Multiples in Practice: Size, Industry, and What the Data Really Says

How to Read a Multiples Table (Including the Ones in This Chapter)

Picture the owner of a commercial services company earning $3 million in EBITDA. At an industry conference he hears that a competitor “got 8x.” Nine months later his own sale process produces two strong offers at 5.5x, and he declines them as insults. Eighteen months after that, with a softer market and a thinner backlog, he accepts 4.9x. The 8x story was true. It was also an enterprise-value multiple on aggressively adjusted EBITDA, for a company twice his size, with 70% of revenue under multi-year contract, and a meaningful piece of the price sat in an earnout that never fully paid. He priced his company on someone else’s headline, and the anchor cost him seven figures and two years.

This chapter contains more numbers than any other in the guide, which is exactly why it opens with the rules for reading them. Multiples are the industry’s shorthand, and shorthand is useful right up until someone mistakes it for the arithmetic underneath. Five habits protect you.

First: a multiple is an output, not a mechanism. No serious acquirer decides to “pay 6x.” They model the company’s future cash flows, the risk to those cash flows, and the financing available against them, arrive at a price they can defend, and the multiple is simply that price divided by earnings. When you treat the multiple as the input, you skip the part where all of the value actually lives. Two companies with identical EBITDA are identical on exactly one line of a hundred-line model.

Second: always ask, a multiple of what? The earnings base matters as much as the number. Seller’s discretionary earnings, reported EBITDA, adjusted EBITDA, and recurring revenue produce very different prices at the same multiple, and a 5.0x offer on a defensible adjusted EBITDA can be more money than a 6.0x offer on a number that will not survive diligence. Confirm, too, whether the multiple describes enterprise value or equity proceeds; the difference is the debt.

Third: averages hide dispersion. Every published average or median compresses a wide distribution into a single number. Around any figure in this chapter, real transactions print far above and far below it, and the useful question is never “what is the average” but “what separates the top of the range from the bottom.” The average of a wide distribution is a poor forecast of any single company’s outcome, including yours.

Fourth: structure inflates headlines. A “7x deal” in which a quarter of the consideration is an earnout that may never pay is not a 7x deal in guaranteed money. Reported multiples, whether in databases or in war stories, usually count contingent consideration at full face value and say nothing about seller notes, holdbacks, or working capital terms. Headline multiples skew high relative to cash at close, systematically.

Fifth: samples are not you. Every dataset describes a specific population over a specific period, and all of them share a survivorship problem: only completed transactions report. The companies that went to market and came home unsold left no trace in any average. A multiples table is the winners’ scoreboard.

One more truth binds the five together: the same company can clear meaningfully different multiples in different processes. Change the buyer mix, the competitive tension, the timing, or the quality of the preparation and story, and the print moves. A valuation predicts; a process decides. Hold every number that follows, ours included, at that arm’s length.

The Size Effect: Why Bigger Earnings Command Bigger Multiples

The most reliable pattern in private-market pricing is simple: the more EBITDA a company generates, the higher the multiple of that EBITDA acquirers will pay. Four forces produce it.

Risk falls with size. A company with $10 million of EBITDA typically has more customers, more managers, more service lines, and more geography than one with $1.5 million. The shock that would cripple the smaller company, a lost top customer, a departed operations leader, one bad quarter, is absorbable at scale. Acquirers pay for durability, and size is a crude but honest proxy for it.

Financeability rises with size. Most acquisition prices are built on borrowed money. Lenders compete harder, extend more, and price tighter for larger borrowers with reviewed or audited financials and management depth, and because an acquirer’s capacity to pay is partly a function of what can be financed, lender appetite flows directly into purchase price.

The buyer universe deepens with size. Below roughly $2 million of EBITDA, the realistic acquirer pool is individuals, search funds, and smaller strategics, many financing through SBA programs or seller notes. Between $2 million and $5 million, private equity platforms begin to compete in earnest. Above $5 million, essentially every acquirer type shows up: platforms, well-funded strategics, family offices. Each added buyer type deepens competition, and competition is pricing power.

Infrastructure exists at size. Larger companies are likelier to have a real management layer, real systems, and financial statements an acquirer can rely on, so the price no longer discounts the cost of building all of that after close.

Notice what this implies: the size effect is a staircase, not a slope. Multiples step up where new buyer categories enter the bidding, which is why crossing a threshold near $2 million or $5 million of EBITDA changes not just the number but the character of the sale process itself.

