Selling an AI or Data Business: Valuation Multiples Explained

How small AI, SaaS and data businesses are commonly valued: SDE vs EBITDA vs revenue multiples, what pushes a multiple up or down, and a worked illustrative example.

Three ways buyers put a number on your company

When you sell a small or mid-sized AI, SaaS or data business, the price usually starts as a multiple of a financial metric. Which metric is used depends mostly on the size and profile of the business.

Seller's Discretionary Earnings (SDE). Common for owner-operated businesses, typically those earning under roughly $1-2M a year. SDE starts with pre-tax profit and adds back the owner's salary and benefits, one-time expenses, and non-cash items such as depreciation and amortization. It answers the question: how much cash does this business produce for one full-time owner-operator?

EBITDA. Earnings before interest, taxes, depreciation and amortization. It's used for larger businesses with a management team in place, where the buyer won't be replacing a working owner. Unlike SDE, a market-rate salary for management stays in as an expense.

Revenue (or ARR) multiples. Common for software businesses with strong recurring revenue and growth, especially when profit is deliberately low because the company is reinvesting. Annual recurring revenue (ARR) is often the headline metric for subscription businesses.

The formula itself is simple:

Enterprise value ≈ metric × multiple

All the negotiation happens over the right metric (and its adjustments) and the right multiple.

What pushes a multiple up

Buyers pay more for lower risk and higher growth. For AI, SaaS and data businesses, these characteristics tend to raise multiples:

• Recurring revenue. Subscriptions and multi-year contracts beat project work.
• Retention. Low churn and strong net revenue retention (customers expanding over time).
• Growth. Consistent year-over-year growth in revenue and earnings.
• Diversification. No single customer making up a large share of revenue.
• Gross margin. Software and data products with high gross margins.
• Transferability. Documented processes, a team that stays, and an owner who isn't the only person who knows how things work.
• Defensibility. Proprietary data, integrations, switching costs, or a niche the business clearly leads.
• Clean rights. Clear ownership of code, models, training data and customer data rights.

What pulls a multiple down

• Founder dependence. If the founder is the product, the sales team and the support desk, buyers discount for the risk of losing them.
• Customer concentration. One or two customers driving most revenue.
• Platform risk. Heavy dependence on one third-party API, model provider or marketplace whose terms could change.
• Unclear data provenance. For AI and data businesses especially, uncertainty about whether training data or customer data was properly licensed or consented is a serious diligence issue.
• Messy books. Mixed personal and business expenses, cash accounting with no reconciliation, or missing monthly financials.
• Declining trends. Shrinking revenue or margins, whatever the cause.

AI-specific diligence questions

Buyers of AI and data companies increasingly ask:

1. What models do you use, and are you on a third-party API, open-weight models, or your own?
2. What does inference cost per customer or per unit, and how sensitive are margins to provider price changes?
3. Where did your training and fine-tuning data come from, and do you have documented rights?
4. Do customer contracts let you use customer data to improve models?
5. How do you handle privacy, security and applicable regulations?
6. What happens to the product if a key provider changes its terms?

Prepare clear written answers before you go to market. They can shorten diligence and protect your multiple.

A worked example (illustrative)

Illustrative only. All numbers are invented to show the arithmetic. They are not market data.

Consider a fictional vertical SaaS company with ML-powered forecasting:

If a buyer and seller agree that a 4.0× to 5.0× SDE multiple reflects this company's growth, retention and risk, the implied value range is:

• 4.0 × $610,000 = $2,440,000
• 5.0 × $610,000 = $3,050,000

A buyer focused on recurring revenue might instead look at ARR. If $1.3M of the $1.45M is recurring, a 2.0× ARR multiple would imply about $2.6M, roughly the same range. Cross-checking methods like this is normal, and big gaps between methods are worth understanding before you set an asking price.

Deal structure matters too. The headline price may include seller financing, an earn-out tied to future revenue, or an equity rollover. A higher headline number with a large earn-out isn't necessarily better than a lower all-cash offer.

Asking price versus selling price

Listing sites usually show asking prices, which are the seller's opening position. Final prices are often different once diligence and negotiation are done. When you compare your business to listings, compare multiples, not dollar amounts. Look at businesses with similar size, growth and revenue model.

How to prepare in the 6-12 months before a sale

• Get monthly financials reconciled and, ideally, reviewed by a CPA.
• Separate personal expenses from business expenses, and document every add-back.
• Reduce founder dependence by documenting processes and delegating key relationships.
• Clean up contracts, IP assignments and data-rights documentation.
• Track the metrics buyers ask for: MRR/ARR, churn, net revenue retention, customer acquisition cost and gross margin.
• Decide what you want: a full exit, a transition period, or continued involvement.

Don't overlook the data itself

If your company has years of operating history in email, files and business systems, there may be value in licensing a de-identified copy of those records to AI developers, separate from the sale. Coordinate any license with your broker and attorney so it doesn't conflict with what a buyer expects to acquire.

You can also try the valuation calculator for a quick SDE × multiple estimate.

General information only, not financial, legal or tax advice. Talk to a qualified business broker, CPA and attorney about your situation.