View demo

Why the current LSP model is obsolete (and how to fix it)

Henrik Lottrup

Author: Henrik Lottrup
CEO, LanguageWire
LinkedIn

The language industry urgently needs to adapt to modern business needs. Henrik Lottrup, LanguageWire CEO & founder, explores why change is needed and how to upgrade LSPs.

Language goes to the heart of business. To sell a product or service, you need to be able to describe it. To draw in customers, you need to tell a story. To engage people, you need to speak their language.

But is the US $50 billion language industry fit for purpose in 2020 and beyond? Can present-day language service providers (LSPs) and translation agencies meet increasing expectations for quality, quantity, speed and cost.

The simple answer is no.

The language industry is not immune to digital transformation, shifting user behaviour and changing expectations. We need to deliver more, at far lower prices, faster and in new digitally driven ways.


Changing models

Before exploring the changes in the LSP landscape, I’d like to examine the evolution of another industry: web development.

Building a website was once the domain of IT services. A brand would work with a web agency to construct a website and continually work with them for every change made to the site.

But at some point in the late 90s, developers started to automate the repetitive task of developing websites and integrating systems. This automated process for creating websites is what came to be known as a CMS (content management system).

Today’s CMSs are an entirely different beast from those operating 20 years ago. They focus on a connected way of managing and analysing experiences. Modern CMSs, often called ‘experience platforms’, now give brands the power to personalise user experiences in real-time and omnichannel.

You might currently be asking yourself how the evolution of the CMS is a relevant analogy for LSPs and translation agencies.

This is an interesting question, because in our industry, we need to switch from being service providers to technology enablers and advisors, much as the CMS developers went through that change themselves.

”It’s not sustainable to act as a call or email centre, connecting businesses with translators in a manual and outmoded way.”

It’s not sustainable to act as a call or email centre, connecting businesses with translators in a manual and outmoded way. That’s why I get multiple approaches every week from small translation agencies looking to sell. They can no longer keep up with customer expectations and don’t have the capital to buy or develop their way out of the hole of unmet demands.

Therefore, the language industry is in need of drastic change and consolidation. Here’s a more in-depth exploration of what I see as the primary change drivers for the language industry.

More content

The Internet and ubiquitous access to information have dramatically shifted the demand for language services.

Brands suddenly need to speak 20 languages fluently.

It’s not just product information and marketing content that need to be in multiple languages. Training guides, customer support documentation, technical manuals, regulatory documentation and much more need to be produced to be accessible for all.

So how can we reconcile the finite number of translators and language experts in the world with an almost infinite demand for multilingual content?

CAT tools, translation memories, terminology databases (Termbases) and cloud-based platforms have helped us to make content workflows more efficient. Still, we need to push even further with advancement in neural machine translation, automated workflows and integration between global content platforms and content repositories that have helped us to make content workflows more efficient.

”These custom NMT engines assist translators and streamline the translation process.”

NMT (neural machine translation) is particularly exciting for us. We have hired a team of data scientists and engineers in Copenhagen to develop our own NMT solution which we train on customer content. These custom NMT engines assist translators and streamline the translation process, increasing the number of words translated per day.

Simplicity and transparency

Taking out layers of complexity and providing users with more transparency in their experience is at the heart of good design. However, this is something that the language industry struggles with.

Sending Word documents to a project manager via email who then contacts translators and passes on comments and feedback between the multiple parties involved is a cumbersome way of working.

”People want a seamless peer-to-peer experience.”

People want a seamless peer-to-peer experience. They also want to have instant answers to questions about projects such as progress, costs, quality and results. We don’t want layers of manual procedures and complexity to keep us from what we need.

If a content creator has a question for their English to Swedish proofreader, then they should be able to contact them directly. And it should be as straightforward as writing a message on WhatsApp. 

If you want to know your language spend for the month, then you should be able to see that at the click of a button. What’s more, you should be able to quickly get a thorough breakdown of your spend to help forecast future spend.

Is there a tight deadline on a more complex project involving multiple processes? Say you’re translating a white paper into eight languages and the deadline is approaching. You want to be able to see exactly at what stage each of those translations is. Is a translation stuck in the validation stage on the desk of a colleague? Is there a bottleneck in DTP? Is Swedish taking a bit longer than the other languages? I want that information to be easily accessible.

Shorter delivery times, higher quality and lower costs

Finally, buyers of language services want better content, with shorter delivery times and at a lower cost.

Businesses realise the difficulty here. They know that quality language services require experienced translators, proofreaders, copywriters, DTP engineers and voice-over artists. They also understand that the more you work with someone, the better they get, but the costs will increase by utilising them. Should they use less experienced people who might be cheaper but have longer delivery times and lower quality?

The competing demands of speed, quality and cost are something LSPs can only solve with technology.

We need to eliminate manual tasks, give customers more control, build connected platforms and rethink pricing models to optimise how we deliver value to brands. This change, by the way, is very similar to the CMS example provided at the beginning of this blog.

Fortunately, we are well on the way to delivering on these goals.

As mentioned, NMT is helping us to translate more than was previously possible. Post-editing of custom-trained NMT also reduces the cost of translating content, while maintaining quality. NMT is a massive innovation for our industry and will soon be incorporated into all language workflows. The custom-trained aspect, where the machine learning algorithm uses existing translations to understand how a particular brand speaks, has delivered a step change in machine translation quality.

Free white paper: Optimise Content Workflows with Machine Translation.

Are you struggling to create more content, in more languages, in less time, and at a lower cost? Read this handy guide on how to streamline your content workflows with Neural Machine Translation.

Another vital improvement is the integration of systems so that users can order translation and localisation from their digital platform, for example. The integration of translation with CMS, DAM and PIM platforms dramatically reduces the complexity and time needed for creating multilingual content. If you’re still sending product information in Excel sheets for translation via email, then you desperately need to look into integration options for your PIM as the ROI is considerable.

We also need to get better at handing over the automated tools we have developed to businesses. Global brands want to bring the management of language workflows in-house, and using global content platforms, such as LanguageWire, are how they can do just that.

”You don’t want to work with price-per-word focused suppliers. Our industry needs to help you maximise the ROI of language.”

You don’t want to work with price-per-word focused suppliers. Our industry needs to help you maximise the ROI of language. Where should you invest your limited budget? How can you create a smart content strategy to meet your customers’ needs?

You should be able to easily create a new language version yourself, focussing on where you create value, making great content for your target audience, analysing and optimising.

Instead of being an LSP or translation agency, we have become something more like a SaaS company that advises you on how to optimise your language workflows.

(R)evolution

And that touches on the main point of this article. We need to continue to evolve and shake up how we do things to deliver more value to our customers.

When we launched LanguageWire in 2000, our goal was to help businesses overcome the complexity of speaking multiple languages. It’s incredible where that simple goal has led us, and the changes we have seen in the language industry over that short period are astounding.

However, you can still see many businesses in our industry doing things in much the same way as they did 20 years ago.

But the rate of change will only accelerate, and our industry will look exponentially different in another 20 years.

How can we help you?

Your journey to a straightforward way of making multilingual content starts here. Tell us about your content frustrations - we would love to build the perfect content solution for your brand.