Billing questions are seen very frequently in the forums of popular translator portals or professional association sites. Often these concern whether translations should be billed by the word or agglomerated words (thousands or hundreds), lines, pages or other units in the source or target text. The answers to these questions reveal a wide range of approaches, which are often dictated by local convention, habit or simple fear. Often it is claimed that a particular method "does not make sense" because of compound word structures or other issues (which is frequently claimed for my language pair, German to English, for example). These claims are interesting, because other successful translators with the same language combinations often view things very differently. Who is right?
Before answering that question, let’s ask what sort of an answer can be trusted most. Would you stake your business on a “gut feeling” recommendation of another translator or translators when many others argue just as vehemently without evidence? Or would you feel more secure using a mysterious online calculator on an agency web site which purports to be based on a large body of text in the respective languages? Or would you prefer to see hard numbers based on your own translations that you have carried out in a number of different fields for various customers? Which approach do you think would give you the best basis for making your business decisions and protecting yourself against getting shortchanged in pricing?
I prefer to deal with real data from my own work. It not only reflects what I have been translating but also what I am likely to be translating in the next year or two. If I make a radical change I can quickly check and make sure that my pricing model is still “safe”.
To answer my own pricing questions, I created a spreadsheet in Microsoft Excel, which allows me to enter the actual data from individual projects and see what the relationship between target and source text pricing would be for words, lines and pages. You can have a copy of this spreadsheet to use yourself by downloading it here.
Because the true measure of price appropriateness is reflected in what one actually earns for the time spent working, a tracking calculator for the hourly earnings on each job was included in that spreadsheet to serve as a "control", but here I would like to focus on the relationship between different unit costing approaches and how to “adjust” prices to the units your prospect or customer is most comfortable with.
When I entered actual data from two types of translation work that I do (one specialized, the other a general category), I discovered some interesting things. I set up my spreadsheet to calculate the relative standard deviations (RSD) of the data, and what I found was that these were generally under 5%. What does this really mean? It means that if the data for your translation jobs is normally distributed, about 60% of the time the “actual” price relationship between two methods will fall in a “band” of plus or minus one RSD around the average, 95% of the time the actual relationship will be plus or minus two RSDs and 99% of the time it will be plus or minus 2.5 RSDs.
Still confused? Here’s a specific example from the downloadable spreadsheet:
If I want to earn € 0.15 per target word for a chemistry translation but the customer I am dealing with wants me to bill in source lines (allowing a “fixed price” to be calculated in advance – line rates are also common in Germany), I enter my desired rate in the little calculator table in cell B20. In cell E20 I see that I need to charge about € 1.27 per source line of 55 characters. How reliable is that figure? The relative standard deviation of the source line to target word ratio is just under 5% (some versions of the calculating spreadsheet in circulation do not include this figure, but it is accurate). This means that 99% of the time if you use this pricing, the “worst “real target word rate you achieve will be about 13.1 euro cents and the “best” you’ll do will be about 16.9 euro cents. If you work consistently with this pricing strategy your average earnings will be 15 euro cents per word. In many cases the bandwidth of variation will be much narrower than the example I have presented here. Compile your own data and see.
But what about the “exception” the fearful translator might say? One indicator of trouble for my language pair if I am using a source word pricing strategy might be an unusually low source word to source line ratio, which would indicate the likely presence of very long compound words in German. What do you do in a case like this? Raise your price if you feel like it. Use real data to show that this text really is different and must be priced differently from other work. Not everyone will agree with this idea, but you may have more success with it than you would expect. The important point here is that, by tracking actual data from your own work, you have a much clearer understanding of when rates may need adjusting.
When you examine your own data you will find that the actual variation in earnings between the calculation methods presented is small in most cases, at least for European languages. If this is not the case for your language, then you will have hard numbers to use in your quotation. Negotiations based on fact often work better, though all this can be greatly outweighed or offset by psychological factors.
By using the rate equivalence spreadsheet or creating your own similar tool, you can navigate the hazards of various quotation methods with greater confidence, quickly determining equivalent rates in the units expected by prospects and customers. This will ensure that you reach your average earning targets and achieve the same average hourly earnings as with your familiar unit pricing. You’ll know how much you have to raise your word price or how much margin you have to reduce it if you are asked to quote by source word instead of target word or vice versa. Now get down to business.