Scott Belsky (CPO Adobe): Collapsing the Talent Stack and the 4P Framework for Building AI Products
Plus why empathy needs to come before vision and how to craft a compelling first-mile product experience
Dear subscribers,
Today, I want to share a wide-ranging interview with Scott Belsky, Chief Product Officer and now Chief Strategy Officer and leader of Design and Emerging Products for Adobe.
Scott is an entrepreneur and product leader who has also advised 100s of startups. His widely-read industry newsletter, Implications, explores how the latest tech advances our daily lives.
I spoke to Scott about:
Why the talent stack is collapsing and what you can do about it
How to improve the first-mile experience of your product
The 4P framework for exploring generative AI products
Check out our full interview below.
Why the talent stack is collapsing and what you can do about it
Welcome Scott! To start, can you share what collapsing the talent stack means?
Sure! In a big company, a team has people across many different functions.
But the best founders that I’ve seen also lead products, write code, design software, and sell to customers. It’s this ability to serve across functions that help startups compete against big companies.
I think a smaller team just has a much tighter feedback loop for decision-making.
So I’ve been thinking about how you can collapse this stack even in big companies.
In startups, people wear multiple hats. But as a company gets larger, the assumption is that people need to specialize. Are you saying that we should question this?
Yes. Let’s say you’re a founder who’s also a great product leader. When your company gets big enough, you start thinking: “I should probably hire a Chief Product Officer.”
You might start believing this even though your superpower is product development.
Well, maybe you don’t have to hire that product leader. Maybe with AI you can still lead the product and be the CEO. I think we’ll start seeing the talent stack collapse like this and it’ll help companies move faster.
How do you advise PMs to prepare for this trend?
At Adobe, we keep our 0-1 teams small to get the benefits of collapsing the stack.
Even in larger teams, my slogan is to “never waste a crisis.” For example, when COVID happened, Adobe was one of the few tech companies that didn’t do mass layoffs. Instead, we reprioritized, shifted resources, and learned how to execute better.
Overall, I favor keeping high-performing teams small for as long as possible.
You founded Behance and now lead Adobe’s product and design org. How do you think collapsing the talent stack will impact the PM and designer relationship?
The controversial answer is that:
The product manager and designer should be the same person.
If you can find someone who can do both — that’s the holy grail. After all:
I think a great prototype is worth 100 meetings.
It’s also important to find product and design leaders who can define a clear vision and narrative. Without those artifacts, the product may end up somewhere entirely different than intended.
Can you elaborate on that?
Sure. I think a strong leader needs to:
Have a vision of what the world will look like in a few years and what their product needs to become to succeed.
Craft a narrative to make people want to bring this vision to life.
Of course, day-to-day execution is important. But to me, it’s these two traits that separate leaders from managers.
Can you talk more about how empathy needs to come before vision?
This is one of my biggest frustrations.
You need to be a little crazy to create something from nothing.
So you declare a product vision, build a team, and set off on years of work. But when you finally launch, the product doesn’t get traction. It’s just 30 degrees off on product-market fit despite your team’s hard effort.
So why does this happen?
Usually, it’s due to a lack of empathy for the customer.
You’ve failed to realize that the end customer needs some gratification to use your product. Your passion has become a red herring.
What should have driven you instead is empathy for the customer and their problem. You should’ve gone shoulder to shoulder with them to identify this problem first before crafting your vision.
You must talk to customers, watch them go about their day, and ask why they’re struggling with something to give you the insight to craft a compelling vision.
I think this comes down to having some humility. Having too much arrogance and thinking “I know what customers want” usually doesn’t end well.
Yes, the cheat code of course is to solve your own problem, like many successful founders do. You're the customer so you collapse the stack of empathy.
How to improve the first-mile experience
It’s easy to build empathy for a small group of power users. But how do you build empathy for new users? If new users don’t like your product, they’ll just leave.
Most products start beautifully simple.
But then, product teams start adding features that power users want. Their products get complicated and new users eventually migrate to newer, simpler replacements.
This is the tragedy of most products in tech.
So how do you keep the product simple for new users?
I think the only way forward is to obsess about your product’s first-mile experience.
Your bar for a great first-mile experience must get higher over time. Early adopters are more tolerant of friction during onboarding. But the early to late majority will more likely just churn.
So make sure you adapt the first-mile experience to changing user profiles and expectations.
What are some tactics that people can follow to create a great first-mile experience?
Keep the first-mile experience of your product simple and introduce complexity in a manageable way as users become more familiar with your product.
Here are 7 specific tactics to do just that:
Optimize for the first 15 seconds. Users are lazy, vain, and selfish in their first 15 seconds of trying a new product. You must overcome these barriers.
Do > show > explain. Doing things for users > showing users how to do things > explaining how things are done. For example, consider including great defaults and templates that let users make progress toward their outcome.