Almost every SME valuation that comes to us for a second opinion turns on the multiple. The methodology is correct, normalisation is defensible, the numbers are clean, but the multiple in use is from a three-year-old report, at an aggregation level that hides 5× variation per sub-segment. The result looks professional and lands 30 to 50% off market.
What a multiple actually measures
An EBITDA multiple is implicitly the sum of three variables: the return a buyer finds acceptable, the growth profile baked into the price, and the risk profile of the specific company against the sector. A sector median of 5.2× with a 3.5× to 7.8× spread already implies a factor of 2.2 between the 25th and 75th percentile. The multiple itself tells no story. The company's position within the spread does, and that position is what most SME valuations fail to substantiate.
Four sources for defensible multiples
- Comparable SME transactions (local, recent, scaled). The gold standard. Real Benelux deals over the last 24 months in the sub-segment and size band, Upswitch publishes aggregated sub-sector multiples as the Upswitch Index and per-country at markets/country.
- Damodaran data. Industry medians and spreads from NYU's Aswath Damodaran. Mostly US-listed, useful for relative ratios and intra-sector dispersion, not for direct application.
- Listed peers (with liquidity discount). Where Europe-listed peers exist, apply a 20 to 35% liquidity discount before using.
- Sector overviews from M&A houses. PwC, KPMG, EY, Argos, Mergermarket. Useful for orientation, rarely granular enough at sub-segment level.
Why sub-segment matching is everything
"Manufacturing" is not a sector. The bucket contains both a high-volume packaging line and a low-volume precision supplier to automotive Tier-1. The defensible multiple gap is a factor of 2.
Worked example. Two businesses, both labelled "industrial metalwork," each €4M EBITDA. (A) Low-volume high-precision CNC for automotive Tier-1, 12-year contract with OEM customer: defensible multiple 6.5× to 8×. (B) High-volume standard steelwork for construction, no contracted customers, cyclical: defensible multiple 4× to 5×. Both get the same NACE-aggregate median in a shallow analysis. The difference between 7× and 4.5× on €4M EBITDA is €10M. Not a detail.
How to match sub-segment in practice
- Sub-sector activity. Not "manufacturing" but "precision CNC for automotive." Not "IT services" but "Microsoft Dynamics implementation partner."
- Size band. €1M EBITDA vs. €10M EBITDA is a size-spring comparison; SME multiples typically compress with size.
- Recurring-revenue mix. 70% recurring vs. 100% project revenue in the same sector justifies 2× difference.
- Geography. Benelux multiples typically run 10 to 25% below comparable US deals. Ignoring this drifts the answer materially.
When a multiple approach falls short
- Owner-operator businesses → SDE, not EBITDA-multiple, is the headline.
- SaaS above $1M ARR → ARR-multiple is dominant.
- Asset-heavy structures (real estate, fleets) → adjusted NAV leads.
- Strong growth where current EBITDA is not representative → revenue × growth-adjusted multiple works better.
Running two or three methods in parallel is normal. Upswitch Market Approach blends them with explicit weighting, no black box.
How Upswitch operationalises this
Every valuation matches the specific business across four dimensions (sub-sector at NACE/SBI level 4 or 5, size band, recurring mix, geography) against our Benelux transactions database. The output is a P25/P50/P75 band plus a coverage score (how many comparables underlie the band) and a sensitivity on the top-3 parameters. A low coverage score is information, not a flaw, it tells you to widen the band and weight a DCF on explicit assumptions in parallel. That is what defensibility means: not pretending precision, but making your real uncertainty explicit. The same dataset is published as the Upswitch Index, deep-link directly to your parent industry, the country-scoped catalog or the methodology underlying every band.
Frequently asked questions
Where do I source comparable SME transactions?
For Benelux, public access is limited. Upswitch publishes aggregated sub-sector multiples at /benelux-multiples. M&A advisors typically work with Mergermarket, Pitchbook or Capital IQ. M&A-house sector overviews (PwC, KPMG, EY, Argos) are useful for orientation but rarely specific enough.
How many comparables do I need to be defensible?
A solid minimum: 5 to 8 transactions over the last 24 months, comparable sub-segment and size band. Below 5, each datapoint becomes a big variable; above 15, dispersion is usually more informative than the median.
Can I use US-listed multiples for a Belgian SME?
With material adjustment, yes. First a geographic correction (typically −10 to −25% for Benelux), then a liquidity discount (−20 to −35% private vs. listed), and always cross-check with local Benelux data. Direct application without adjustment overstates by 50%+.
What if my sector has no recent comparables?
Fall back on a blend of Damodaran aggregates (for relative ratios), listed peers (with liquidity discount), and a DCF on explicit assumptions. Be explicit about lower certainty, a defensible 4× to 7× band is more honest than a fixed 5.5×.
Upswitch is the M&A infrastructure layer for the European SME economy. Defensible valuations and structured transaction matching for the lower mid-market.
Continue reading
See it on the Upswitch Index
Live multiples for the sectors this article touches
Each link opens the live published EV/EBITDA, EV/Revenue and P/E bands per business type. Anchored at the right parent industry on the Upswitch Index.
Manufacturing
Sub-segment dispersion is huge. Aggregate manufacturing multiples are almost always wrong.
Open on Index→
IT & SaaS
ARR vs. services revenue mix swings the multiple by 3 to 6×. Mix-discipline is the headline.
Open on Index→
Healthcare practices
Regulated practice-economics keep the band tight; comparability is the lever.
Open on Index→