I've had some version of this conversation dozens of times in the last few months. A support leader is trying to make the case for new tooling, or headcount, or both. They know their team drives value. They can feel it. But when they sit down to build the business case, they get stuck.

They're reporting in a language their CFO doesn't speak.

Most support teams track CSAT, first response time, and resolution time. These are useful operational metrics. They tell you how your team is performing day-to-day.

But they don't connect to anything your CFO tracks.

Your CFO thinks in CAC, retention rate, expansion revenue, and margin. That's the language that unlocks budget. Everything else is noise to them.

When you walk into a budget conversation leading with "our CSAT is 94%," you've already lost the room. Not because it doesn't matter, but because it doesn't translate. It's like reporting temperature in Kelvin to someone who only understands Fahrenheit. The data is correct. The communication is broken.

Four metrics that translate

Same work your team already does, reframed in terms your CFO already tracks.

1. Support-touched conversion rate

What percentage of trial users, free users, or prospects who interact with your support team go on to convert?

This is one of the most undertracked metrics in SaaS. Most companies can tell you their overall conversion rate, but very few segment by support interaction. When you do, the number is consistently higher for the support-touched group.

Someone who reaches out during a trial is actively trying to make the product work. They're invested. A fast, helpful answer at that moment removes the last friction between them and a purchase.

2. Support-influenced retention

What's the retention delta between customers who received fast, high-quality support and those who didn't?

There's a difference between support that fixes the immediate issue and support that helps a customer get more value from the product. A customer asks how to export a report. One agent gives them the steps. Another agent gives them the steps and notices they haven't set up the integration that would eliminate the manual export entirely. Same ticket, same resolution time, completely different outcome for that customer's long-term success with the product.

When teams actually segment their retention data by support quality, the second group renews at a higher rate. Most teams never segment for this, which means the revenue impact of good support stays invisible to the people approving budgets.

3. Cost of inaction

Your CFO thinks about what support costs. Flip it: what does bad support cost?

A support leader knows they need better tooling or more people. Their CFO asks, "Is it really that much better? Why should we spend this?" And the support leader doesn't have a dollar figure for what's already being lost.

Every churned customer is lost expansion revenue and a higher acquisition cost to replace them. Bad support doesn't show up as a line item. It shows up as churn that everyone blames on "product-market fit" or "competitive pressure."

The cost of inaction is the number your CFO has never calculated but will immediately understand.

4. Agent leverage ratio

What does one support person actually produce? Not tickets closed or average handle time. Customers retained, product gaps identified, escalations prevented before they became fires.

The simplest version of this ratio: take the revenue your support team helped retain (from your cost of inaction calculation) and divide by support headcount. That gives you a dollar figure per person that your CFO can track over time. When you invest in better tooling or add a senior hire, that number should go up. That's the story.

AI and automation are changing the math here. They can absorb volume, and plenty of teams are leaning into that. But the companies that frame AI purely as a cost reduction tend to end up rehiring once quality drops. The right question isn't "how many people can we cut." It's "what could this team produce if they weren't buried in repetitive work."

The business case framework

You have the metrics. Here's how to structure the actual document. This part matters more than people think. I've watched good numbers fail because they were presented as a wall of data instead of a story a CFO could follow.

I talk to support leaders who are literally building these right now. The ones that land tend to follow the same storyline: this is what we have today, here's what we've evaluated, here's my recommendation. The sections below walk through that storyline, and Sections 2 and 3 are where you'll build the actual numbers for the four metrics above.

Section 1: Current state

What your team handles today. Volume, channels, headcount, tools. Just the facts. Include growth trajectory: if ticket volume grew 40% last year and headcount grew 0%, say that.

Where to find it: Your support platform's analytics dashboard. If you don't have clean data, a one-week manual count is enough to establish a baseline.

Section 2: Revenue connection

Map your support touchpoints to the four metrics above. You don't need all four. Pick the two that are strongest for your business.

For product-led growth companies, support-touched conversion rate is usually the strongest argument. For enterprise SaaS, support-influenced retention carries more weight. If your CFO is fixated on headcount efficiency, lead with agent leverage ratio and cost of inaction together. Show what you're losing now, then show what better tooling unlocks.

How to get the numbers: For conversion rate, pull a list of converted accounts from the last quarter and cross-reference which ones had a support interaction during their trial or evaluation. You'll have your baseline in an afternoon. For retention, compare renewal rates for customers with high support satisfaction versus those with low scores or no support interaction at all. Segment by resolution quality, not just speed. For agent leverage, take the retained revenue figure from your cost of inaction work and divide by support headcount.

