Olark Accessible AI Chatbots’ Post

One of the challenges of building great software is building great product-feedback loops. Inside Olark one of our challenges is to build great feedback loops for assistive technology, it is not easy. In Ben's recent post, he more about this challenge, and asks for help.

I've struggled on my own enough, and decided it was worth asking for help... ....One of my goals is to build a product-feedback loop that helps our team build not just check the box accessible software, but GREAT accessible B2B software. We've worked with accessibility agencies (Allyant WCAG certified), we've hired testers (Fable), and we have many customers that have stated accessibility requirements as part of their procurement process -- Thank you Higher Ed and Public Sector and Non Profits! BUT. Very few of these organizations that care about accessibility in their procurement process seem to employ folks with disabilities in roles where they would do direct customer service over chat or manage an AI tool to do direct support over chat. They are often employed in more specialized accessibility related roles. GREAT SOFTWARE requires real feedback from real users. I want Olark to be GREAT at building AI-assisted Customer service software that sets the standard for accessible B2B customer service software. Right now we check the box, but are falling short of where I want us to be --- a tight feedback loop of product improvement from domain experts in customer service -- who use assistive technology every day. So here's my ask, if you were in my shoes and wanted to find organizations who not only employed folks who uses assistive technology, but employed them in a role where these folks did in house customer service -- who would you talk to? Note: I am aware of a couple of outsourced call-centers, but in these environments there's often not a lot of room for innovation and clients often dictate role and the software. I am also aware that the unemployment rate for folks with disabilities is ~2X folks without disabilities and likely is under reported -- which does help explain some of the complexity of creating these product feedback loops. Thoughts? Mike, Meryl, Abid, Jennison ? Timothy? Kyle? Preety? #a11y #accessibility

  • In product development we use a build (arrow) measure (arrow) learn loop.

In this image I've inserted arrows between Build and Measure, inserting Testers, Actual Users, and Actual users with Accessibility needs to emphasize that testers who are not actual domain users are not enough to build a great usable product.

To view or add a comment, sign in

Explore topics