Product Discovery Without Access

At my first startup, we built a consumer smart home product without ever meeting the users. A consequence of building a white-label solution for another business to sell.

That Business-to-Business-to-Consumer (B2B2C) approach is a sneaky go-to-market model. It feels like you’ve found a workaround for costly consumer distribution with a few salespeople and a great RFP submission. You don’t need to raise a bunch of money or create a clever viral hack to get started; you just build a great product and convince a customer that you’re the right partner. Suddenly you have access to millions of potential users and endless marketing budgets.

Those are real benefits. But here’s the downside: you are trading connection for reach.

Your relationship is with the customer, but all the growth potential comes from the user, someone you might never get to know. And that creates a problem that most startup advice doesn’t address: how do you build a great product for people you can’t talk to?

The Discovery Problem

Continuous discovery assumes access. Regular user interviews, rapid experimentation, tight feedback loops. You watch people use your product, hear what confuses them, ship a fix, and check if it worked. The whole approach depends on closing the distance between you and the people you’re building for.

B2B2C breaks all of that.

In B2C, the feedback loop is a circle. You ship, you hear, you learn. In B2B2C, it’s more like a tunnel. Every layer between you and the user shapes what gets through. The signal degrades at every step. By the time it reaches you, “The login button is hard to find” has become a strategic requirement for “Biometric Authentication.”

At my startup, we could only see user actions once they started installing the product. Everything before that was invisible. We didn’t know if someone bought it and left it in a box. We didn’t know if they tried to install and gave up before our system registered them. We didn’t know what prompted them to buy in the first place. We were building for users whose journey was already half over before we could see them.

So don’t confuse a rollout with full product-market fit, because there are two markets you need to fit. That’s why B2B2C models will feel successful before they are a success. You need to stay skeptical and design your product from the ground up to keep learning inside a slow-moving system.

Create Your Own Access

Before you accept the black box, see if you can build a side door.

The most direct approach is to maintain your own users. A free tier or a cohort that doesn’t go through your enterprise customer. These users aren’t meant to scale your business, but they’re your lab. You can talk to them. You can watch them. You can ship something and hear about it the next day. They’ll give you a glimpse into what’s actually working.

But they’re not a free resource. They require separate support, possibly free access, and real attention. And because they have a direct relationship with you, their experience won’t perfectly match your B2B2C users, who are behind your customer’s brand and support team. What you learn is valuable, but it’s not the same as being in the room with the end customer.

The other window is pilots. During a pilot, your customer and your user actually overlap. The enterprise is motivated to learn because they haven’t fully committed yet. The people running the pilot are usually closer to end users than whoever will manage the relationship later. And you have permission to participate by offering direct customer support.

Treat pilots as research opportunities, not just sales milestones. Sit in on user sessions if you can. Build relationships with the people running the pilot and the people participating, not just the executives who approved it. Ask for the unfiltered version of how it’s going.

This window closes fast. Once you’re “live,” you become infrastructure and your role changes. So get everything you can while the pilot is still running and discovery is still expected.

Build Feedback Into the Contract

Don’t wait until after the deal is signed to start thinking about what information you’ll need to improve your product. You need access to customer data and the ability to run tests. The only way to guarantee that right is to make sure the contract covers it.

That means thinking beyond the features list and asking: how do we continue learning once we’re live? Push for terms that allow for anonymous user data to be shared, light A/B testing, and feedback loops that don’t require prior approval.

You also need to think about the unintended consequences of contract terms. For instance, if you charge per active user, your customer has no reason to tell you about the users who never activated. It seems like a neutral term until you realise it blocks you from understanding customer behaviour.

If a customer flat-out refuses any feedback access, pay attention. What you see in the negotiation is what you’ll get in the relationship.

Instrument for Discovery

With B2B2C, you might never hear from users at all. Your product has to tell you what they won’t.

You’ll need to instrument everything far earlier than a B2C startup might need to. In B2C, you might ship a half-baked feature knowing you’ll hear about it immediately and fix it next week. In B2B2C, a problem might stay invisible for months, buried in a support queue you don’t have access to, or dismissed as user error by someone who’s never seen your roadmap.

The goal is analytics that tell stories, not just aggregate clicks, even when privacy constraints mean you’ll never see an individual user’s session. Track repeated attempts at the same action. If a user tries something three times in a row, that’s frustration you can measure without asking them. Track features that get disabled after a period of use. If someone turns off an automation feature after two weeks, they’ve made a judgment. You don’t know why, but you know something failed to deliver.

Customer Support as a Discovery Channel

You won’t get bug reports from users or long, thoughtful emails about what’s confusing. But the support staff at your customer will. Support hears the raw truth about your product daily. If you can connect with them, you’ll tap into the closest thing you have to direct user feedback.

Ask for anecdotes. Support remembers the moments when users were confused or delighted. These stories will let you see problems your hard metrics will miss. Focus on making their lives easier. Update your support tools when they need something. Offer to write up FAQs or review support documents. You’ll learn how your product is actually being used, not just how you think it should be.

Keep Your Champions Engaged

Your customer can insist a feature is critical, the most important thing in the world, and you have no way of knowing if that’s real user demand or one mid-level manager’s pet project.

This is why champions matter for discovery, not just sales. They can be your proxy for learning what’s actually happening. Help them see why understanding end users is good for them, not just for you. If they’re bought in, they’ll spend their capital on getting closer to customers.

But make sure they bring you problems, not solutions. Push for the unprocessed version. The actual complaints and raw feedback, not their analysis or feature request. Sometimes you’ll have to build it anyway to maintain the relationship. Just don’t confuse doing someone a favour with product validation.

The Trade-off

B2B2C is a frustrating way to build a product. You will make decisions based on guesses. You will build things nobody uses. You will watch problems unfold and not understand why until months later.

The model gives you massive distribution before it gives you understanding. If it gives you understanding at all. Most companies don’t survive the wait. The ones that do build systems to learn despite the obstacles.