# Build a SaaS marketing model

> Build the model that turns traffic, conversion rates and sales cycle into pipeline and ARR, with the eight inputs that matter and sensitivity ranges.

Source: https://saas-marketing.net/guides/saas-marketing-model/
Topic: SaaS Marketing
Type: guide
Published: 2026-09-11
Last updated: 2026-09-11
Publisher: SaaS Marketing (saas-marketing.net)
License: CC BY 4.0. Quote or republish with attribution and a link to https://saas-marketing.net/guides/saas-marketing-model/

## Short answer

A SaaS marketing model is a spreadsheet that converts eight inputs into forecast ARR: qualified traffic, visitor to lead rate, lead to opportunity rate, win rate, average contract value, sales cycle days, gross margin and gross revenue churn. Every input comes from your own trailing four quarters rather than a benchmark post. The model's real job is to expose lag, since spend in January produces revenue in April, and to rank which input moves ARR per dollar spent moving it.

## Key takeaways

- Eight inputs run the entire model. Everything else on a marketing planning spreadsheet is derived from them or is decoration.
- Spend in January becomes revenue in April at a 90 day cycle, so this quarter's budget cannot fix this quarter's revenue gap.
- A $24K ACV company needing $1.2M new ARR needs roughly 96 marketing-sourced opportunities and about 25,000 qualified visitors a year.
- Raising ACV 21 percent adds more ARR per dollar of effort than raising traffic 36 percent, and it lands two quarters sooner.
- Teams over-invest in traffic because traffic is the only input marketing can move without asking sales or the CEO for permission.
- Present a range with an assumption sheet, not a point estimate, or the first miss costs you the model's credibility permanently.

---

Most SaaS marketing plans are a list of activities with a budget stapled to the front. A model is a different object. It is a file where you change one cell, watch forecast ARR move, and can explain out loud why it moved. Build it once and next year's planning meeting stops being an argument about opinions and becomes an argument about three numbers.

The arithmetic is not hard. Eight inputs, a lag structure and a sensitivity tab. What makes it uncomfortable is that an honest model tells you things about your own team's priorities you would rather not hear.

## The eight inputs, and where each number actually comes from

Eight numbers drive the whole file: qualified traffic, visitor to lead rate, lead to opportunity rate, win rate, average contract value, sales cycle days, gross margin and gross revenue churn. Everything else on a planning spreadsheet is derived from those eight or is ornamental.

Each input has a source system and a predictable failure mode. Get the source wrong and the model produces a confident number resting on a definition nobody in the company agrees with.

Two of these will start a fight. Win rate is the first, because sales counts it on closed opportunities and marketing tends to count it on everything created, and the gap between those two definitions is usually 10 points or more. Average contract value is the second. Use the median by segment, not the blended mean, or one Salesforce-sized deal makes your whole funnel look twice as efficient as it is.

Gross margin belongs in a marketing model for one reason. It sets the ceiling on what you can spend to acquire a customer. At 80 percent margin a $24,000 contract throws off $19,200 a year of gross profit. At 55 percent it throws off $13,200, and a CAC that was fine becomes a CAC that breaks the business.

If marketing and sales cannot agree on what counts as an opportunity in one ten minute conversation, stop building the model. You are about to forecast a unit that two teams measure differently, and every variance review for the next year will end in a definitional argument instead of a decision. Write the stage definition into the model's notes tab and have both leaders initial it.

## Why January spend cannot fix a Q1 revenue gap

Because the money moves through a pipeline with a fixed transit time. At a 90 day median sales cycle, spend in January creates opportunities in January and February and produces closed revenue in April and May. The revenue that lands in March was paid for in December.

This is the single most useful thing a model does, and it is the thing teams resist hardest. Every February someone asks whether an extra $80,000 of paid search can rescue the quarter. The model answers no, in writing, with dates. That answer saves more money than any optimisation you will run all year.