Two third-party reference points are worth carrying, both read with the habits above. GF Data, an ACG company that tracks private-equity-sponsored transactions of $10 million to $500 million in enterprise value, reports that average purchase-price multiples held steady at 7.2x adjusted EBITDA for full-year 2025. Note the population before you borrow the number: institutional deals starting at $10 million of enterprise value, at or above the top of many lower-middle-market outcomes. A $2 million EBITDA company treating 7.2x as its own comp is committing the sample error described above. At the smaller end of the market, the IBBA and M&A Source Market Pulse survey of business brokers and M&A advisors reported Q3 2025 median multiples of 2.0x SDE for businesses under $500,000 in value, 2.8x SDE for $500,000 to $1 million, 3.3x SDE for $1 million to $2 million, 4.0x EBITDA for $2 million to $5 million, and 5.3x EBITDA for $5 million to $50 million deals. Those are medians, each compressing its own wide spread, and the real message is the ladder itself: pricing climbs with size all the way up the market.

The table below shows typical enterprise-value multiple ranges by EBITDA size for lower-middle-market services businesses, based on FIH transaction experience; every company prices on its own facts.

Why Industry Moves the Number

Industries price differently for structural reasons, not fashionable ones. Five characteristics do most of the work: expected growth, the share of revenue that recurs without being re-sold, capital intensity (how much of EBITDA must be reinvested just to stand still), cyclicality, and fragmentation, because fragmented sectors attract consolidators and consolidators are a standing bid. Every sector paragraph below is those five variables in different proportions.

The table below shows typical multiple ranges by sector for lower-middle-market companies, based on FIH transaction experience; every company prices on its own facts.

Business services. Commercial cleaning, testing and inspection, environmental services, staffing, agencies: the category is broad, but the pricing question is always the same. How much of next year’s revenue is already sold? Route-based and contract-based models with auto-renewing agreements price at the top of the range because the acquirer is buying a book of committed revenue rather than a sales engine that must refill itself every year. Project-driven and relationship-driven firms price lower, and firms whose delivery depends on scarce credentialed people carry a discount for labor risk. Capital intensity is low, which flatters cash conversion, and most services niches are fragmented enough that active consolidators are always shopping for well-run companies of platform size.

IT services and managed service providers. Acquirers sort this sector with one blunt instrument: the share of revenue that is contracted monthly recurring revenue, as opposed to projects and hardware resale. High-MRR providers with per-seat or per-device contracts, low churn, and a standardized technology stack price toward the top of the range and attract the consolidators who have been rolling up the sector for years. Project-heavy IT shops price like ordinary services businesses. And at the small end, an owner-operated MSP where the founder holds the client relationships and the senior engineering knowledge prices on SDE like any other owner-run company, whatever its technical credentials.

Software and SaaS. The exception on the earnings base: growing software companies are typically priced on annual recurring revenue rather than EBITDA, because acquirers are buying growth and the income statement is suppressed by reinvestment. Typical ARR multiples run from 2.0x to 8.0x and beyond, a range that wide because growth rate, net revenue retention, and gross margin do all of the work. A fast-growing product that keeps and expands its customers is a fundamentally different asset than a slowly leaking license base, and the market prices them that way. Mature, modestly growing software with real profits trades on EBITDA like a services business. No sector in this guide has a wider gap between its best and worst outcomes. Chapter 19 reads valuation, diligence, and structure through the technology lens in full.

Healthcare services. Demographics supply the demand story, recurring patient relationships supply the revenue quality, and a decade of consolidation supplies the buyer depth: dental, veterinary, behavioral health, home care, and physician specialties have all seen sustained roll-up activity. The cap on pricing is reimbursement. Private-pay and commercially insured models price higher than models dependent on government reimbursement rates that can change by rule. Compliance and licensure diligence runs heavier here than in any other sector in this chapter, and the owner-clinician who is also the practice’s production engine is healthcare’s version of a key-person problem every sector shares.

Manufacturing. The spread runs from commodity job shops quoting work order by work order to proprietary niche products with engineering content, certifications, and pricing power. Acquirers price capital intensity explicitly: EBITDA overstates free cash flow in machine-heavy operations, so sophisticated buyers underwrite EBITDA minus normalized capital spending even when the headline is an EBITDA multiple. OEM customer concentration is common and gets priced or structured around. Aerospace, medical, and defense certifications, documented processes, and a real second layer of production management push companies toward the top of the range; dependence on the owner’s tribal knowledge pushes them toward the bottom.

Distribution and logistics. Position in the flow of goods is everything. Distributors with exclusive lines, technical value-add (kitting, vendor-managed inventory, application engineering), and contracted customer relationships earn the top of the range; pure pass-through distribution with thin margins prices near the bottom, because there is little in it an acquirer could not replicate with a purchase order. Working capital is the sector’s sleeper issue: inventory and receivables absorb cash as the business grows, and the working capital terms of a distribution deal move real money. Asset-light brokerage models and asset-heavy fleet operations price on different logic and should never be compared casually.

