Two products. One sells. The other stays.

Every B2B company is quietly shipping two products.

The first is the marketing site, the demo, and the pricing page. Carefully art-directed, opinions on every pixel, signed off by the founder. This product wins the meeting.

The second is the actual thing customers use every day. The dashboard. The settings page. The onboarding flow. The empty screen they see on a Tuesday morning before any data has loaded. This product wins (or loses) the renewal.

They’re rarely made by the same hands. And almost never with the same care.

The marketing site got the customer. The product keeps them.

Marketing buys the meeting. Pricing converts the meeting. The product is what makes a customer renew nine months later. And product, in B2B, is rarely “the cool feature” — it’s the screen the analyst opens every Monday, the admin panel the ops lead lives in, the onboarding flow new hires get dropped into.

None of that is in the demo. All of it decides whether the contract becomes a renewal.

Why this happens

• The marketing site is a sprint.One designer, one voice, one shippable scope. Everyone agrees on what “done” looks like.
• The product is forever. It grows one feature at a time, each one signed off by a different PM, drawing on whatever pieces are closest at hand. Nobody owns the whole shape of it.
• Templates make “good enough” the easy path. Drop in a stock table, drop in a stock modal, ship the ticket, move on. Every feature lands at 70% of finished. The 30% gap is invisible to any one team and obvious to every customer using the product daily.

After eighteen months, the marketing site looks like a flagship. The product looks like 2018. Both are doing exactly what the team optimised for.

What “good” looks like inside the product

I don’t mean visual polish. I mean the specific decisions that almost never get made when nobody’s in the room thinking about the product end-to-end:

• Hierarchy that survives real data.Most B2B dashboards look fine in mockups with three rows of data and break the moment a real customer’s 87 projects load. A designed product decides what to show, what to fold, what to hide as the data scales.
• Empty states that aren’t apologies.The first thing a new customer sees is a screen with no data on it. That screen is the most important screen in the product. Most teams ship it as a grey rectangle with the word “empty”.
• Settings, billing, and admin pages that respect the user. These are used by your most engaged customers (often the ones paying you the most). Treating them as second-class screens tells those customers exactly where they sit in your priority order.
• Onboarding that gets out of the way.A B2B buyer who already decided to try your product doesn’t need to be sold to again. They need to be unblocked. Most onboarding flows do the opposite.
• Density without clutter.Marketing pages have generous whitespace because they have one job. The product has to fit a lot on a screen because real users are doing real work. Different problem, different rules — and most B2B products use the marketing-page rules because they’re what the team practiced on.

The cost of getting this wrong

Customers don’t cancel because your landing page got worse. They cancel because the product they were promised felt like extra work to use, every day, for nine months.

And here’s the harder cost: when the product feels unloved, your team stops loving it too. People stop suggesting improvements. PMs stop asking what would make a feature actually good. Everyone settles for shipping the next ticket. The bar drops without anyone noticing the moment it happened.

What changes when you treat the product as the actual product

• The dashboard, settings page, and onboarding get the same care as the landing page. Different rules — same level of intentionality.
• The design system stops being a folder of marketing pieces and starts including the things that actually run the product: dense tables, multi-step forms, permissions, audit logs.
• Empty states, error states, and edge cases get designed instead of invented on Friday afternoon by whoever drew the short straw.
• The team stops apologising for the product on customer calls.

How to start, in one sprint

You don’t need a design team to fix this. You need one person to own the product surface end-to-end for two weeks.

Pick the single most-used screen — the dashboard, the admin home, or the onboarding flow. Redesign it as if it were going on a marketing site. Same standard, different rules. Ship it.

That one screen will reset the bar for the rest of the product. The next features will start matching the new bar without anyone being told to, because the new bar is now the closest reference. The redesign propagates by gravity instead of mandate.

That’s the unlock. The marketing site is the trailer. The product is the film. They don’t need to look the same. They need to be made by the same kind of attention.