Build the lag as a cohort offset, one row per month, with columns shifting right. January spend sits in the January row and its revenue appears in the April column. Do not apply a single quarterly average, because that smooths away the exact thing you are trying to see. Our [marketing pipeline forecast calculator](/calculators/pipeline-forecast-calculator/) does this offset for you if you want to sanity check the structure before committing to a spreadsheet.

Organic has a second lag stacked on top. Content published in January will not rank until roughly April and will not reach steady traffic until August, so an SEO investment made in Q1 produces its first closed revenue in Q4. If your model treats organic and paid as the same shape, it is wrong, and the [organic traffic forecast calculator](/calculators/organic-traffic-forecast/) exists because that ramp curve is different enough to deserve its own math.

**16 months** Median CAC payback across B2B SaaS, which sets the bar your model has to clear

## A worked model: $24,000 ACV and a $1.2M new ARR plan

Here is a complete pass for a Series A company selling at $24,000 ACV with a 90 day median cycle and a $480,000 annual marketing budget. The board approved $1.2M of new ARR. Marketing carries 60 percent of it and outbound plus partners cover the rest.

Work backwards from the revenue, then forwards from the traffic, and see whether the two ends meet.

| Stage | Calculation | Result |
|---|---|---|
| New ARR target | Board plan | $1,200,000 |
| New customers needed | $1,200,000 divided by $24,000 | 50 |
| Marketing-sourced share | 60 percent of 50 | 30 customers |
| Opportunities needed | 30 divided by a 31 percent win rate | 96 |
| Leads needed | 96 divided by a 16 percent lead to opportunity rate | 600 |
| Qualified visitors needed | 600 divided by a 2.4 percent visitor to lead rate | 25,000 |
| Monthly run rate | 25,000 divided by 12 | 2,083 visitors, 50 leads, 8 opportunities |

Ninety six opportunities. That is the number the whole plan hangs on, and it is eight a month, which is small enough that a single bad month is visible immediately. Two and a half wins a month. A marketing team that understands its model stops reporting session counts and starts reporting whether opportunity eight landed by the 28th.

Now the economics. $480,000 of loaded marketing spend divided by 30 marketing-sourced customers gives a $16,000 CAC. At 78 percent gross margin each customer returns $18,720 of gross profit a year, or $1,560 a month, so payback lands at 10.3 months. That sits comfortably inside the roughly 16 month median Benchmarkit reports for B2B SaaS, which is the sentence your CFO will actually repeat to the board.

New ARR is not growth. If this company starts the year at $2.4M ARR with 12 percent gross revenue churn, it loses $288,000 before adding anything. With $180,000 of expansion, $1.2M of new ARR becomes $1.09M of net growth. A model that stops at new ARR overstates the year by roughly 10 percent every single time.

## Top down and bottom up have to meet within 15 percent

Build both numbers separately, then put them next to each other and measure the gap. Top down starts with the revenue target and divides down. Bottom up starts with what each channel realistically produced last year plus a defensible improvement, and adds up.

| Channel | Opportunities, bottom up | Basis |
|---|---|---|
| Organic and content | 34 | 2025 actual of 27, plus 9 months of ramp on 18 new pages |
| Branded and non-brand paid search | 26 | Current spend held flat, cost per opportunity up 8 percent |
| Review marketplaces | 14 | G2 category placement maintained at current budget |
| Webinars and partner co-marketing | 11 | Four events, historical 2.75 opportunities each |
| Product-qualified signups | 9 | Self-serve trial volume converted at last year's rate |
| Bottom up total | 94 | |
| Top down requirement | 96 | |

A two opportunity gap is noise. Sign it and go. When the gap is 20 or 30 percent, you have a real decision, and there are only four honest ways to close it: more budget, a higher win rate, a higher ACV, or a lower target. Inflating a conversion rate until the columns agree is the fifth way, it is the most common, and it is how marketing leaders lose their jobs in month nine.

Put the gap on the first slide of the planning deck. A marketing lead who arrives saying "the plan is 22 percent short and here are the three ways to close it" is running the conversation. One who arrives with a model that happens to balance perfectly is going to be asked how, and will not enjoy the answer. If you are assembling the surrounding document, the [SaaS marketing plan template](/templates/saas-marketing-plan/) has the section order that keeps the gap visible rather than buried.

