Translation is like any job: you have to start somewhere. And the day you actually start you will probably find you do not have much of a clue of where you really stand. I was very confident on my language skills, had done a fair amount of Internet research and even had the leisure of knowing that I had another job to pay the bills no matter what. However, those first jobs were quite daunting, full of ad hoc decisions and lots of unknowns.
I opted to take this stage as a sort of internship, where I made little money (certainly a lot less than I had hoped for) but got to assess where I stood, map a route for what lay ahead and generally position myself with relation to a market that I knew little about and which knew even less about me.
My research showed that 0.20 dollars per word was a good rate and that 0.03 dollars per word was lousy. I chose a middle road, and when agencies got back to me asking, "How about half that?" I had to make tough choices. I intended to make a career in translation, and I decided that I was only as good as the jobs I managed to get. It would not help me to stay at home doing nothing, just waiting for a well-paid project. At the beginning, these were mostly cheap jobs, which was bad. But they brought me a bunch of interesting things beyond money.
When you are starting out, in translation as in any other freelance field, you need to make a name for yourself. On Proz, that is largely measured in Willingness to Work Again (WWA) points. My first jobs brought me a few of those: people who said I had done well and they would be happy to give me more work in the future.
Further, these jobs allowed me to assess certain market characteristics. I soon realized that there is a lot of demand for German-English technical translations, particularly cheap ones. I could get lots of work down that alley... but would not be able to make much out of it: I found it boring, was unproductive since I had to look up lots of concepts and probably did not deliver the kind of quality I would like to be known for.
I realized that, although I am fully bilingual in English and Spanish, it was not easy to convince potential clients of that. For some reason, many find where you were born more important than where you got your education, for example. I spent the most important part of my formative life in Britain - three years of secondary school and a degree in Philosophy, Politics and Economics from Oxford - but for many I am still only a Spanish native, just like someone who got a degree in translation from Madrid or Buenos Aires. Mmm... Again, I was only as good as the work I could get.
Most importanly for the future, I realized that I would need to assess any potential project in terms of my own productivity rather than rates. I am a decent German translator, and I certainly have loads of experience at it, but I cannot translate much more than 500 words per hour from German. When I work with English, Spanish or French, my output is often close to 1,000 words per hour. So surely I would have to charge clients almost twice as much to translate from German as I could afford to charge in other languages.
The same went for fields of specialization: I was particularly good at the humanities, marketing, finance and business-related translations, and the day I got a choice I should probably stick to those.
That is the thing about internships: you usually do not make much money, but if you take them seriously you learn a lot.