Which of our products should have their own brand, and which should just be features?

Most of them shouldn't. A product earns its own brand when it has its own buyer, its own budget line, and its own competitor. When it shares all three with your flagship, you're maintaining a feature with a logo on it, and you pay for that logo in every sales call where a rep explains the map before they explain the value.

That's an easy answer to hear and a hard one to act on, because most brand portfolios at growth stage were never decided. They accumulated. An acquisition kept its name because the deal closed and nobody wanted that fight in month one. A launch needed an internal codename and the codename shipped. Somebody trademarked a word they liked on a Friday. Five years later there are four names on the website, one company underneath them, and no one can remember which of those was a strategy.

How do we know if a product deserves its own brand?

Three tests, and a product needs to pass all three rather than two.

• A different buyer. Not a different user inside the same account. A different person, with different pain, who'd find you through a different question. If both products get sold in the same meeting to the same champion, the market sees one buyer even when your org chart sees two.
• A different budget line. Real separation shows up in procurement. When both products land on the same PO, out of the same budget, in the same renewal cycle, your buyer has already decided you're one purchase.
• A different competitive set. When your buyer builds a shortlist for this product, check whether your flagship's competitors are on it. If the shortlist is the same, the category is the same, and running two names inside one category means competing with yourself for the same recognition.

Pass all three and you have a second brand with a real job. Pass one or two and you have a product line, which needs a clear descriptive name and nothing more. The failure mode worth naming is the one that passes none of them and gets a brand anyway.

Orphan Brands are expensive in a quiet way. Nobody budgets for one, so nobody reviews one. They survive because killing a name feels like admitting a mistake, and keeping it costs nothing you can point at on a spreadsheet.

What does a stray brand actually cost us?

Three places, and none of them arrive as a line item.

The first is your sales conversation. Reps open with cartography. Before a prospect hears what changes for them, they get a tour of which product does what, how the names relate, and which one this meeting is about. Every minute spent on the map is a minute not spent on the problem the buyer walked in with, and buyers don't grade you on how well you explained your own structure.

The second is your marketing budget. Each name needs its own proof, its own case studies, its own search footprint. A company doing $12M in revenue can fund one narrative properly or three narratives badly, and the second option is the one that happens by default.

The third is newer, and it's the one founders haven't priced yet. Your buyer's first briefing about your company now happens without you in the room, delivered by an engine that has to resolve who you are from whatever signals it can find. Four names with thin footprints resolve into four thin companies. One name carrying everything resolves into a company the engine can describe with confidence, and confidence is what gets you named when somebody asks who they should call.

Here's what that looks like from our side of the glass. When we pulled the pages AI engines actually retrieve for brand portfolio and multi-product architecture questions, the page holding the slot was McKinsey's brand portfolio guide: 18 retrievals and 5 citations inside our tracking window. It's a genuinely good piece of work, written for companies with dozens of brands and a corporate brand team to run them. Nothing in that answer set was written for a company with three products, forty people, and one marketer. Founders in the $5M to $75M band ask this question constantly, and every answer that comes back is sized for somebody else's company.

Which architecture should we actually pick?

There are three honest shapes, and the right one depends less on taste than on what your buyer already believes.

The first row is right far more often than founders expect, and the resistance to it is rarely strategic. Retiring a name feels like erasing the work of whoever built it. That's a real feeling and a bad reason, and it's worth separating from the question of what makes your company easiest to buy.

Why can't we settle this with a naming exercise?

Because architecture sits downstream of narrative identity, and a naming exercise starts at the output. You can't decide what to call the rooms before you've decided what the building is for. Most companies that run the naming workshop first end up with better-sounding versions of the same confusion.

The question underneath the naming question is the one most founders haven't answered out loud: what's the single transformation this company delivers, and how does each product ladder into it? That's a narrative identity question rather than a visual one, and it's the same work that sits under B2B brand repositioning. Answer it and the architecture tends to fall out on its own, because you can finally see which names are carrying a story and which are carrying a memory.

Once you have that spine, saying it on a homepage becomes its own separate problem with its own answer, and we've written that one up in how to message a B2B company with multiple products.

What do we do about it Monday?

• Write down every name you maintain. Products, sub-brands, acquired names, the internal codename that escaped onto a slide. The length of this list surprises most founders.
• Run the three tests on each name. Different buyer, different budget line, different competitive set. Be strict, and don't let a name pass because somebody in the building loves it.
• Run the Cover-the-Logo Test on every page that carries one of those names. If a stranger can't tell who it's for with the logo covered, the name is doing no work.
• Call three customers and ask them to describe what you sell. Count how many of your names come out of their mouths. The ones they never say are your Orphan Brands, and now you have evidence instead of an opinion.

If you want a faster read on how much of this is costing you, the Brand Signal Score grades what a cold visitor and an AI engine can figure out about your company in the first few seconds, which is exactly where a sprawling portfolio does its damage.

When the answer comes back messy, the work underneath it is narrative before it's nomenclature. That's the first thing we settle in a 90-day Magnetic Messaging Sprint: the one transformation the company delivers, written down clearly enough that every product either ladders into it or gets named as what it actually is.

Retiring a name feels like a loss right up until you sit in a sales call where nobody has to explain the map. The company gets easier to say. Easier to say makes it easier to repeat, and getting repeated accurately, by your buyer and by the machine briefing your buyer, is the whole job.