A straightforward way to conceive of a product is as a bounded collection of features. For example, a personal financial management product’s features include monthly budgeting, real-time spending tracking, and integrations with bank data sources. Features like this are often listed on marketing materials. They make good bullet points.

This conception of product-as-feature-set is fine, accurate, and easy. Again: straightforward. But it’s inadequate for top-grade product folks like the Pollinator community—especially when AI changes the fundamental economics of product-making.

An analogy: think of how you value a beloved pet. Now, you might say, “but my dog isn’t like finance software! My pooch Duchess is so much more than that!” To that I say: yes, exactly. A good product delivers value beyond its feature set, just as a family pet is more than its observable characteristics.

For example, my own mixed-breed dog Teddy weighs in at forty pounds, and he has short brown fur, double dew-claws, and weirdly small ears. I joke that he’s a champion-grade example of the American Brown Dog. Teddy does not subscribe to The Pollinator; a feature that he lacks is the ability to read.

But that’s not enough to really understand Teddy. In addition to his physical features, Teddy has a presence and a temperament. He is loyal. He is friendly. He is generally pretty chill, unless deer wander through the yard; then he lets us know. More than that: he plays a role in our family. His presence is calming. He keeps us active. Years of accumulated interaction, care, training, companionship, and trust create a relationship. That, not the attribute list, is why a pet relationship is so meaningful.

What’s this have to do with managing a product? Sharp product managers know that “observable features” aren’t actually how a wise business thinks about its products. If your business leaders are making decisions about where to invest the company’s capital, they usually aren’t deciding based on proposed features, as cute as your dog might be.

See, a business prioritizing product investments isn’t buying a set of features, it’s making a bet on value sustained over the course of a relationship. Products exist to create value for users in a way that also creates value for the organization that builds them. The best products, and the best product managers, strengthen that exchange.

So of course, care about the product’s story, capabilities, functions, and interfaces. But when making funding decisions with business leaders, step up to a new altitude. That’s where your business partners already operate. Don’t focus on the next feature to add; focus on the next problem to solve. With AI-enabled development, that’s where your own value as a product thinker increasingly resides.

Your product has dashboards and screens and integrations and interfaces. Those are its features, just like a dog has brown fur and small ears. But more than that, your product comprises the value that you discern through your experimentation and judgment. Just like the dog is a very good boy.

On to the Garden,

Do you enjoy reading The Pollinator? Have you heard me (Jason Scherschligt, editor) speak at a recent conference, like Open Source North or Product Camp Twin Cities)? With Product Explorations from SDG, I’m available for speaking engagements on product topics at your company. I’ve joined department lunch and learns, product team off-sites, or leadership retreats. Participants enjoy it—or at least they say they do. I know I do!

Learn more here: Product Explorations from SDG

Swedish product consultant Marcus Castenfors describes a scenario recently encountered by product and design legend Steve Blank. Blank teaches in an entrepreneurship program at Stanford. Using AI capabilities, Blank's students are now demoing complete, polished products, not just showing mere wireframes or Powerpoint-ware like they once did. But Castenfors (and Blank) notes that the high fidelity of these AI-assisted demo-able products has actually changed the type of insights students receive--and not for the better.

This observation isn’t new, but Castenfors connects it to various cognitive biases: confirmation bias, automation bias, the IKEA effect, and the illusion of explanatory depth.

This suggests the importance of unpolished, low-fidelity prototypes and early drafts. Their rough edges say "we are still working on this, and before we get too far, we want to learn from you." A hypothesis: the more finished a product or system appears to be, the less likely early reviewers are to participate in improving it. And a challenge: how might we take advantage of GenAI's incredible ability to produce working demos quickly, while we still invite customers and early reviewers to shape the product alongside us, through genuine, actionable feedback and insights?

In this recent live talk, Claire Vo says she is “coming for the roadmap.” And come for it she does. She starts with the provocative notion that, given AI’s ability to provide “limitless engineering,” product managers may have written their last roadmaps.

Over the course of her talk (and through her other public content) Vo adroitly recasts the product manager’s classic domain and artifacts—backlogs, roadmaps, vision statements, feature lists—to fit the speedy engineering capabilities opened up by AI-assisted product development. She herself builds and ships at rapid pace, but she doesn’t make the mistake of thinking that this rapid building is the real differentiator in a modern product pro’s work.

Instead, Vo identifies the most important intersection between software product management and AI: value creation and judgment. Her talk is also chock-full of quips and aphorisms that I swear should get emblazoned on a product manager’s bumper sticker. Here’s one great quote.

Golden, huh?

Melissa Perri’s 2018 book Escaping the Build Trap has a firm place in the Product Management literary canon. It’s a book that reminded product managers of the “trap” that ensnares many product teams: we build the thing we originally conceived, without relentlessly testing for its value.

Here David J. Bland, a well-known consultant and author (Testing Business Ideas), interviews Perri about the notions she first explored in that book. Their discussion focuses on how Perri’s conception of the build trap changes in an AI-enabled environment. Unsurprisingly, Perri notes that even as AI revolutionizes the product development process, it makes the build trap even more tempting. But she also identifies an escape route: smart people who are equipped to “steer the AI.”

I also appreciate that towards the end of the interview, Bland and Perri discuss AI and organizational factors, concluding that “AI won't fix broken organizational systems.”

Bland’s a good interviewer; he’s systematic but collegial. Perri is an even better interviewee.

If you are wondering what breed of dog is best for you, check out the American Kennel Club’s Breed Selector. Through a wizard-like interface, it asks questions about your preferences and lifestyle and returns a list of dog breeds that might suit you. You can even prioritize the categories to really nail the recommendations. Unfortunately, it doesn’t include mixed-breed mutts like my own dog. But Teddy’s one of a kind, anyway.

Check it out at https://www.akc.org/breed-selector-tool/