Some success and a little reflection…

We dropped a press release yesterday about a couple of our recent successes since launching eleven weeks ago. It’s great to finally be able to share a story with people outside the company that shows what we’re doing here is actually for real.

I was talking with some folks the other day and was remarking that of all the industries/spaces I thought I’d end up working (playing) in, the localization/translation space never would have come up, even 2 years ago. The post-bubble years certainly consisted of a lot of fumbling in the dark, trying to figure out where to go next – through some luck and a lot of serendipitous occurrences I ended up here.

As a technologist it’s a really interesting space to be in right now. It’s a neat mix of tools and technologies I hadn’t really ever imagined existing, combined with a bunch of processes I’m surprised there aren’t technical solutions for. It also very much feels like an industry on the verge of a total technical transformation (and all indicators are that the leading edge is already well on their way over that hump).

It’s certainly looking like we’re in for an interesting ride, both “we” in the Clay Tablet sense, and “We” in the industry sense. Thanks for reading – and I look forward to carrying on the conversation as we go forward.

– Ryan

Technorati: , , , ,

Using flags as language selectors – and the real problem they may be hiding.

Patrick Hall over at Blogamundo put up a good post the other day about why it’s a bad idea to use flags as language indicators (“Flags are a dumb way to represent languages” – Nov. 21/05). I agree with his points whole heartedly but I also think the issue also speaks to the confusion between the two types of translated/multilingual websites.

Certainly the argument against flags holds up when you look at countries like Canada or Belgium. Both are officially bilingual countries – so when you show someone a Canadian flag as a language option they aren’t sure whether to expect an English site, French site or a completely separate site that caters to both languages.

Herein lies the problem hiding under the surface – when starting out to create a “multilingual” site you need to consider very carefully the structure of its multilingual elements. Should you be creating a single website that is translated into multiple languages, or is multiple sites, each localized to their geographic region what you really need?

Take the translation industry for example. Europe views translation very differently from how North Americans view it. North America, with a few exceptions (i.e. Quebec), is very much in the midst of moving from translation as a novelty to translation as a business decision that everyone should be considering. In Europe translation just is – It’s just a part of doing business in Europe.

So, keeping in mind that geography could change how your message is received by site visitors, a company needs to weigh whether that perception shift is great enough to justify splitting off a site to specifically address that need.

At the end of the day you essentially have two high-level paths you can go down:

  1. “Multilingual” site – one source language translated into many other languages

    PROS: Simpler architecture, only one content source to manage

    CONS: Only one message is presented to different markets, changes can be costly when you have to update all languages each time

  2. “Regional” subsites – a series of sites that are delineated by region (Continent, Country, State/Province, etc.)

    PROS: Gives you messaging flexibility, each site would need to support fewer languages so changes wouldn’t have as big a translation impact. Gives a very inclusionary message to regions who you create a sub-site for

    CONS: More websites to juggle/manage, similar changes to core information may require updates to multiple sites if content can’t be shared


It’s a tough decision to make when starting to think of building a multilingual site – but it’s very important to consider this decision. At the outset the Mutlilingual approach will likely look far more appealing financially but if, in the long run, you’re fairly sure you’ll end up with regional sites it’s a good idea to consider buying a tool that will let you grow into that stage. From an application perspective the applications that support these types of websites are very different, and many CMS’s just weren’t designed to handle multiple sites, especially in multiple sites in multiple languages.

So the question comes back to flags. If you do go with regional sites are Flags a good way to get people to select their language? Purely language? No. Regionally? Perhaps.

The challenge with using graphical images to differentiate items is there’s no logical way to process them. You could line up all of the flags ordered by country name, but that’s a big leap for a user to make. It also potentially alienates users.

Quite often I’ll go to a site and there will be a US flag and a French flag. What message does that send to a Canadian visitor (Anglo or Francophone)? I would suggest use flags only in the following conditions:

  • If a regional site services more than one country either abandon flags altogether or, show all the flags that are covered by that region – i.e. don’t just put a French flag up if that’s where your office is if their mandate is to do business across Europe. Bad: You alienate users from other countries, Worse: They think you don’t service their market and find someone who does.
  • Use flags in a graphical context (i.e. a world map) so users can easily locate their region and then flag.

Either way, you must provide some textual way of selecting a regional site (i.e. a drop-down). Make sure the drop-down is ordered by the alphabetical order of the native language version of the country name (some countries names may change first letter from English to their native tongue and English speakers won’t be looking for that name).

Technorati: , , , , ,

Multilingual Blogging – what’s out there?

Well I figured I’d do some searching to see if anyone had “solved” this challenge, or at the very least come up with a good workaround.

Almost immediately (i.e. first hit on Google) I thought I’d stumbled on someone with the solution. Proclaimed as the “first multilingual blog” in a press release, I immediately thought “cool”. Unfortunately when you click through to www.multilingual-search.com what you encounter is a bunch of English posts with language options of…

Yeah, English. :(

Thinking I’d maybe just hit them in a bad streak of English only content I scrolled down a bit to see other posts. About halfway down I encountered a post that was tagged as being in both English and Bulgarian. The “holy grail” was right there! I clicked on the [Bul] and was promptly greeted with “Page not found”. I don’t speak any Bulgarian – but I’m pretty sure that wasn’t the content I was looking for.

