Organization & Role
Helpshift (acquired by Keywords Studios) is a mobile-first customer support platform. As their first Content Designer, I created and managed all in-app copy and knowledge base content.
Project
I wrote the in-app copy for SensAI, Helpshift’s AI-powered engine for automatically classifying support tickets into categories, prior to its release.
Problem
To be successful in the multi-step onboarding process, customers need to learn three new key terms. The challenge was presenting users with all this new information in a way that felt digestible, and which followed the chronological setup process so they would learn exactly what they needed to know each step of the way.
Prior to the content revision phase, this was the placeholder content written by the Product Manager:
Approach
Users won’t see this screen until they’ve already paid to unlock this feature, so there isn’t a need to sell or market it here; it was taking up valuable real estate in a space where they could start learning how to set it up.
With this in mind, my goals with the copy revisions were to:
- Describe the primary use case so they could adapt it to their needs
- Define the three new terms that will be important for them to understand (labels, Models, and datasets) throughout setup
- Provide any troubleshooting tips needed to unblock them at key points of friction
It was a lot to accomplish in one paragraph, so I made a recommendation to the PM and Designer to break up the introduction modal into three panes.
Outcome
Once approved, working within this new design flow, I re-architected this content into a three-part introduction:
I used the titles of each pane to state the primary benefit of the feature and bolded the new term for them to pick out, just in case they are moving fast and skim these lines.
Defining one new term in each slide felt like the best way to avoid overwhelming users with a bunch of new information in one go – this description also flows chronologically with how you set up the feature.
The last screen touches on the fact that this feature requires maintenance – the benefits of doing this work are addressed first, then the work needed is added after. A common point of feedback in early beta was that customers didn’t realize they had to do this maintenance until after setup, so I added this to avoid catching them by surprise.
A couple of other points about the copy decisions for SensAI:
- Users will only ever see these slides once, so I kept it light with no links to avoid breaking the flow.
- We make decisions about capitalization based on which features are branded as unique to Helpshift and which are not. This is why you see ‘Language-based Model’ (a unique aspect of the SensAI engine as opposed to our competitors) capitalized, but not ‘label’ or ‘dataset’.
Results
This updated copy was shipped during our beta; all of our beta customers were able to successfully self-onboard with SensAI with minimal assistance needed.
