Why buyers start with metrics, not the pitch
A founder knows the product inside out. A buyer has a few hours to decide whether to spend weeks on diligence. Metrics are how they make that call. Clean, consistent numbers say the business is run with care. Missing or inconsistent numbers say the buyer will have to rebuild the picture themselves, and they price that work and that uncertainty into the offer.
The metrics below are grouped in the order most buyers review them: revenue, retention, unit economics, AI specific costs, and operational health.
Monthly recurring revenue and annual recurring revenue
MRR is the recurring subscription revenue the business earns in a month, with annual plans converted to monthly equivalents. ARR is MRR multiplied by twelve. Both exclude one-off fees, setup charges and usage overages that do not recur predictably, unless the seller clearly shows them as a separate line.
Buyers want MRR at the end of each month for at least twelve months, so they can see the trend rather than a single point.
MRR movements
A single MRR number hides what happened inside it. Buyers ask for the breakdown each month.
- New MRR from new customers
- Expansion MRR from upgrades and added seats
- Contraction MRR from downgrades
- Churned MRR from cancellations
- Reactivation MRR from returning customers
This breakdown shows whether growth comes from new sales, from existing customers or from both, and whether losses are rising.
Revenue growth rate
Month-over-month and year-over-year growth in MRR. Buyers look at the trend and at what drives it. Growth from organic channels and word of mouth is valued more than growth bought with paid acquisition that would stop if spending stopped.
Customer churn and revenue churn
The share of customers and of recurring revenue lost each month. Our guide on SaaS churn rate explains the definitions and the maths in detail.
Net revenue retention
Net revenue retention, or NRR, measures how much revenue a group of existing customers generates now compared with a year ago, including expansion, contraction and churn but excluding new customers. Above 100 percent means existing customers grow faster than they leave. Buyers treat NRR as one of the clearest signs of product value, because it shows customers choose to pay more over time.
Cohort retention
The share of each month's new customers still paying after one, three, six and twelve months. Cohorts show whether retention is improving with newer customers, and they reveal early drop-off that a blended churn rate can hide.
Average revenue per account
MRR divided by the number of paying customers. Buyers use it to understand the customer base and to check whether price increases or a move toward larger customers are working.
Customer acquisition cost and payback
Customer acquisition cost, or CAC, is sales and marketing spend divided by new customers acquired in the same period. Payback is the number of months of gross profit from a customer needed to recover CAC. For a founder-led business with mostly organic growth, CAC may be low and hard to measure; say so and explain the channels.
Lifetime value
Average revenue per account multiplied by gross margin and by average customer lifetime. Buyers compare lifetime value with CAC to judge whether growth can scale profitably.
Gross margin
Revenue minus the direct costs of delivering the service, divided by revenue. For classic SaaS, direct costs are mainly hosting and support. For AI SaaS, model costs often dominate, which is why the next group of metrics matters.
AI specific metrics buyers now check
AI products carry costs and risks that traditional SaaS metrics were not designed to show. Buyers increasingly ask for these numbers separately.
Inference cost share
The share of revenue spent on requests to the AI model provider, measured monthly. A low and stable share means the business keeps most of each dollar. A rising share means margin is eroding as usage grows. Buyers often ask for inference cost per customer or per active user, broken down by plan, to see whether any plan loses money.
Gross margin after model and hosting costs
Gross margin with inference, embeddings, vector storage and hosting all counted as direct costs. This is the margin buyers actually use when they value an AI product. It is often lower than the margin founders quote when they treat model spend as a general expense.
Usage limits and heavy user exposure
Whether each plan has usage limits enforced in code, and what share of model spend comes from the heaviest users. A plan without limits exposes the business to a few customers who cost more than they pay.
Model provider dependence
Which providers and models the product uses, what share of requests goes to each, and how hard it would be to switch. Buyers also check whether current costs include promotional credits that will expire.
Quality and evaluation
Whether the business keeps evaluation sets, tracks output quality across model changes and owns the prompts and data that make the product work. Evidence of a repeatable evaluation process reduces the risk that a provider update breaks the product after the sale.
Operational and traffic metrics
- Traffic and its sources, verified at the analytics source, with the share from organic search, direct, referral and paid
- Trial or demo to paid conversion rate
- Support volume per customer and median response time
- Owner hours per week, and what those hours are spent on
- Uptime and incident history
- Refund and dispute rates at the payment processor
These numbers tell a buyer how the business runs day to day and how much work they are taking on.
Metrics for micro SaaS and young products
Not every business has two years of clean history and a finance team. Buyers of micro SaaS and young AI products know this and adjust what they ask for, but a few numbers remain essential at any size.
- Monthly MRR and customer count since the first paying customer
- Churn by month, with the absolute number of cancellations next to the percentage
- Model and hosting costs by month, taken from provider invoices
- Owner hours per week and a short list of recurring tasks
- Traffic by source for the last six to twelve months
For a small business, a short and honest set of numbers verified at the source is worth more than an elaborate dashboard built on estimates. If some data does not exist, say so and explain how you will collect it during diligence. Buyers usually respond better to a clear gap than to a number they later find was guessed.
How to present metrics so buyers trust them
- Use one definition for each metric and keep it the same every month
- Show at least twelve months, monthly
- Explain spikes and drops in one or two sentences each
- Separate recurring revenue from one-off income
- Reconcile metrics with payouts from your payment processor
- Verify at the source rather than sending screenshots
Sellers who prepare these numbers before listing shorten diligence and spend less time answering the same questions for every buyer. On Acquisitions, revenue, churn and traffic are verified through read-only connections to your payment processor and analytics, so buyers see the same source data you do. Read more on Revenue verification, and use the SaaS due diligence checklist to see what comes after the first review.
FAQ
MRR and its trend, churn, net revenue retention and gross margin come first in almost every review. For AI products, inference cost share and gross margin after model costs sit right next to them.
Anything above 100 percent means existing customers grow faster than they leave, which buyers value highly. What counts as strong differs by customer type, so buyers compare with similar businesses rather than a single benchmark.
Yes. Buyers treat inference and other model costs as direct costs of delivering the product. Counting them as general expenses makes gross margin look higher than buyers will accept.
At least twelve months, monthly. Two or three years of history makes trends clearer and supports a stronger position in negotiation.
Many sellers share verified headline metrics publicly and keep detailed breakdowns, customer data and the business name behind an NDA. On Acquisitions, the NDA gate controls what each buyer sees.