(Update: Andy from Multilingual Search let me know that the Bulgarian translation issue was a bug and it’s fixed now – which I’ve confirmed.)

I came across many other sites that were about globalization translation and multilingual blogging, but, like my site, they were all in English.

A glimmer of hope this morning – I just came across a plug-in called “Polyglot” for WordPress. It talks mostly about “localizing” your blog and seems to be quite focused on the wrapper (i.e. non-content text) of the blog. After navigating through a few more layers of blogs & other sites I encountered “Multilingual“, a plug-in for WordPress that allows you to post in multiple languages. The only limitation that I can gleam from the website is that it’s geared more for those who know two languages and want to blog in both – i.e.. do their own translation. (I’m going to drop these guys an e-mail and see if that is the case.)

I also encountered a few plug-ins for Machine Translation – but, yeah, not wanting to go there yet…

What I still haven’t found though is any sort of translation workflow solutions. Something that spits out your content to someone who’s willing to translate for you and then makes that translation available – I’ve got an idea on that, and depending on how the next 24 hours shape up I’ll try and get it on the blog tomorrow – otherwise I’ll crack it off this weekend and post Monday.

– Ryan

Technorati: , , ,

Major site content revisions – when do you translate?

Coinciding with the article in CSN this month we also put a new revision of our site up. It’s a fairly major overhaul, focusing a little less on our content management system and more on our translation workflow gateway – which connects to content authoring and management systems and ties them directly into a TSP/LSP’s workflow.

Interestingly enough this revision brought about some interesting questions/considerations when it comes to a small business trying to run and multi-lingual website – the biggest being, “when do you send the content out for translation?”.

Based on feedback during sales and partnership efforts we recognized that some of our messaging probably needed to change – what resulted was a fairly large change in how our site is structured, a generous revision to existing content and the addition of several new pages. Changing our source language (English) is no big deal (ah the wonders of a CMS), and with our Gateway getting the content translated is just as easy (just click “Send for Translation”). The challenge though, is that clicking “Send for Translation” has a fairly serious financial impact (we’ve made it easy, not free :) ) – especially for messaging that may or may not be quite right yet.

So the question comes back to – when making a large change in the content of your site when is the right time to send content out for translation?

Of course the ideal is to finish your source content, send it out for translation and launch all languages at the same time. For more established companies, who’s messages and marketing components are more established, and less likely to change this is fairly easy to accomplish. But for a small business, or startup, who may be in a much more turbulent space, it’s very likely that their messaging will initially change frequently for a while and over time settle down. Constantly sending out revised content to translators can quickly add up – and for some companies the costs could become crippling.

From what I see you have three choices:

1. “Turn off” the other languages until you’re ready to get them translated
2. Retain the old versions of the translated sites and update your source. Translate when you’re confident the content is stable for a bit.
3. Bite the bullet and just keep sending those revisions out.

For small “tactical” changes in content I believe option #3 is really the best way to go – a website should be dynamic and elements of it should change from time to time. It’s important that when companies consider moving to a multilingual website they also budget for the iterative changes that will incur some translation work.

I also think #3 is the way to go if your existing content has become completely out of date to the point of being incorrect or misleading. Happy customers or prospects aside there could be serious liability risks for your organization in some cases.

For now we’ve chosen to go with Option #2 for our site. While we’ve made major revisions to the site, the fundamental messaging from the old site is still accurate. Our products haven’t changed dramatically, just the way we’re talking about them.

What are the risks? Well, If we’ve really nailed the messaging this time around on the English site we run the risk of missing an opportunity because a non-English visitor hasn’t seen the improved messaging. Other than that I don’t expect we’re risking much else. Information is still available on our site in 5 additional languages (English, German, Spanish, Japanese, Mandarin), and those visitors are getting accurate information about our products. Our goal right now, provided the messaging works, is to push the new site out to other languages within the next few weeks – it’s still very important to us to maintain a consistent multilingual presence.

As for #1, personally I think it’s only an option in very few circumstances:

  • Your organization just can’t afford to translate your site anymore and content is getting to the point of being inaccurate or misleading.
  • Your organization is moving away from operating/making an effort in the regions you hoped to service through that language.

In either case, you’re company is getting to a point where it’s hand is somewhat forced. In the first circumstance it’s understandable that an organization may come to the point where it’s just not fiscally feasible to maintain a multilingual site – unfortunately at that point there are probably much bigger things to worry about as well. In this situation you’re likely also experiencing circumstance #2 – when things get tight an organization needs to focus, it’s likely that international sales efforts will be reevaluated.

If you’re purely in circumstance #2 it does make sense to stop spending money translating for a market you’re not interested in pursuing at the current time. The key though is “Current time”. Remember that you may decide to turn your focus back on that region again. My recommendation if you find option #1 as your only choice is to take the last of your translation budget and make a graceful exit from that language. Consider condensing your basic pitch or product info to a single page or maybe just a page with a message that indicates why that language is no longer being updated.

At the end of the day you just don’t want to come across as not caring – you never know when opportunity may arise in that region again – the last thing you want them to remember is how you just vanished the last time.

Technorati: , , ,