Guides
Translation agency trends: six shifts to test in quote logs
Use your quote log to test six translation agency trends across service mix, batch size, asset requests, locale demand, multimedia, and language-access work.
What to take away
- Treat translation agency trends as hypotheses to test against your quote log, not claims that every market is moving the same way.
- Six possible shifts are worth testing, and your records can show whether each is appearing in your business.
- A mix shift can matter even when total volume stays steadywork may move between service lines rather than disappear.
- Test any trend claim with three questions before you act on it.
- Your quote log records requests reaching your agency; consistent tags make patterns easier to interpret.
How to test a trend claim
Before you plan anything around a claim you read somewhere, ask three questions.
Test a trend claim
Who measured it, and how?
treat as advertisement
check your quote log
Who measured it, and how?
Does it appear in my own quote log?
What would it change on Monday?
Six shifts to test
1. The mix moves before the volume does
Work does not usually vanish. It moves: from full translation toward post-editing, from documentation toward interface strings, from long batches toward frequent small ones.
Six shifts and what they change
Shift
- Mix moves
- Recruit post-editing skills
- Release cycles
- Per-batch setup cost dominates
- Clients own assets
- Settle ownership and export position
- Markets not languages
- Recruit in-market reviewers
- Multimedia share rising
- Build line or subcontract openly
- Institutional language access
- Ask client's authority
What it changes
- Mix moves
- Price line separately
- Release cycles
- Small frequent work unprofitable
- Clients own assets
- Markets not languages
- Multimedia share rising
- Institutional language access
- Check what work pays
Check it in your own data: compare the share of your revenue by service line this year against last. A falling line and a rising line together is a mix shift, not a decline.
What it changes: the skills you recruit for and translation agency pricing and margin guidance. Post-editing has its own definition of done and its own staffing, and treating it as discounted translation is where agencies lose margin quietly.
2. Release cycles replace publication dates
Buyers who once shipped a manual now ship continuously. Content arrives in small batches, often weekly, with no natural freeze point.
Check it: count your projects per client and the average size of each. Rising count with falling size is this shift arriving.
What it changes: per-batch setup cost becomes the dominant cost. If your process assumes a large project, small frequent work will be unprofitable at the same rate.
3. Clients start owning their linguistic assets
A client-ownership shift to test is whether procurement teams ask about memory and termbase ownership, export formats, and what happens on exit. Make the answer concrete: test TMX (Translation Memory eXchange) for translation-memory data, TBX (TermBase eXchange) for terminology, and XLIFF (XML Localization Interchange File Format) for localization content with a sample export and re-import.
Check it: count how many of your last ten client agreements or questionnaires raised it.
What it changes: your answer needs to be settled and stated. It also affects pricing, because a client who owns the memory has a stronger argument for match discounts.
4. Buyers ask for markets rather than languages
The distinction between a language and a locale reaches more buyers every year, usually after a bad experience.
Check it: count how many recent requests specified a country or region rather than only a language.
What it changes: you need reviewers in the markets you serve, separate style guides per market, and quotes that name markets explicitly.
5. Multimedia grows as a share of what needs translating
More content is video and audio, which brings subtitling, captioning, and voice work into requests that used to be documents only.
Check it: count requests that mention video, and how many you declined or subcontracted.
What it changes: script freezes, reading-speed constraints, and a different supply pool. It is a translation agency service-package build, not an add-on.
6. Language access requirements in institutional settings
Language-access obligations depend on the country, sector, and institution. In the United States, Title VI of the Civil Rights Act of 1964 and Section 1557 of the Affordable Care Act prohibit discrimination based on national origin, including discrimination based on the ability to speak English; Section 1557 also prohibits discrimination in health programs and other services that receive federal financial assistance, as KFF's summary of these protections explains. More specific obligations are set by law and by the relevant authority.
Check it: ask institutional clients to identify the applicable law or program, the responsible authority, and any written plan, notice, form, or process they follow; record the exact names they provide.
What it changes: the work is often recurring and often needs both written and spoken services. Never state such a requirement to a client as fact yourself; direct them to their own counsel or regulator and quote what that body says.
What outside sources can tell you
For occupational context, not current industry demand data, the Bureau of Labor Statistics profile of interpreters and translators describes both professions in a United States context, including the spoken and written distinction and the role of research, terminology work, and subject specialization.
Treat that profile as occupational background, not an industry demand series; it does not establish whether any of the six shifts is happening in your market.
For general business questions a shifting market raises, including structure, insurance, and growth planning, the SBA guide to planning and growing a business sets out topics to settle with a qualified adviser where you operate. It is general planning guidance, not a source for translation-industry demand trends.
Read your own quote log
Quarterly quote log review
- Read every quote issued
- Read every quote lost
- Tag service line
- Tag locale set
- Tag content type
- Tag reason for outcome
- Flag refused work as demand
That exercise answers demand questions no article can. It tells you which lines are being asked for, which are being declined on price, and which requests you are refusing because you cannot staff them. The last group is the most valuable: refused work is demand you already have.
Make the log comparable by recording client, quote date, service line, locale, requested volume, quoted amount, outcome, decline reason, and available capacity. Calculate win rate as won quotes divided by quotes marked won or lost, and group declined quotes by reason; keep no-decisions separate. Where file analysis is available, track translation-memory match share as matched source words divided by total source words; use delivered revenue from project records, not quoted value, to compare revenue by service line.
What each shift means operationally
| Shift | What to change first |
|---|---|
| Mix moves toward post-editing | Write the definition of done, and price the line separately using the pricing and profit guide |
| Small frequent batches | Reduce per-batch setup cost in the process described in onboarding and delivery |
| Clients owning assets | Settle your ownership and export position, and state it publicly |
| Markets rather than languages | Recruit in-market reviewers, as covered in the hiring and training guide |
| Multimedia share rising | Build the line properly or subcontract it openly |
| Institutional language access | Ask the client's authority, and check what the work pays using the break-even arithmetic |
Rising demand in a pair also shows up as pressure on what linguists will accept, which is the early signal described in the translation agency labor-cost guide.
Common questions
Is machine translation reducing demand for human work?
It is changing what the human work is on a lot of content and leaving other content untouched. Measure the mix in your own revenue rather than accepting either the optimistic or the pessimistic version.
Should I invest ahead of a trend?
Only where the investment is useful anyway. Recruiting a reviewer in a market you already serve is useful whether or not the trend arrives. Buying tooling for a line you have never sold is not.
Why are there no percentages in this article?
Because this article is a method for testing demand against your own quote log. An industry percentage without a stated method would not establish what is happening in your service mix, client base, or market.
What is the single most reliable demand signal?
The work you turned down. It is demand you already reached, and it is the only signal on this page you do not have to interpret.
