In this weeks Product Warrior podcast titled “Things product managers forgot” we speak with Mark Rogers, senior product manager at Grapeshot. As always a big thank you to MindTheProduct and Knowit for supporting us.
Mark shares with us a story from his time as senior product manager at BrandWatch, involving the launch of a new image recognition based service using latest AI techniques.
Marks team built a new image recognition feature for the popular social brand tracking solution called Brand Watch. They assumed early on that a completely new dashboard was required for customers to handle image based data instead of text. The early learning suggested their new interface negatively impacted the customers workflow. The customers problem was the same regardless of text or image. A rich interface was difficult to let go of and some of it was built to show off the capability, but the core workflow was the more original style of dashboard.
Often we have the opposite challenge, we aim to re-use code and UX, Mark shares an example where on the surface a solution feels appropriate, due to same data models and workflows, but he demonstrates how nuances can demand very different designs.
The team wanted to give their customers the ability to see the original source of a viral image in social media. Technically this appeared feasible. The reality of the real world data found it was very difficult, the teams user behaviour assumptions were wrong. The real data meant identifying the true source was not feasible.
As the new tools went live with real usage there were some challenges requiring the team to listen carefully and react quickly. One example was manga cartoon eyes being mistaken for pepsi logos.
A lean experimental approach drove his team to a successful launch which was well received by the Brandwatch customers.
Make sure you subscribe to never miss a Product Warrior update.