Explainability and Product Context - Follow Up Story
AI evaluates your product before people do
What happened after writing a recent Product Management IRL article? These insights are for paid subscribers to Product Management IRL.
This week’s backstory is about putting explainability in product context:
What Prompted This Article?
Recently I asked AI about my product. The response was mostly accurate. But the answer wasn’t convincing for a potential customer.
This led to a new question:
Is there anything I can do as a product manager to “make” AI give a more appealing answer for people thinking about buying my product?
It turns out there isn’t a clear answer about improving what AI thinks about our products. And there is very little information about measuring the perception of AI over time.
When I think like AI, then I come back to product context. AI finds all the material on my product and then responds to questions.
After some research, I found this type of material is more likely to be used by AI:
Recent stuff - material created or updated in the past month
Outside reference - established publishers that reference your material
Social media - Reddit and YouTube especially
Each of these items would ideally point to my product context.
But what product context matters?
The items that make a difference to potential customers are:
clear product positioning in product material
material that covers how to configure and use the product
showing product outcomes for customers
Here is a visual to show before and after product context explainability is added.
In my product, this information is easiest to cover in the product FAQ. I know well-established products have whitepapers to cover many of these aspects. For most product managers, there is a limited budget for whitepapers.
FAQs are usually created and maintained by product managers. After sending this article out, I focused on product context maintenance and the FAQ.
