Cost calculation – per page or per word

Indeed, the source text was typewritten, all words covered the same amount of space on each page, and had letters of equal size.

Now, source texts for translation are completely different. They are created by computers using a variety of text and graphics editors. Thus, calculation per page has become meaningless. A ‘page’ is considered as a very conventional concept. The page of a PDF file may include as few as 10 words or as many as 2000 words and this is a normal situation.

Therefore, most of the translation community started to use word counts as a basis in their work at the end of the last millennium. This shift has also led to the emergence of special translation software which allows the translator to store previously translated texts in a database and automatically substitute them into the translation if this segment was previously translated (as a rule, a sentence constitutes a segment).

Let us consider the translation of a contract. Basically, each company uses rather standard contracts; the subject of the contract, special terms, and specifications may differ. For example, 20% of the contract’s text may be changed, whereas the other 80% of the text, as a rule, remains unchanged. Then it’s just a matter of mathematics. What would happen if the company concludes 4 contracts per day and all of them require translation? What happens if we count the number of pages? The translator will count the number of standard pages in each contract and multiply them by the number of contracts.

Then suppose we apply the Translation Memory technology and calculation based on word count. In this case, the translation of one contract is required. As for the rest of the contracts, the program will count only new sentences that were not present in the first contract. This method of calculation based on word count makes it possible to greatly reduce the cost of translation.

Also, the concept of Fuzzy Matches may be applied if the cost calculation is based on the number of words (used in the segments where sentences coincide by 80% or more). This is common with the titles of contracts, where only the contractors are different.

The translation program also takes this into account and recognizes them not as new words, but as so-called Fuzzy Matches. Translation of Fuzzy Matches is not paid at a 100% rate, but at a rate of 30 to 90% according to the degree of coincidence. This is also very convenient for the translation of standard texts.

Calculations made on the basis of Translation Memory technology without the use of calculation per page are fully justified for translations of technical texts, financial and legal texts, medical documentation, model reports, protocols, conclusions, and so on.

Translation Memory technology does not provide such tremendous advantages for the translation of literary texts and marketing materials. However, calculation based on word count makes it possible to better assess the cost of work compared to calculation based on the number of standard pages even for these types of texts.

Consequently, we and most of the translation community recommend calculating the cost of translation using special software. It will result in a higher level of transparency and accuracy with respect to the cost calculation.