Developments in writing-assistance technology Public  

New tech and investment in writing assistance

Following the announcements at the TAUS annual conference in Vancouver, we explored, and reported on, two important trends: 

  1. A shift in LSP technology – MLVs future-proof their management systems and make sure they support many new services (not just translation), and;
  2. An historical milestone reached by machine translation – more than 1 trillion words translated every day

But there is one more trend left to explore – developments in writing assistance technology.

From CAT-tools to computer-assisted writing tools

While most professional translators work in CAT-tools, and technical writers have generally agreed to switch to XML editors that can reuse phrases and ready-made parts of text, the vast majority of copywriters still work in Microsoft Word and Google Docs without any significant computer help.

Enterprise content strategists often consider content creation and translation as part of an overall “center of excellence“, rather than two separate functions, so they are interested in new tools for content creation. Tech companies see an opportunity here. 

Professional tools for corporate writers aim to have the following features:

  1. Content scoring with AI (terminology, emotion, style, readability)
  2. Automatic suggestions to improve
  3. References to relevant content and integrations with external knowledge databases
  4. Integration with terminology and translation memory, and;
  5. Workflow management

New competition enters the corporate writing assistance space

At TAUS Vancouver, two new content tools have been announced by translation tech providers. Smartling has shown the first screenshots of Draft, a Google Docs plugin for text quality, while SDL has hinted at SDL Content Assistant. Smartling Draft scores text quality in Google Docs and provides suggestions for text readability and use of the company’s terminology. The aim is to create better content and sync with translation steps for better translation memory reuse. The writer gets suggestions to improve the text and as she does, a large gauge on the right tilts towards green.

When writing content, there is an element of a game for the human and for the underlying computer algorithm – there is a lot to learn about creating an engaging writing style. Smartling Draft syncs with the glossary in Smartling’s translation tool so that both the writer and the translator share the same list of terms and speak in the company’s voice. It will also be possible to publish content to the CMS from this tool, or send it to translation. Smartling Draft is somewhat similar in its idea to Acrolinx, launched in 2002, which also has a scoring indicator on the right side of the UI, and provides a list of suggestions. Acrolinx can also check the translation memory and suggest changes to the text to improve reuse of the memory during the translation step, thereby reducing the cost of localization even as the text is being created.

Smartling Draft user interface (click to zoom)

Acrolinx interface (click to zoom)

In a stint of similar thinking, SDL Content Assistant, which launched on November 7 of this year, can provide content summaries for writers, highlight facts, quotes, and figures, and can suggest relevant information from social media. Congree, a new entrant to corporate writing assistance (and an offshoot of Across Systems in Germany) will compete with Acrolinx mentioned above. New entrants will also compete with the San Francisco-based TMS Qordoba “Text Intelligence” product. The field is quite small at the moment. In fact, all the main players taken together most likely have around 500 corporate clients between them. Priced around USD 30 – 50k a year, these products together should be making up to USD 15 million.

Investors show interest in writing assistance

In October 2018, Congree became independent from their mother company backed by “private investors from the IT industry”, while Qordoba’s tool, launched in 2017, received USD 11.5 million in series B funding (October 2018). The fact that Smartling and SDL launched content tech, and that there have been investments into Qordoba and Congree, there is clear evidence that this niche is very dynamic at the moment.

In the future however, corporate writing assistance products created by localization people may become overshadowed by consumer products that come from industry outsiders. To put things into perspective, the Grammarly tool, founded by three Ukrainian entrepreneurs, has more than 7 million individual users and has received USD 110 million in funding. There are also other tools to consider, including Hemingway Editor, and Readable App.   

At the moment, however, tools like Grammarly don’t have have a strong link with translation. They may not allow for organization-specific style guides, they may not have a robust API to connect with a TMS, and they may not be able to leverage TM content and terminology. If at some point, in response to growing customer requests, Grammarly integrates with translation systems for terminology and translation memory, it will be difficult for smaller companies founded by the industry natives to compete with such a presence.

One thing is for certain – investors and technologists want corporate copywriters to eventually drop plain MS Word and Google Docs and begin working in more professional environments so that the art of writing will be backed up by science.

Stay up to date as Nimdzi publishes new insights.
We will keep you posted as each new report is published so that you are sure not to miss anything.

  • Sign up
Password Strength Very Weak
Lost your password? Please enter your username or email address. You will receive a link to create a new password via email.
We do not share your personal details with anyone.