Weighted Word Count Calculator

A weighted word count (also called an effective or leveraged word count) is a raw word count adjusted for how much of the text a translator actually has to translate from scratch. When a translation memory already holds a translation for a segment, the linguist reviews rather than retranslates it, so that segment is charged at a reduced rate.

CAT (Computer-Assisted Translation) tools such as Trados, Phrase and memoQ analyse each file against the translation memory and report the words in each fuzzy-match band. Applying a discount grid to those bands gives the weighted total that quotes and purchase orders are actually based on.

The maths is straightforward. For each band, multiply the word count by the band’s weight (a percentage of the full new-word rate), then add up every band:

weighted words = Σ (words in band × band weight %)

A band weighted at 100% is charged in full; a band weighted at 30% is charged at just under a third. The total cost is then the weighted word count multiplied by your rate per word.

The match percentage measures how close a translation-memory segment is to the new source text. The closer the match, the less work is required, so the deeper the discount:

• Context matches (101%) and repetitions — identical segments, including surrounding context. Usually accepted with minimal review.
• 100% matches — an exact match from the memory; the linguist reviews rather than retranslates.
• High fuzzy (95–99%) — very close; only minor edits needed.
• Low and minimal fuzzy (50–84%) — progressively more reworking; charged closer to the full rate.
• No match / new words — translated from scratch and charged at the full rate.

Weighting produces a fairer price for clients and a transparent breakdown for linguists: repeat and previously translated content is not paid for twice, and everyone can see exactly how the total was reached. It also makes quotes comparable across projects, because the effective volume — not just the raw word count — drives the price.

The catch is the manual work: exporting the analysis, applying the grid in a spreadsheet and re-keying the totals into a quote. VectorLingo removes that step by reading the analysis directly from your CAT tool and applying your rate card automatically, from quote through to invoice.

A raw word count is simply the number of words in the source file. A weighted word countis that same file after the translation memory’s match analysis is applied — so repeated and previously translated content counts for less. The two numbers can differ enormously: a 10,000-word file with heavy repetition might have a weighted count of only 4,000 words.

Quoting on raw words alone ignores CAT-tool leverage and either overcharges the client or underpays the translator. Weighted words are the number that a translation quote calculator actually needs to price a job correctly.

Each CAT tool runs its own file analysis against the project’s translation memory and exports the match bands in a slightly different format — Trados as an XML analysis report, Phrase as a CSV export, and memoQ as a CSV with a two-row header. All three report the same underlying data: word counts per fuzzy-match band.

Agencies then apply their discount grid to that export — either by hand in a spreadsheet, or automatically, as this calculator does — to get the weighted total. For the workflow around managing that analysis inside each tool, see our guides to managing projects in Trados, Phrase and memoQ.