Let me set the scene. I recently had a three way conversation with a very valued and long established customer and a senior executive from a machine translation technology company. The meeting was called to discuss the possibility of using machine translation to help bring down the costs of localisation (currently a $4m annual spend for my customer) and was prompted by a senior executive at the company who had heard at a recent conference (as one does!) of the enormous and instant savings that can be made in this area. My customer contact was simply told to “make it so”.

After 45 or so minutes of a concerted and collaborative effort on the part of the MT company representative and myself to convince the customer that her material was neither suitable or large enough to warrant even the investment of a pilot project (I am not joking here!) she decided to go ahead and spend the $20,000 it would cost to do such a pilot anyway.

Such was the power of the original suggestion that a usually controlled and circumspect professional was willing to disregard the people who she knew would know better and try to reach for the fantasy world of “free” high quality translation. Wishful thinking in the extreme.

Maybe you have been asked to investigate machine translation as a potential cost saving vehicle for your company or maybe you’ve seen enough of the web based free apps (here’s one!) that accurately translate “can you direct me to the nearest train station”  in almost any language, to believe that it’s not a leap from that to the full blown free translation of your UI, Help and Documentation. Or maybe you’ve just seen enough evolution in technology in the last five years to make you believe that this must be possible soon, if not now.

Whatever you’re reasoning, here are the “undeniable truths” that you’ll have to get your head round before you can even begin to think about putting any plan into action:

1. There are companies far, far bigger than yours who have invested millions of dollars into machine translation over the last ten years who still will admit to being a ways from having a viable solution (certainly in a general sense). Microsoft is one (if you’re looking for one!).
2. Any machine translation system needs a whole lot of coaching and coaxing to even begin to generate results. By this I mean it needs hundred of thousands of paired segments (or matches) so it can learn what to do and what not to do. You need to pay for this coaching, most likely by the hour. You’ll need to do this for every language you want to make savings on.
3. A typical “industry standard” minimum source word count needed to  justify any savings is 1 million words.
4. Some language pairings are more “advanced” than others. This is because some languages (Asian languages like Simplified Chinese for example) are simpler in construct (lack of gender and cases). German, for example, is much more difficult to program for. On an interesting side note, Arabic to English is a good pairing, following the multi million investment on behalf of the US government as part of their fight against terrorism.
5. Once a machine translation system generates a translation, it needs to be post edited (by a human!). Basically, the task of the post-editor is to edit, modify and/or correct pre-translated text that has been processed by a machine translation system. More on post translation can be found here if you’re interested. Suffice to say that the “better” translators tend to avoid post editing duties, with some preferring to work with the original source rather than trying to understand the machine output when trying to get a sense for what is being said in the text.
6. In all likelihood the adoption of any machine translation system will result in restrictions being placed on your source material. Machine translation systems need simple constructs (or at least as simple as possible) to generate results. You need to know that the quality of your source material will be affected accordingly.
7. On a related point, the subject matter needs to be “straightforward”. If your material is in any way complex you’re probably wasting your time.
8. On a positive note (I need to appear balanced on this subject!!) some companies have found that they benefited from machine translations in applications where the material will not be seen (or seen very rarely) by customers. Knowledgebase systems for example (where you have a company wide HR database or something similar (that is for internal company use only)) might lend themselves to a possible benefit.

In summary, my advice to anyone who is considering a machine translation trial, is to talk to people who have benefited from it. Be suspicious. Assume what you’re being sold is being talked up. Get references, and be sure to check that their material is similar or has a similar application, to yours.

And even then you should realise that it’s a long, hard and expensive road to when those machines start spitting out your  material in multiple languages.

At least that’s how it looks from where I’m sitting…….

I was coming out of a hotel last week when I came across this scene in the parking lot. Just look at it – how did they park like that?! Now you may think, “well, the lot was full and that’s just the way it goes”.

But actually, the parking lot was half-full [for you pessimists out there, it was half-empty]!

It’s a picture that doesn’t need words. So how did it come to pass that no one was willing to budge an inch? I’ve no idea, but it does remind me of negotiating; well, how not to, I guess.