## Which input actually moves ARR, and what each one costs to move

In a multiplicative funnel every input is mathematically equal: lift any of the five volume inputs by 10 percent and ARR moves 10 percent. So the interesting question is never which input has the biggest coefficient. It is which input you can realistically move, by how much, at what cost, and how long it takes to land.

Run sensitivity on achievable twelve month ranges, not on identical percentage bumps. Here is the same worked company, one input at a time, holding everything else at base.

Traffic wins on the raw ARR column and loses on every other dimension. It costs $140,000 to buy $258,000, so you net $118,000 and you wait seven months for the first dollar. Raising ACV costs a pricing project and a fortnight of arguing, delivers $149,000 of new ARR, and then does the thing nothing else on that table does: it lifts the existing base at renewal. Apply a 21 percent list increase to 100 existing customers on a $2.4M base and you are looking at several hundred thousand dollars more, arriving without a single extra visitor.

That is the case for putting ACV and win rate at the top of your sensitivity tab. They are cheap, fast, and they compound into the retained base rather than only into new business.

Sensitivity analysis makes every input look independently adjustable. They are not. Push ACV 21 percent and your win rate usually falls a few points and your sales cycle stretches, because you have moved upmarket without saying so. Model the second-order effect or you will book the ACV gain twice. In practice, budget a 2 to 4 point win rate decline against any double-digit price increase and see whether the move still clears.

## Why marketing teams over-invest in traffic anyway

Because traffic is the only input a marketing team can move without permission. Win rate needs sales to change how they demo. ACV needs the CEO and product to change packaging. Lead to opportunity needs SDR routing that marketing does not own. Traffic needs a writer and a budget line, and marketing controls both.

So the org chart, not the arithmetic, decides where the money goes. Every quarterly plan I have reviewed at companies under $10M ARR allocates 60 to 75 percent of incremental budget to the top of the funnel, and the sensitivity tab almost never justifies it. The model does not fix this by itself. It just makes the bias visible enough that somebody in the room has to defend it.

The counter-move is simple and political. Put the sensitivity table in front of the CEO and sales leader together, and ask which of the five inputs each function is going to own this year. You are not arguing for less content. You are arguing that the two cheapest levers sit outside your department and need an owner assigned in the same meeting as your budget.

## The tab structure that keeps the file auditable

Six tabs, in this order, with no formulas that reach backwards. Anyone should be able to trace a number in Outputs to a cell in Inputs in under thirty seconds, because that is exactly what a CFO will do in front of you.

**Building the file**

Keep the file in one place and version it monthly rather than continuously. A model everyone edits live becomes a model nobody trusts by March. If you would rather start from something already wired up, the [SaaS growth model template](/templates/growth-model-spreadsheet/) has these six tabs built, and the [SaaS marketing budget template](/templates/saas-marketing-budget/) handles the loaded-cost allocation that feeds the Channels tab.

## How to present it to a CFO without losing the room

Lead with three numbers: CAC payback in months, marketing-sourced pipeline coverage against plan, and contribution after gross margin. Those are the three a finance leader can act on. Sessions, impressions and MQL counts do not appear anywhere in the first five minutes.

Then give a range, never a point. "Between $640,000 and $790,000 of marketing-sourced ARR, with the midpoint at $715,000" survives a miss. "$715,000" does not, and the first time you land at $680,000 the model loses its authority permanently.

Name the two assumptions the range depends on and state what would make you revise. Something like: this holds if win rate stays above 28 percent and the sales cycle stays under 105 days; if either breaks for two consecutive months, I will reforecast in week three of the following month. A CFO who has been handed a marketing forecast before will notice immediately that you told them what would falsify it, and that is the whole basis of the trust you are trying to build.

I do not need marketing's forecast to be right. I need to know which two numbers to watch so I find out it is wrong in April rather than September.

