Bootstrapping a Language Product Company Using Services from London, Then Taking it Public and Scaling It to $450M: SDL CEO Mark Lancaster (Part 3)

Sramana: Can you give us an example of the types of products that would be labeled as language technology products?

Mark Lancaster: There are a lot of products that fit in this space. We started out by building products that would make translators more productive as they translate. We have gone on to create many more enterprise translation technologies.

Sramana: What does it entail to take a software package and convert it into a different language? Aside from pure natural language translation, what other nuances and intricacies are involved?

Mark Lancaster: It’s a very big topic. What people like to think about is automatic translation. That is machine translation and we have been investing in that technology for about 15 years. That is a very specialized area. There are three companies in the world that have the best scientists working in this space: ourselves, Google, and Microsoft. We all invest in statistics machine translation. We own FreeTranslation.com, for example, but we don’t make money from that.

We use our automated translation technology in all of our language products. When a translator is translating, they can get a rough translation from a machine and then, review it to make it perfect. It is difficult to get perfect translation in a machine because of all the linguistic challenges. However, it is very valuable in the translation space. I think people are fairly unimpressed with that kind of stuff. The Google and Microsoft technologies are very good, but i’ts just a message within the ecosystem. It plugs in and translators can work with the automated translations in their world. It’s not really that useful. People use Google and FreeTranslator when they want a quick translation. That will give a rough translation with a few words wrong. Users look at that and think the translation is not that good when in fact it can be very effective. We translate billions of words and provide the technology to 70 of the top 100 brands in the world. Any large company will be using our technology or services in some shape or form.

There is also technology to help translators work more effectively. When you need multiple formats, we can allow them to work with all of those. They can save everything they have translated, apply automated translations, etc. That technology plugs into very big workflows. If you are someone like HP or Microsoft, then you are translating upwards of 60 languages and millions of words per day. Does not matter if that is software text that lives in a product or if it’s content on a website or user manual, it’s content that has to be translated. There is an awful lot of content and it has to go through the appropriate workflow. If it’s high quality marketing, it goes through human translators. If you are reviewing website feedback, it can go through an automated format. The large system integrators have predefined workflows and we plug into probably 10 or 15 different content feeds which automatically get dispersed around the world.

This segment is part 3 in the series : Bootstrapping a Language Product Company Using Services from London, Then Taking it Public and Scaling It to $450M: SDL CEO Mark Lancaster
1 2 3 4 5 6 7