To me, negotiation in this business is all about building relationships. Unlike buying a car, when we negotiate contracts for our localization services, it rarely happens in isolation; we will work with Company A or Company B, etc. on a continuous basis - and probably with the same people with whom we negotiated in the first place. So if we effect bad relationships with these people, it can come back to haunt us at a later time.

Essentially, negotiating will involve a series of events and therefore we will want those events to be good ones for both sides. Now, take the guys in the cars; this hotel they parked in is a very popular one, so the chances are high that they will be back again sometime in the future.

Will they have learnt from this visit? Hopefully.

A series of good events/negotiations allows us to build relationships with companies we do business with; part of the negotiation process involves us [and them] looking at the bigger picture; can we come to an agreement that will benefit the business of both sides? And that does involve now and again, some creativity and flexibility. Which is fine, we love flexibility!

In this case, all it required was for one driver to figure out how their actions would affect the others around them. But I would suspect this wasn’t up for negotiation…

Now, talking of flexibility, it’s time to go squeeze into my car

So you’ve decided. Those pieces of Marketing collateral just have to be translated for the German, Japanese and Chinese markets. Just send it out to us and you’ll hear back sometime yesterday when it’s finished, right? Maybe not – so here’s your quick 4-point checklist for eternal enlightenment.

1. Do You Really Know What You Want?

Ok, so you have the collateral. But have you checked with any in-country people [your Distributor, your Sales office] that this piece of collateral is locale-ready? Maybe they’ll want to re-write parts of it to make it more specific and maybe some parts just don’t wash locally [those stories about Coca-cola taglines in China are still legends].

Are there any screen shots in the piece that require either part of the relevant software to be translated, or the shot to be “faked” with some clever artwork? And do you really, really want that Flash Demo localized?

2. Which Part Of The Content Is Most Important To You?

When you’re writing the content for a piece of collateral you may be able to pick some terms or words that are key to your marketing messages. You’ll use these over and over again in a campaign or over several campaigns. Once you’ve indentified these, we can use them to create a “Glossary” of key terms that can be re-used in subsequent campaigns, saving you on costs. And also they will be a valuable starting point for any new language translations you decide to add along the way.

It's Not As Cold As It Seems, Go Ahead Try It

3. Who’s Approving Who?

Have you considered how the approval process will work – if we translate the marketing collateral, who will approve the final translated content – will it be an in-country Distributor, a Sales person on the ground or will you have someone in corporate HQ give it the thumbs up? Who you choose [if anyone] to approve final content, will affect Point 4……

4. How Much Time Have you Factored In?

So you want the job completed yesterday….Ok, well, here’s a rough example of how long “yesterday” will actually take. Say you have 1,000 words of copy, one or two graphics/screenshots in the piece and an in-country Partner who will sign off that the translated piece is good for the local market.

1,000 words translation = 1 day

Graphics manipulation = 1/2 day

Review [by Partner] = 1 day [Just think time difference!]

Implement review feedback and approve final copy = 1 day

Total:  3.5 days

But of course being a good project manager as well as top notch Marketing person, you’ve padded out the schedule and told everyone it will be 4 to 5 days turnaround time!

This is just a quick overview of what is entailed [no mention of file formats, source files, etc], but it should give you Marketing folks some baseline for judging the time and steps involved in the translation process. Now then, about the costs…….

Having recently trawled through the sites and other promotion material of some of our competitors (as one does) I was struck by the fixation so many of them seem to have with the concept of “biggest”.

One company is the “world’s biggest independent localisation service provider” ; another boasts of the fact that they are “the largest independently owned localisation company on the west coast”. Guess the company claiming the first title doesn’t reside on the West Coast then.

Does it really matter who is the biggest in a particular “class” when you’re tiny compared to other players? Surely, if size is really important, you don’t want to boast of being the tallest dwarf.

The truth is that size really doesn’t matter as much as some people (the tallest dwarves) might think. Most savvy content provider project managers just need to know that you have the capacity and knowledge to get their stuff done. After all the bigger a provider is, the greater the chances are that they will have at least a couple of larger accounts that will inevitably draw the most attention and the best resources. Would you really want to place your $100k of annual business with a company that has you languishing in the lower quartile profile of their customer portfolio? Will their size fit your needs then? Chances are that the service you receive will be commensurate with where your business falls on their scale.

