RWS paid 2x EBITDA for Acolad. Vitruvian won't say what it paid for Smartling.

What happened, six weeks apart

3 August 2026. RWS Holdings plc (AIM: RWS) announced via RNS a binding agreement to acquire Acogroup, the parent of French-headquartered language service provider Acolad. Enterprise value: £22.4 million (€26.0 million), stated as 2x adjusted EBITDA for the year to 30 September 2027. Total consideration, including about £17.8 million (€20.6 million) of cash already inside the business at completion, is £40.2 million (€46.6 million). The filing discloses Acolad's trailing financials directly: revenue of £182 million and adjusted EBITDA of £13 million for the year ended 31 December 2025. RWS CEO Benjamin Faes and Acolad CEO Bertrand Gstalder are both named and quoted. Funding is existing credit facilities, no new debt. Completion is expected in RWS's H1 FY27 at the latest, subject to French regulatory clearance.

15 September 2026. Smartling announced that Vitruvian Partners, described as "a leading international growth-focused investment firm," had taken a majority growth stake, buying out existing investor Battery Ventures. CEO Bryan Murphy stays in place. The release states Smartling serves "more than 1,000 global brands" and translates "billions of words annually," and names Vitruvian partner Tomer Yosef-Or, who is quoted saying "translation is one of the largest industries being reshaped by AI." No price, no percentage stake, no revenue figure, no valuation appears anywhere in the release.

Both targets compete in the same category RWS itself tracks: in BLOMEGA's earlier read of the 2026 Nimdzi 100, about 60% of the top 100 language service providers were active buyers and only 8% wanted to exit. These two deals are two instances of that same buying wave, disclosed in opposite ways.

What the filings actually say, side by side

One deal is a UK-listed company's regulatory disclosure. The other is a company-issued press release distributed through a newswire. That difference in obligation is most of the reason the two tables below look so different.

Two different things are being bought

The low multiple on Acolad is not a mystery once you separate what each buyer is acquiring. RWS is buying trailing revenue and EBITDA in a business whose growth has stalled enough that the market prices it near asset value. Vitruvian is buying a stake in a company whose growth equity investors do not need to justify the price with a public multiple, because there is no public shareholder base to answer to.

This is not evidence that AI-native platforms command richer prices than traditional LSPs, and we are not claiming that from two data points. It is evidence that a buyer comparing "the multiple the market is paying for localization companies in 2026" from press coverage alone will only ever see the regulated side of that market. The unregulated side, which includes most private growth-equity deals in this category, simply does not publish the number being compared.

What this means for anyone pricing a localization deal or vendor

A buyer or a localization lead benchmarking "what does a language service provider sell for" against public reporting will systematically see only the low end: public and quasi-public strategic deals with disclosure obligations, priced near asset value on trailing EBITDA. The growth-equity deals that might be setting a materially higher bar for AI-native platforms are invisible in that same sample, not because they are rare, but because press releases are voluntary.

For a vendor fundraising or selling: a strategic acquirer with public disclosure duties will put a number on the table that becomes permanently searchable, including the multiple. A growth-equity buyer has no such requirement, and sellers who prefer the terms stay private can route the transaction that way. Acolad's owners got a public 2x-EBITDA valuation on the record; Battery Ventures' exit price from Smartling did not.

For procurement comparing providers: Acolad's own disclosed numbers, £182 million revenue against £13 million adjusted EBITDA, a 7.1% margin, are themselves a data point about where AI-driven price pressure has already landed on a traditional, headcount-heavy LSP. That margin, not the acquisition price, is the more useful comparison figure for a buyer evaluating any mid-tier provider's unit economics.

Check it yourself

• RWS's RNS filing, with the full disclosed terms: investegate.co.uk, "RWS agrees to acquire Acogroup," 3 August 2026.
• Smartling's press release, with everything it does and does not disclose: Fortune / EZ Newswire, 15 September 2026.
• Recompute our two ratios yourself: EV £22.4m ÷ trailing EBITDA £13m = 1.72x. EV £22.4m ÷ revenue £182m = 0.123x. Both are arithmetic on RWS's own disclosed figures, not RWS's stated multiple, which is 2x against a forward EBITDA figure for the year to 30 September 2027 that RWS does not itself publish.
• Future RWS filings, to see whether the pattern holds on the next deal: search "RWS Holdings" on investegate.co.uk.
• BLOMEGA's earlier read of consolidation across the top 100 language service providers: nimdzi-100-2026-language-industry-mid-tier.

What would prove this wrong

If RWS's next disclosed acquisition of a traditional, non-AI-native language service provider prices above 3x trailing EBITDA before the end of Q2 2027, the "cheap scale" reading of this deal does not generalize and was specific to Acolad's own growth stall. If Vitruvian, Smartling, or Battery Ventures discloses a price or valuation figure for this transaction before that same date, the opacity was a choice about this announcement, not a structural feature of growth-equity deals in the category.