Marketplaces Were Built Around Products. The Next Generation Will Be Built Around People.

Financial marketplaces were built close to the transaction because that's where the economics worked. Rank for best personal loans, capture someone already shopping, compare products and monetize the click. It was a great model.

AI is changing the value of that position. Answers and basic product comparisons are becoming abundant. I believe the more valuable position will move upstream: understanding what someone is trying to accomplish before they know which product they need.

That's a fundamentally different marketplace.

The old model starts too late

Affiliates historically optimized around high-intent searches because those consumers were easier to monetize. Someone searching best balance transfer cards is worth more today than someone asking how do I lower my monthly payments?

That economic incentive shaped the internet's financial marketplaces. They built enormous content libraries around products, ranked for high-intent keywords and created increasingly efficient funnels from search to click.

The model works. The problem is that AI is becoming very good at the information sitting on top of it. Definitions, basic education, product research and even comparisons can increasingly be generated on demand.

The strategic question becomes: what remains valuable when the answer itself becomes cheap?

I think it's understanding the person asking the question.

Start with the outcome and work down

Consider someone who says, “I need to lower my monthly payments.”

A product-first marketplace wants to route that intent toward a product category. But there isn't enough information yet to know whether that person should consider a personal loan, balance transfer, home equity, refinancing, changing repayment behavior or doing nothing at all.

Before recommending anything, you need context. What payments are creating pressure? What debt do they have? What's their credit profile? Do they own a home? Are they optimizing for monthly cash flow or total cost? How quickly do they need relief?

That creates a different journey:

The product still matters. It just comes later.

And the data created before the product may ultimately be more valuable than the click itself.

Keywords built the last generation. Decision journeys may build the next.

Search marketplaces learned which keywords indicated commercial intent. The next opportunity is understanding which prompts, answers and attributes indicate where someone is in a decision.

Someone starts with “I'm struggling with my credit card payments.” What do they ask next? Which answer changes their direction? What additional information helps distinguish one path from another? Which attributes eventually predict the strategy that makes sense?

At enough scale, those interactions create a learning loop:

That is very different from knowing that someone who searches best debt consolidation loans is likely to click on a lender.

It also requires more than putting an AI interface on top of an existing marketplace. The difficult part isn't generating conversational answers. It's knowing what questions need to be asked, what information matters and how those answers change the decision.

That requires domain expertise, structured data, consumer trust and enough real interactions to learn from.

Incumbents have a difficult tradeoff

Existing marketplaces can move upstream, and many will try. But their biggest strength also creates tension.

Their funnels were built to monetize product intent efficiently. A truly outcome-first experience sometimes has to conclude that a different product is better, the consumer isn't ready, or no product is the right answer at all.

That's difficult when the existing economics reward getting someone to a monetizable click.

This is one reason I think new entrants have an opportunity. They can design the journey around the decision from day one rather than retrofit it onto a product funnel.

It's also one of the bets we're making with Rose Cove. We're starting with outcomes like lowering monthly payments, getting out of debt, or borrowing cash, then working down toward the appropriate decision. We're early, and there is plenty left to prove.

But the thesis is simple:

The first generation of marketplaces got very good at answering, “Which product should I buy?”

The next generation needs to answer, “Given my situation, what should I do?”

That isn't a better comparison marketplace.

It's a different starting point.