TMS 2024 - Bigger Than Ever

by Konstantin Dranch & Jourik Ciesielski

This article outlines where translation and localization systems stand at the end of 2023. Jourik and I collected fresh figures on this market by interviewing founders, and we found unexpected results.

While LSPs are reeling, TMS are growing fast! It’s time to revise the market size from $0.2 to $0.33 billion. The opportunity is bigger than we envisioned before, and some companies are making a killing in the localization SaaS.

1. TransPerfect Surprise

The largest LSP in the world turns out to be the second-largest localization tech company in the world. We learned that their license revenues are at least $90 million, with over half generated by localization software (TMS, MT, website localization, app localization). TransPerfect never broke out the technology side from their billion-dollar revenue before, so the world didn’t know exactly how big their localization suite is. Well, now we do!

While RWS’s Language and Content Technology division is even bigger - circa $171 million in 2023, the TMS part is comparable with TransPerfect. Furthermore, RWS’s tech division has seen an erosion of its market share over the last few years, and TransPerfect grew. Buying TransPerfect’s TMS might be more of a closed-door enterprise deal than getting a Trados license, looks like localization’s biggest company sure knows how to be big in software as well as in services.

2. Fastest growing

• WeGlot has charted the sky like an inverted meteor. After raising $50 million in funding, this lofty French company grew sales from $11 to $27 million in two years, full-on SaaS, according to the founder Augustin Prot. That is most likely the fastest organic growth spurt in the history of TMS. How could that happen? Compared to mature enterprise TMS with heavy linguistic features, Weglot sells an easy-to-use website localization product to marketers and website owners. They solve a complex problem in a simple way, and their massive online marketing generates 0.4 million visits per month according to Similarweb. The outcome? A very scalable business that requires neither multi-year RFPs nor armies of enterprise salesmen. Investors must be giddy with anticipation - this one is the rare time when a big funding round in localization immediately paid off.
• Crowdin - went from $1 to $7 million in sales in 4 years. The commercial register in Estonia where the company’s key legal entity is registered, shows $5 million in sales last year and $7+ million projected for 2023. The most likely driver behind Crowdin’s rapid rise is its first sales team, and upped pricing to levels more comparable to the competition. I.e. the product was successful before, it was simply offered below its real value. Once the team got the taste of business, figures started mounting up. In parallel, very quietly, and without advertising at localization conferences, Crowdin has become the largest localization connector box in the world, with 600+ integrations advertised. Should they all work, it would mean that Crowdin currently offers more connectors than all other TMS and middleware combined.

3. Subtitling and dubbing tools

We estimate that both Ooona and CaptionHub have broken the $3 million barrier in SaaS.

• CaptionHub focuses on localization teams and referral partnerships with TMS and video hosting platforms.
• Ooona has become the go-to business management system for media LSPs and now expands to end clients.

Media localization technology has become crowded faster than it became big, but these two companies seem to have found a product/market fit better than others.

4. Steady growth

1. Phrase - revamped the team, added Orchestrator integration middleware, relaunched the LSP marketplace, and played around more with own machine translation products.
2. Lokalise - scaled down the team after a few years of venture-funded extravaganza, and immediately jumped onto the GPT train. “Lokalise AI” helped generate new sales in the hype wave.
3. Smartcat - upped pricing, focused on the e-learning industry, and now sells software, and payments, and offers services to create localization groups from the freelancers available on their marketplace. This is a strategy unique to the TMS space most likely inspired by Upwork.
4. XTM continued organic growth and showed signs of an appetite for more acquisitions. However, after the founder Bob Willans stepped back from the wheel, and veteran CRO Shamus Dermody left for a startup, we saw XTM in the field a bit less.
5. memoQ, the only top player still fully in the hands of the founders, diversified into a new type of business software and launched memoQ RFP. The translation software beloved by linguists and gaming companies grew somewhat less aggressively than the competition, and it remains slightly dependent on selling perpetual licenses to small and medium businesses. It’s good for the industry to have a mature TMS that doesn’t charge as much as the others and doesn’t bill you per word.

5. New Ambitious Entrants

These two companies have <$1 million in SaaS product revenue, but their marketing and product strategies allow them to start competing with the established players.

• Bureau Works - a vision that started 10 years ago now resulted in a bootstrapped TMS funded with LSP profits. The company endured a multi-year lean marathon to develop the product to maturity and to bring it into the saturated localization software market. This year, BWX spared no expense and made sacrifices to enter the racetrack with a thunder.?

• Tarjama Cleverso - the only localization product out of the MENA region with a level of product, commercial, and organizational maturity to start playing in the global RFPs. After years of toil and significant investments by Amethis fund, this tale of magical Orient displays digital might that sets them apart from any local competitor.

Product Trends

1. Peacockish GenAI. 2024 promises to be the year of the GenAI and LLM breakthrough in the TMS market. Crowdin and Trados launched AI assistants; memoQ, Bureau Works and Smartling blended GPT with translation engines, and LILT added a full-on LLM studio and a writing tool. At the most recent Adobe Summit, Jonckers presented a use case where they injected terminology in existing translation memory units. TMS rushed to add OpenAI-based features to their product portfolio, but we haven’t witnessed any significant impact on the client business yet, with a potential impact on translation memory, terminology, and quality management. For now, it’s tests and demos. For tomorrow, buyers expect TMS to integrate LLMs as the go-to-engines powering localization programs.?
2. Re: Integromats. With the launch of Blackbird and Phrase Orchestrator in 2023 and the pool of integrations at Crowdin reaching 600, the TMS market attempts to solve connectivity once again for the non-technical stakeholders. No-code integromats promote the ability to build workflows easily, bypassing heavier TMS products, especially in a no-human-in-the-loop scenario. In the past, multiple companies attempted this challenge with “headless TMS” and connector boxes, but all products remained tiny compared to non-localization integration platforms (IPaaS) such as Mulesoft, Zapier, Make, and Snaplogic. The next year brings in a fresh take on the problem.

1. Data-driven localization & and cost reduction. Phrase and memoQ relaunched their analytics suites to allow clients to customize reporting and track KPIs. We anticipate other TMS providers to follow suit, the purpose being to enable users to make informed decisions about their localization strategy.
2. Overcrowded media localization tech. More TMS providers, from memoQ and Trados to Smartcat and XTM, moved into subtitling by adding live video previews, although no speech-to-text and text-to-speech functionality yet. AI dubbing remains the prerogative of specialized products. We’re losing count of products with “dub” and “.ai” in the name, including deepdub.ai , dubdub.ai , Dub Dub Dub, dubformer.ai , and dubverse.ai . While buyers are still suspicious of the robot quality and major studios such as Paramount are only starting with tentative synthetic voice RFPs amidst the Hollywood actor strikes against AI, the vendor side has already turned into a red ocean.

This is our first published collaboration with Jourik Ciesielski . Stay tuned for more!