In almost any service industry you could name, the smaller, “boutique” type, bespoke provider wins on product and service every time. It has to, almost by definition.  Marry that with an acceptable cost and you’ll never go “large” again.

Trust me

At least that’s how it looks from where I’m sitting…….

What, where, how and also when… all these questions have their answers only clicks away using search engines but could we get answers with less clicks?

When the web was young… it is still young but I mean younger… there were no search pages but directories of web sites like our old good Yahoo. It was good if you knew in which category you had to look, a bit like (or exactly like) the yellow pages (or whatever colour they might be in your country). It was not automatically updated but you had to submit your web site and wait for Yahoo to check it. Apart from few specialised directories, this model of “search engine” is not used anymore.

Then came Google… well not exactly, Altavista was first, it was the first engine to send a “crawler”, a little virtual robot jumping from website to website via links to index all the pages… Google started in 1998, four years later. Google success is in part due to Altavista failure, after being bought by Compaq, it was converted to a portal, that was the killer. Now altavista is owned by Yahoo, ironic, no?

So from 1999 to today (I am writing this in 2009, in case you are wondering) Google kept its number one ranking in the hall of fame of search engines… but in 10 years or so, have the results improved? You can say that you have more results but are they useful? After you have typed in your request and clicked “Google Search”, how many pages will you open before finding what you want? A lot, I know!

So NOW is when the future of search is starting, where you get what you are looking for, not a whole bunch of results to be sorted to find what you are looking for.

The future will hopefully look like WolframAlpha (WA). You have to know what you are looking for before using it and, as it is still at a beta stage, there are lots of things it can’t help you with. WA will compute the results and will give you what you are looking for and not links to pages where you might find what you are looking for.

Results for "ireland"

The simplest test is to try a country, for example “Ireland”. If you search for “Ireland” in google you would get 248,000,000 results, a lot to go through. The same search in WA will show you maps, will tell you how small it is (70,280 km²), how many people enjoyed its climate, even the GDP per capita! For all these details you can, obviously, know the sources to make sure how accurate and old they are.

It handles also very well any mathematical requests, from a simple calculation to more advanced formulas. It likes dates, do you want to know which day of the week you are born and how long it has been in days? Easy!

Are you on a diet? Yes, this is still related to the same subject, bear with me… Imagine your breakfast was 1 orange, 2 apples, 1 glass of milk and a slice of bread… how many calories is that? And fat? Well WA will tell you exactly: 414 calories and 4 g of fat.

Into music? How does A F D C# B A D sounds like? Not bad actually.

There are plenty of examples, per topics, on the web site for you to experiment.

It is not perfect and can’t replace Google just yet but can WA replace Google one day? It would take time and Google might try to beat them, they are still behind with their attempt but I am sure this is something they are working on to keep the #1.

One other big problem is the name, Wolfram Alpha is not as catchy as Google .

If the term “characters” is lost in translation for you, then think “sociable, genial, funny” and “larger than life”.

Actually, 140 characters in a bar sounds like a bar anywhere downtown Dublin on a Friday evening [even in these recessionary times]. I guess they asked for a Guinness and a Guinness Follower [chasers are so 90s].

Ah yes, that looks like 140 characters

It’s amazing what messages you can pack into one tweet, even when 140 characters seems a small amount. I “watched” a whole Boston Celtics playoff game a few weeks back on Twitter; by the end of it I was a nervous wreck, but not only did I get the score updated constantly [they won that battle, if not the war] I was completely immersed in the game as if I was in the TD Banknorth Center.

But anyway, what about other uses – say, using Twitter to order beers – maybe it’s happening in your area already? I was in a bar in Estonia a few years ago where you could order a beer via the telephone on the table – what about @estoniabeerdrinker – 2 pints for table 5 please?

If you have kids, how about tweeting them [come on, they all have mobiles or laptops!]; @MumAndDadNeedPeace – Time for bed, finish your homework. NOW!