Construction and trades. Project revenue is the penalized revenue: every January the backlog must be rebuilt from zero, and acquirers discount for it. The premium sits in service and maintenance. HVAC, plumbing, and electrical companies with dense books of recurring service agreements price like services businesses, which is precisely why consolidators have pursued them so aggressively. Backlog quality, margin discipline in bidding, bonding capacity, end-market cyclicality (municipal, commercial, or residential), and licenses held personally by the owner are the recurring pricing levers. Two contractors with identical revenue can sit at opposite ends of the range on revenue mix alone.

E-commerce and consumer. The widest dispersion for its size band of any sector here. A brand with genuine customer loyalty, repeat purchase, first-party customer data, and diversified channels is a durable asset and prices like one. A traffic-arbitrage operation dependent on a single marketplace or a single advertising channel can lose its economics to an algorithm change it reads about after the fact, and acquirers price that fragility with low multiples, heavy earnout structures, or both. Advertising cost sensitivity, inventory working capital, and platform concentration are underwritten hard. This is the sector where the difference between a brand and a sales channel becomes a seven-figure question.

The Smaller End: Why SDE Multiples Look Low (and Are Not)

Below roughly $1 million to $1.5 million of earnings, the market prices owner-operated companies on seller’s discretionary earnings, and typical SDE multiples in our experience run 2.0x-3.5x. Owners who set that beside the EBITDA multiples above conclude that the small end is brutally cheap. Most of that conclusion is an illusion, because the bases differ.

SDE adds the owner’s full compensation back into earnings; EBITDA charges the business for a market-rate manager. Work the arithmetic. A business generating $900,000 of SDE sells at 3.0x, or $2.7 million. To restate the same deal on an EBITDA basis, deduct, say, $175,000 for the general manager the acquirer must hire: $725,000 of EBITDA. The identical $2.7 million price is now 3.7x. Nothing changed except the denominator. Comparing an SDE multiple to an EBITDA multiple is comparing fractions with different denominators and calling the smaller one unfair. The Market Pulse medians quoted earlier show the same transition inside a single dataset: the survey itself switches from SDE to EBITDA as deal value crosses $2 million, exactly where the market does.

Some of the discount is real, though: small owner-run companies do price lower in real terms, because they carry precisely the key-person and concentration risks the size section described, and no restatement of the denominator removes those.

The Spread Within an Industry Beats the Average Across It

If the sector table tempts you to shop industries (“healthcare gets 8x, we should buy a clinic”), stop. In our experience, the spread between a well-run and an ordinary company in the same sector is consistently wider than the difference between sector averages. Industry sets the neighborhood; the company picks the house.

That spread is not noise. It has named causes: revenue quality, retention, concentration, key-person dependence, margin structure, systems. Chapter 9 takes each one apart and shows how acquirers price it, because the gap between 4.0x and 7.0x is the most actionable money in this guide.

Multiple Expansion: The Re-Rating Acquirers Underwrite

One more pattern explains a great deal of acquirer behavior. Because multiples climb with size, an acquirer who buys a $3 million EBITDA company and builds it to $8 million expects to sell not just more earnings but a higher multiple on all of them. That re-rating, called multiple expansion, is underwritten in nearly every private equity model, and it is why platforms pay premium prices for scarce platform-quality companies and then buy smaller add-ons at lower multiples to blend down their average entry cost.

Owners can capture part of this arbitrage by selling into it. A consolidator underwriting an exit re-rating can rationally pay forward some of it today, which is why the right moment in a sector’s consolidation wave can matter as much as the company’s own numbers. And an owner sitting just below a size threshold faces a genuine question about whether to sell now or grow across the line first, a question with honest arguments on both sides. Chapters 15 and 16 give owners and acquirers, respectively, their playbooks.

The Bottom Line

• A multiple is the output of cash flow, risk, and financing math, never a price you can select from a table.
• Before comparing any two multiples, confirm the earnings base, the structure behind the headline, and the population the number came from.
• Size is the most reliable multiple driver: in FIH’s experience, typical services ranges run from 3.5x-5.5x at $1M-$2M of EBITDA to 7.0x-10.0x+ above $10M, and pricing steps up where new buyer types enter.
• Industry sets the neighborhood, but the spread inside a sector is wider than the gaps between sector averages.
• SDE and EBITDA multiples use different denominators, and converting between them removes most of the apparent cheapness of small companies.
• The dispersion around every average is where the money is, and Chapter 9 covers the KPIs that decide which side of it you land on.