A request for five languages at 4.45pm
A client emails a zip file with twelve documents and asks for a price into French, German, Spanish, Polish and Arabic, 'as soon as possible'. Your project manager downloads the files. Three are PDFs, one is a scanned image, two are Excel files with hidden columns. They create a project in Trados Studio or memoQ, run the analysis, copy the word counts and fuzzy matches into a spreadsheet, look up the client's rates for each language pair, check whether any language falls under the minimum fee and work out whether the requested deadline needs an urgency charge.
An hour later the quote goes out. By then the client has had two other quotes back, one of them from an agency with an instant online price.
Why quoting is slow
- Every quote starts with manual file handling: downloading, unzipping, converting.
- The CAT tool analysis is run by hand, and the results copied out by hand.
- Rates vary by client, language pair, subject area and service level, and live in spreadsheets.
- Minimum fees, repetitions and fuzzy match discounts are applied by the PM from memory.
- Deadlines depend on translator capacity, which is not visible at quote time.
None of these steps need a project manager's judgement for a standard request. All of them take a project manager's time.
What slow quotes cost
Lost work, above all. Direct clients often ask several agencies and go with the first credible answer. Project managers spend time quoting jobs that never come in, while live jobs wait. Mistakes creep in, such as a missed file, a wrong rate or a forgotten minimum, which either loses money or causes an awkward revision. And clients get different pricing logic from different PMs.
| Quote step | Manual | Automated |
|---|---|---|
| Receive files | Download and unzip from email | Uploaded through intake form or portal |
| Prepare files | Convert by hand | Converted, scans flagged for OCR |
| Analysis | Run in CAT tool, copy out | Run via CAT tool or TMS API |
| Pricing | Look up rates, apply discounts | Rate card and match bands applied |
| Deadline | PM's guess | Suggested from word count and your throughput rules |
How we build quoting
- Clients upload files through a quote form on your website or a client portal, choosing languages, subject, service level (translation, translation plus revision, certified) and deadline.
- Emailed requests to your quotes inbox are handled too: attachments are extracted and the request is parsed into the same fields for a PM to confirm.
- Files are converted where needed, and anything that cannot be analysed, such as scanned images, is flagged for a PM rather than silently undercounted.
- Word counts and match analysis are run through your CAT tool or translation management system, such as Phrase, XTM, memoQ server or Plunet, using its API, against the client's translation memory.
- Your rate cards are applied per client and language pair, with repetitions and fuzzy match bands, minimum fees and urgency surcharges as you define them.
- A quote draft is created in your template or your management system, and the PM reviews, adjusts and sends it.
- Simple, standard requests can be sent automatically if you choose, with anything unusual always routed to a PM.
Your prices and discount rules stay yours. The build applies them consistently and shows the PM the working.
What changes for your PMs
Quotes go out much sooner after the request arrives, often before the client has heard from anyone else. PMs spend their time on the quotes that need thought, such as complex formats, unusual languages or large projects. Pricing becomes consistent across the team. And accepted quotes can flow straight into project setup, because the files and requirements are already in the system.
Is this your agency?
- Every quote starts with downloading and unzipping attachments.
- PMs copy analysis figures from the CAT tool into a spreadsheet.
- Rates and discounts are applied from memory or a separate sheet.
- You lose jobs to faster quotes.
- Different PMs price similar jobs differently.