One more presentation detail. Bring the paid comparison. If you are asking for $140,000 to grow organic, show what the same $140,000 buys in paid search in the same period using the [SaaS PPC budget calculator](/calculators/saas-ppc-budget-calculator/), and be honest that paid wins on speed and loses on cumulative return by around month 18. The [SaaS SEO ROI calculator](/calculators/saas-seo-roi/) runs the other side of that comparison. A marketer who volunteers the case against their own request gets asked fewer hostile questions.

## Where the model is genuinely weak

It cannot model brand, and it cannot model the deals that arrive because someone read a post in 2024 and remembered you in 2026. Self-reported attribution surveys routinely return 25 to 40 percent of new customers naming a source your CRM never recorded, which means every model of this kind under-credits the slow channels and over-credits the last click.

It also breaks in the first twelve months of a company's life, when the sample is too small for any rate to be stable. Below roughly 30 closed opportunities in a segment you have an anecdote, not a conversion rate. For very small teams and pre-revenue products, the arithmetic here is theatre, and the [micro SaaS marketing playbook](/playbooks/micro-saas-marketing/) covers what to do instead until you have enough deals to measure.

And it says nothing about which tactics to run. A model tells you that you need 96 opportunities and that ACV is your cheapest lever. It is silent on how to get them, which is what [SaaS marketing ideas](/guides/saas-marketing-ideas/) and the broader [SaaS marketing](/saas-marketing/) hub are for.

## What to do this week

Open a blank sheet and fill in the Inputs tab only. Eight cells, each with a source and a date. Most teams discover at that point that two of the eight have never been measured, and finding that out is worth more than any forecast the finished file will produce.

Then build the Cohorts tab, because the lag structure is the part that changes behaviour fastest. Once your leadership team has seen in writing that December spend pays for March revenue, the mid-quarter panic budget request stops happening. Everything else on the file can wait a fortnight.

## Frequently asked questions

### What is a SaaS marketing model?

It is a spreadsheet that turns marketing inputs into forecast pipeline and ARR using your own measured conversion rates. Traffic flows to leads, leads to opportunities, opportunities to closed revenue, offset by the sales cycle so revenue lands in the month it will actually close. Unlike a marketing plan, a model lets you change one assumption and see the revenue consequence immediately.

### How many inputs does a SaaS marketing model need?

Eight is enough for almost every company below $50M ARR: qualified traffic, visitor to lead rate, lead to opportunity rate, win rate, average contract value, sales cycle days, gross margin and gross revenue churn. Models with thirty inputs are not more accurate, they are just harder to audit. Add channel-level splits before you add new stages.

### How do you forecast ARR from marketing spend?

Work forwards from spend to qualified traffic to leads to opportunities to wins, multiply wins by average contract value, then shift each monthly cohort forward by your median sales cycle. A 90 day cycle means January spend produces April revenue. Subtract gross revenue churn on the existing base to get net ARR growth rather than new ARR.

### Which marketing input has the biggest effect on ARR?

In pure percentage terms every input in a multiplicative funnel is equal, so a 10 percent lift anywhere moves ARR 10 percent. The difference is cost and lag. Average contract value and win rate move for the price of a packaging change or a sales enablement project. Traffic costs real money and arrives six to nine months late.

### How do you reconcile a top down revenue target with a bottom up marketing model?

Build both, put them side by side, and measure the gap as a percentage. Inside 15 percent, pick the bottom up number and move on. Beyond 25 percent, someone has to change something structural: headcount, pricing, win rate ownership or the target itself. Never quietly inflate a conversion rate until the two numbers agree.

### What sales cycle number should the model use, the mean or the median?

Use the median for the base case and the 75th percentile for the conservative case. The mean is distorted by a handful of deals that took fourteen months, which pushes your forecast revenue months later than reality for most cohorts. Show both in the model so sales cannot claim the timeline was invented by marketing.

### How do you present a marketing model to a CFO?

Lead with CAC payback in months, marketing-sourced pipeline coverage against the plan, and contribution after gross margin. Show a range rather than a point, name the two assumptions the range depends on, and state the trigger that would make you revise the forecast. A CFO trusts a model that says what would falsify it.