[Just make sure to turn the mobile or laptop off after 1am to avoid being used as a taxi service].

I can see endless uses and of course, despite the fact that it seems like we’re not communicating in the way we knew, it looks like Twitter has us all talking. Constantly.

Now then it’s time to find where those characters went [for research purposes, of course]!

Speak English? Then Book This Train

2 or 3 days a week, I commute to the office by train [a great way to travel, believe me]. Armed with an online timetable I can pick several times to catch the train, or book tickets. Easy, eh? Extremely. But then again, I speak English.

As I work in the translation business I was curious – what about all the tourists Ireland wants to attract in from other European countries? Which language on www.irishrail.ie would they pick? Um, well check it out and you tell me.

OK then, what about all those thousands of business people and tourists who use the Heathrow and Gatwick Express trains in the UK every year; people who have limited English and who want to book online in advance? No problem, they can go to www.heathrowexpress.com or www.gatwickexpress.com and pick any language they like. As long as its English….

And then there’s the hundreds of thousands of people who use the French train system every year www.sncf.com – they’re stuck with….no, maybe you didn’t guess it! French, English and German!

For a small investment, maybe some of these rail companies could boost their sales by making it easier for people all over the world to enjoy some track time? Qui sait? Mais, vive la France!

Iota LS typically stay out of the localisation tools debate. Sure, we have an opinion (on everything!), but when it comes to choice of tools we tend to work with what our customers prefer; after all they pay the bill.

The uproar caused by the introduction of what is essentially the latest version of Trados however has caused enough concern among the people who work for us, that it seems appropriate to at least ask the question if the benefits outweigh the cost in this case. Even if you leave aside the discussion about the relative merits of its features, the fact that it will be seemingly impossible to upgrade to the new version AND still maintain a foothold in the old technology will cause untold problems for the translation footsoldiers who have to cater for the widest spectrum of tools and versions in order to service their customers. SDL have moved to quell the discontent but stuff still doesn’t smell right.

For more in-depth material, go to the TW_users group on Yahoo Groups. You’ll find one or two unhappy bunnies there.

I’d imagine that publishers and content providers need to think long and hard before taking this particular plunge. Many fine translators will not follow and some loss of linguistic continuity might well be expected. And as I understand it, there literally is no going back on this one.

It seems to me that it’s always clever to at least talk to the people doing the work before deciding how it should be done!

At least that’s how it looks from where I’m sitting…….

Addendum: July 14th 2009:

A quick trawl through the web today and a poke to many of the translation contacts I have worked with for many years revealed no progress in the situation. I’d imagine that SDL will try to ride this one out, hoping that as the top content providers succumb to the “lure” of the 2009 product, the ducks that line up behind them (the MLVs and in turn SLVs and freelance translators) will follow suit and any unhappiness will just die down over time; as befits the typical modus operandi of a “dominant” player in most markets.

What they should do, of course, is to actually listen to their users, rethink the whole situation, call a time out and revise their plans and products.

For those of you that feel compelled, you can find about a million discussions on the issues though a simple Google search. Here are a couple; one here and another here that I read with some interest.

For Christmas 2008 I received the wonderful gift of a Garmin Forerunner 405, the ultimate tool in exercise timekeeping.

Gone would be the days of my old stop watch, and my black notebook filled with entries only I could understand, outlining date, time of run and distance (approximated of course). Now I could relax, knowing that everything I could possibly want to know about my run would be recorded, on a website. Speed, distance covered, a map of my run, pace, heartrate, elevation and other attributes were now magically uploaded to my own space on the web for me to pour over. I was in detail heaven.

Except that’s not how it went. What had gone from a 7 mile run in 1hr 4mins (on a good day!) was now broken into mile splits and I was set against the virtual partner whom I became determined to “beat down” at every opportunity.  Due to what I am convinced is a slight case of OCD on my part, my once enjoyable runs (that’s how they look when I remember them now!) have become an hour filled with “check the watch…how far have I run?…how fast am I going?….what do I need to do to beat my best?….what’s my heart rate?”.

These devices are amazing. Get one if you’re so inclined and able. But don’t let the compulsion for detail take the enjoyment out of the activity.