Helping users create content and refine over time

Helping users create content and refine over time

Saving translations is the first step for DeepL to enable people to give and get feedback on translations and it positions DeepL as more of a ‘place where people do work’ — instead of being a transaction in their journey.

Role: Sr. Product designer, Researcher

Team: 1 Product Manager, 1 Developer

Other collaborators:

  • 2 Designers to create a vision for a new navigation and logged-in experience

  • 4 PM’s to coordinate efforts and prioritization

  • 1 User researcher to refine my research plans

In 2017, DeepL was as a scrappy startup that launched with ‘fast, accurate, and secure translations’ that surpassed the quality of Google Translate.

In 2017, DeepL was as a scrappy startup that launched with ‘fast, accurate, and secure translations’ that surpassed the quality of Google Translate.

In 2022, better translation quality was no longer good enough to drive the growth DeepL wanted. Now as a scale-up, DeepL needed to improve its offering as it shifted its focus to B2B customers. What can be built in addition to its core translation experience to add value for these customers?

With no clear path forward, I discovered new opportunities by running user interviews and drove this project to save translations and ultimately influenced product strategy — bottom-up.

What were we already hearing from users?

What were we already hearing from users?

“We’re struggling with version management”

“They’re getting it read through [by a native speaker] if [the communication] important or delicate”

“We’re using DeepL API to translate confluence pages and collaborate on writing
documents”

Users talked about getting feedback from others (despite trusting in DeepL’s translation quality). What could be driving this need?

Users talked about getting feedback from others (despite trusting in DeepL’s translation quality). What could be driving this need?

No matter how good machine translation can be, users still wanted the reassurance of being able to have a trusted friend or colleague to help review their text.

For others, their text was too long to create and refine in one sitting and they needed to work with others on it.

Time to run research! 🧐

Time to run research! 🧐

I found that collaborating on translations was an opportunity that wasn’t satisfied.

I found that collaborating on translations was an opportunity that wasn’t satisfied.

Specifically, users were trying to satisfy these 3 jobs:

  1. Co-creating a document with others

  2. Getting feedback and proofreading help from others

  3. Preparing and understanding source material for translation (Not a good fit for DeepL)

This meant features like sharing, version history, live-editing, and commenting would all be useful extensions of the translation experience

However, we didn't have the resources to built it all out at once. So, what instead?

However, we didn't have the resources to built it all out at once. So, what instead?

So we set off to build an MVP, while rallying other teams to build towards a future vision

So we set off to build an MVP, while rallying other teams to build towards a future vision

Discover, and save a translation for the first time

Re-organize their saved documents

And finding the documents once again

Why was this opportunity valuable for other teams and the company as a whole?

What could a vision for the future look like?

What could a vision for the future look like?

This vision was well received and multiple product teams have added it to their 2024 roadmaps

This vision was well received and multiple product teams have added it to their 2024 roadmaps

Our MVP hit our initial adoption metrics, and we've continuously usability tested to tweak the initial design.

In the meantime, I continue work with two other designers to break this vision down this larger change into testable smaller parts. At the time of writing, collaboration is one of DeepL’s big bets and multiple teams are working towards this shared goal.