SlatorCon London 2024 featured a panel discussion focused on the business benefits of machine translation (MT) and language AI. Three experts from DeepL and Clarivate explored the aspects of compliance, quality, and ROI and discussed how MT is used across a range of organizational settings to enhance operational efficiency.
David Parry-Jones, CRO at DeepL, Joy Uzuegbu, Product Marketing Manager at DeepL, and Chloe Ratcliffe, Director of Translation at Clarivate, exchanged views as panel experts. They were joined by Slator’s Esther Bond, who moderated the discussion.
Parry-Jones explained that in addition to DeepL’s several hundreds of thousands of power users across organizations, the company is increasingly observing that “end users within those organizations are reaching outside of businesses to our consumer product to get translation.”
On the topic of compliance, he added that DeepL aims to provide a route for these end users to access MT in a way that is compliant with their business’s practices, touching on the phenomenon of “shadow IT” — which can impact how organizations select which technologies they are using. Uzuegbu elaborated, saying that it’s important to be vigilant about how translation data is handled to protect clients’ information and comply with regulatory standards.
For Clarivate — a major global information services provider focusing on the areas of academia and government, life sciences and healthcare, and IP — safe practices are paramount. As Ratcliffe observed, “The best way to control [employees’ use of public machine translation] is obviously to try and bring all the users into your controlled platform, and have everything running through that.”
Adding controls to employee access to machine translation introduces a layer of cost for a business that wasn’t present before. Yet, as Parry-Jones remarked, “There may be increased cost, but what is the adverse cost of non compliance?”
ROI and Requirement for MT
From cost to quality, DeepL’s Uzuegbu said that everything the company brings to market is subject to an “extremely rigorous process that is backed by a network of language specialists.” What is more, customers are displaying an increasing “need for translation that is customized to the organization that’s using it,” for example with glossaries, Parry-Jones added.
Ratcliffe also highlighted the critical balance between maintaining high translation quality and managing costs, pointing out that “if the quality isn’t there and if the end output isn’t what you need, everything else is wasted.”
Central to the discussion about quality and cost is the concept of return on investment (ROI). Parry-Jones shared some data points from a recent Forrester’s Total Economic Impact (TEI) study commissioned by DeepL that interviewed a number of anonymous customers. Overall, he said, “I think the definition of ROI around deploying MT is very clear. The requirement for it […] is very demonstrable because people are consuming it outside of your organizations, whether you want them to or not.”
A key theme for the panel was ease-of-access to machine translation. As Parry-Jones said, “you need to give translation to people at the point of use where they want to use it.” Translation professionals and power users of DeepL have a specific set of needs and features but “most users in most organizations want to consume it in their productivity suite.”
Clarivate’s Ratcliffe echoed this perspective: “Really, if you want people to be compliant with the way of working, then it’s got to be easy to use. If you can get something quicker elsewhere, then you’re going to go and do that,” she said.
Rounding out the discussion, the panelists fielded a number of audience questions around the topics of LLMs, data security and customized translation.