers.
4. New customers times first order value gives revenue.
5. Revenue times gross margin gives contribution.
6. Contribution minus spend gives net return for the month.
7. Spend divided by new customers gives cost per customer.

A worked illustration

The numbers below are invented to show the arithmetic and are not a benchmark. Assume monthly spend of INR 200,000, a cost per click of INR 25, a landing page conversion rate of 4%, a lead to customer rate of 10% and a contribution of INR 6,000 per customer on the first order.

• INR 200,000 divided by INR 25 gives 8,000 visits.
• 4% of 8,000 gives 320 leads.
• 10% of 320 gives 32 customers.
• 32 customers at INR 6,000 gives INR 192,000 of contribution.
• Cost per customer is INR 200,000 divided by 32, which is INR 6,250.

On first order alone the month shows a net return of minus INR 8,000. That is the signal to bring in repeat purchases: if the average customer buys again, the contribution per customer rises and the programme crosses into profit. The pro forma shows exactly how much repeat behaviour the plan depends on, which is the conversation finance wants to have.

Three scenarios

Run the same model with the weakest input moved up and down. In the illustration, change only the landing page conversion rate:

Scenario	Conversion rate	Leads	Customers	Cost per customer (INR)
Low	3%	240	24	8,333
Base	4%	320	32	6,250
High	5%	400	40	5,000

One variable at a time keeps the table readable. Choose the variable the plan is most sensitive to, usually conversion rate or cost per click, and say which one it is.

Presenting it for approval

Lead with the base case in one line: spend, customers, cost per customer and payback month. Follow with the low and high cases. Then show the assumptions block, because the first question will be where the numbers came from. Close with the review rule: the pro forma is rebuilt from actuals each month, and a channel that misses its cost per customer for two months in a row is cut or fixed.

Keeping the pro forma honest

• Never copy a conversion rate from an article. Use your own data or a test.
• Keep new customer revenue separate from repeat revenue.
• Show the month the programme pays back, not only the total at year end.
• Update it with actuals every month and keep the original, so you can see how far your assumptions were off.

A related question is how much of company revenue to commit in the first place, which is the subject of the budget structure and of any benchmark you choose to use.

Where each assumption comes from

Finance will ask for the source of each number, so record it beside the input. A cost per click can come from the ad platform's own estimate or from last quarter's account. A landing page conversion rate comes from your analytics for the same page, or from a small test. A lead to customer rate comes from your sales records, split by lead source if you have enough data. Average order value and margin come from the finance team.

When two sources disagree, use the lower rate in the base case and show the higher one in the high scenario. Mark every cell as measured, estimated or assumed, with a colour code. The reader then sees at a glance how much of the projection rests on evidence and which cells to test first.

The input your pro forma probably leaves out: the assistant's answer

A lead to customer rate is partly a measure of what buyers believed before they spoke to you. Those beliefs now come from ChatGPT, Gemini and Perplexity answers as much as from your website. If an assistant names three competitors when a buyer asks for the best option and leaves you out, the pool of leads that arrive already convinced is smaller than your model assumes. AI Knows Us measures which businesses those assistants name, and the free scan at aiknowsus.com gives you a result you can add to the assumptions block as a tracked input.

Questions about building a marketing pro forma

What is the difference between a pro forma and a forecast?

A pro forma is a forecast built on explicit, adjustable assumptions, laid out like a financial statement. The assumptions are the point.

How many months should it cover?

Twelve months monthly for the main case. Add a three year view only if customers repeat or contracts run long.

Which numbers do I use if I have no history?

Run a small paid or outreach test for two to four weeks, take the measured rates, and label them as test figures in the model.

Should the pro forma include salaries?

Include them if the person's time is part of the programme. Show them on their own row so the reader can see the cost with and without.

Who should approve it?

The person who owns the revenue target and the person who owns the cash. When both sign the assumptions, the later results stop being a debate.