man driving vehicle near tree

Understanding your time/activity sheets: getting more out and analysing them to propose a simplification

At the end of the calendar year I sign off my time/activity sheets to say that they represent an accurate record of the work that I have done during the course of the year. There are two elements here: time physically clocked in, and the activity performed. Perhaps it is a good comparison to think of how London taxis charge: by time when stood still, and by distance covered when the wheels are turning. Translation is the wheels turning, the other jobs are the cab being stationary.

Freelancers probably consider it slightly differently. Time is more important where they provide added value services (e.g. revision, transcreation, terminology, database maintenance). In contrast activity/performance is more important when providing a basic translation service (whether charging by characters, words, lines, pages etc.). That said, the latter has a time element in terms of whether a certain job is worth the hassle. The taxi comparison also breaks down as added value services are also productive, and not just idling in traffic. In particularly, database maintenance ensures a leaner translation memory rather than an out and out larger quantity of translation units.

My annual breakdowns of timesheets since 2014 have shown that I spend around 2% of working time on translation on database-related activities including terminology. This can be in the form of dedicated terminology hours, alignments or TM maintenance.

How has my time and activity logging changed?

I physically sit in the banking supervision department, but also have responsibility for outsourcing for other supervisory departments. After a couple of years’ experience of outsourcing, it became apparent that small jobs cost more in terms of the procurement process. In terms of time, they often take longer than had I just got on and translated them. I used historic timesheets to show the hidden cost of procurement instead of me logging to a different cost centre. Doing so reduced the number of items translated externally. The time saving for the procurement of such translations in turn freed up time to handle short pieces. By doing so, I proved that some translations, e.g. press releases were “too small to outsource”.

A further simplification was possible once I had multiple years’ of data. Initially, activity logging for translation was broken down into three sub-activities (translation/revision/other) in an extra level of detail in the activity booking. In some jobs this resulted in triple the number of bookings. After a number of years I proved that the level of the individual sub-activities remained pretty constant. Broken down by different sub-activities I consistently posted approx. 80% for translation, 17% for revision, and 2-3% for database maintenance and terminology work each year. The consistent level was instrumental in reducing three booking codes to one and reducing the number of activity logging bookings by 75%. It was also possible to bundle jobs that were otherwise realistically almost “too small to log”.

But what happens if the goalposts move?

When my employer changed website CMS, I developed a way to handle translations of pages through the code view. This meant that jobs weren’t broken down into translation, revision and then layout, in my web editor guise. However, when the site changed editing system with the CMS, this did affect my workflow. Fortunately, I was able to then book the layout/CMS work as external communications (web editing).

This also meant that I could see how the change affected my activity. The extra layouting step’s impact has diluted my percentage of translation bookings out of total bookings. The consequence has also been an increase in overtime hours. I’ve tried numerous work-arounds, but they haven’t been effective handling Unicode code bindings. I am able to show the effects of the workflow change impact my activity logging.

Does terminology work get marginalised?

A mere 2% of translation activity being specifically devoted to terminology is a very low amount. I concede that this is one of the shortcomings of being a SPLSU. Considered over a working year, (for sake of argument 220 working days), it accounts for only one working week. However, this amount doesn’t fully reflect any ad hoc terminology work on the fly during a translation job. I add a lot of barebones items (i.e. source and target terms, reliability score, supervisory area, and whether a term is inclusive).

Termbase items quickly add up, but revision and completion (e.g. phraseology, synonyms, context, definitions) only happen after delivering the translation. I would love more time for terminology, but am realistic that added detail is something that only I benefit from. Customers are naturally happy that I use the correct terminology, but have little interest in the research behind it. Terminology work can become hard to rationalise. For them translation is like a swan gliding effortlessly across a lake – they don’t see below the waterline.

Perfect efficiency: exploding the myth

When I deliver my time and activity logging, total translation hours do not tie up 1:1 with my total working hours. It would be a myth/illusion if they did! Time and activity logging also reflects non-translation related activities (e.g. attending courses, working groups, internal/external communications tasks). A translator would need to work up to a 10 hour day to have 8 hours of pure translation productivity. This confirms that your mileage may vary even if benchmarks about daily translation output apply (see for example this blogpost).

In my early days at my current job, a common question was about how much translation was possible each day. I took a simplistic approach: of sustained output of 2,000 words a day. With an 8 hour working day, that meant the output was a steady 250 words an hour, right? Jein. 250 words an hour might be the average, but output could fluctuate between 100 and 800 words per hour. Text complexity, TM maturity and subject familiarity all affect output rates.

As a freelancer, customers and agencies set the deadlines and I worked long hours. I became all too well aware of burst and steady translation output rates. I learned how to work on extensive multi-day/week/month projects and how to set deadlines. The secret was to have enough of a buffer to deliver slightly ahead of schedule. Lengthy projects were vastly preferable to very short pieces (“snippets”). Agency “ticking clock deadlines” were never a favourite. For long jobs there was also due warning about their arrival.

Managing workplace wellbeing: handling peaks and troughs

Time and activity logging provides an added dimension of understanding to recognise peaks and troughs. Having experienced fast-moving black swan events, I now know what effect they have in terms of shockwaves. Typical black swan events include insolvency of a bank, national lockdowns and events impacting financial market stability. Years of bookings also show when to take extended periods of annual leave. How? The type of projects I have to handle provide tell-tale signs of how busy the office is. Past time and activity logging shows that August is usually filled with low urgency projects. This makes it easier to take annual leave. Time logging shows that it is easier to leave the office earlier during the summer. However, it also confirms that I do not often adopt the POETS approach.

Visited 6 times, 1 visit(s) today