If your systems aren't connected, start with a sample of 50 churned accounts and 50 retained accounts and manually check for support history. Imperfect data that tells a clear story beats perfect data you never collect.

Section 3: Cost of doing nothing

Model what happens if nothing changes. If volume is growing and headcount isn't, show where that line crosses. Where response times degrade and quality drops and churn starts accelerating.

How to get the number: Start with how many customers churned last year and their combined annual revenue.

Then sample 10-15 of those churned accounts and check their support history. How many had a negative interaction, or needed help and never got it? Apply that percentage to your total churned revenue. That's your support-influenced churn number.

Now add what it would cost to replace that revenue using your CAC, because your CFO is already thinking about that. The real cost of inaction isn't just what you lost. It's what you'll spend in sales and marketing to earn it back.

Section 4: The ask

Be specific. "We need more investment in support" is not an ask. "We need two senior support engineers and a workflow automation tool, costing $X annually, which we project will improve retention by Y points and save $Z in reduced churn" is an ask.

If you're pitching headcount alongside tooling, the framing that works is: "here's what our current team could produce if they had the right tools and one or two more senior people." That makes the spend feel like an unlock. The moment it sounds like "we need more people," you've lost the thread.

How to build it: Pull pricing for the tools you've evaluated. Get salary benchmarks for the roles you need. Then map each line item to one of the metrics from Section 2. Every dollar should point back to a specific projected improvement. If it doesn't, cut it from the ask or save it for a later phase.

Section 5: ROI framing

The strongest version I've heard went something like: "Yes, we're going to invest in this, but it means we save money here, here, and here." Just a clear return on a specific investment. Tie every dollar you're requesting to a projected return. Your CFO doesn't fund "better support." They fund fewer churned accounts and faster conversions.

How to build it: Take the cost of your ask from Section 4 and put it next to the cost of inaction from Section 3. That's your basic ROI story: we spend X to stop losing Y. If Y is bigger than X, the case makes itself. If it's close, layer in the conversion rate uplift or the agent leverage improvement to widen the gap.

Section 6: Measurement plan

How you'll prove it worked. Define the metrics you'll track, the baseline you're starting from, and the timeline for results.

Ninety days is the right window. Long enough to show a trend, short enough to maintain urgency. Commit to a check-in at 30 and 60 days with preliminary data.

How to build it: Pick two or three of the four metrics from this piece, lock in your current numbers as the baseline, and define what success looks like at 90 days. Keep it simple. A one-page table with metric, baseline, target, and check-in dates is all you need. If you overcomplicate the measurement plan, it signals that you're not sure what you're measuring.

It shows confidence in your own proposal, and it gives your CFO a low-risk way to say yes. They're not committing to a permanent cost increase. They're approving a 90-day experiment with defined success criteria.

Before you hit send

The document is the foundation. The conversation is where it lands or dies.

Know what your CFO is already worried about before you walk in. If they're under pressure on margins, lead with cost of inaction. If the board is pushing on growth efficiency, lead with support-touched conversion. Same data, different entry point. The business case doesn't change. The first slide does.

Have your 30-second version ready. If your CFO stopped you in the hallway and said "what's this proposal about," you should be able to say: "We're losing $X in preventable churn. For $Y, we can cut that by Z%. Here's how I'll prove it in 90 days." If you can't say that without looking at the document, you're not ready to send it.

Anticipate the AI question, because it's coming. "Can't we just automate this?" The answer isn't no. The answer is: automation handles volume, but the retention and conversion impact comes from the moments that need a human who understands the product and the customer. Your business case should already show this in the agent leverage ratio. If it doesn't, go back and add it.

And don't send it cold. Ask for 20 minutes. Walk them through the story. A business case that arrives as an attachment gets skimmed. One that gets presented gets discussed.

The hardest part of this isn't the math. It's accepting that your CFO is never going to learn to care about CSAT. That's not their job.

Your job is translation. And honestly, even with this framework, it's still hard. I've seen support leaders put together a great business case and still get told "not right now" because of timing or budget cycles or a CFO who just doesn't get it yet. You do the work, you bring the numbers, and sometimes the answer is still no.

But the ones who keep coming back with this data, same framing, updated numbers, eventually stop having the argument at all. The language starts sticking.