Do we fix our positioning, our messaging, or our website design first?

Position first, message second, design last. That's the order, and it's the answer to the question most growth-stage founders are actually asking when they start collecting rebrand quotes. Position is the decision about who you're for and what you believe. Message is how that decision gets said, format by format. Design is how it looks once the first two are settled.

Almost nobody runs it that way. The order gets inverted because design is the only one of the three that produces something you can look at on a Friday. Positioning produces an argument. Messaging produces a document. Design produces a page, and a page feels like progress.

Why does the design-first instinct feel so reasonable?

Because the symptom is visual. The homepage looks tired next to the competitor who just raised. The deck has four different fonts in it. Someone on the board asks why the site looks like a 2019 Series A. Every one of those observations is true, and every one of them points at the wrong layer.

A redesign makes an undecided company look more confident about nothing in particular. The type gets better. Photography gets better. The sentence at the top of the page still says the same unownable thing it said before, now set in a nicer typeface. Six months later the pipeline hasn't moved and the conclusion is that the agency was bad. That's the pattern we wrote up in why our website redesign didn't change our pipeline, and the agency usually wasn't the problem.

What do 47 homepage audits actually show?

We've scored 47 B2B homepages with the Brand Signal Score, a 19-criteria rubric across narrative clarity, trust, AI-readiness, and conversion. Twenty-six of the 47 are healthtech, nine are cybersecurity, five are AI or data platforms, four are B2B SaaS, three are fintech. Twenty-five of them disclosed a funding round, and those rounds add up to roughly $2.26 billion.

Median score: 50 percent. Twenty-three of the 47 landed below half, with a range from 14 to 81. Every one of these companies has a designed website, built by someone competent, paid for with real money.

Here's the part that matters for the ordering question. The single most common top gap, in 12 of the 47, was that nobody had named an enemy or challenged a status quo. Nine more failed on differentiation by belief rather than features. Nine more led with a solution instead of the buyer's problem. Six had no owned frame or distinctive language. Those four gaps sit in the position layer, and together they account for 36 of the 47 audits.

Now the flip side, which is the genuinely interesting number. The most common strength across the whole set, appearing 21 times, was that the hero section names who the product is for. Most of these companies got partway through step one. They can tell you their buyer. What they can't do is tell you what they believe about that buyer's situation, or which way of doing things they're arguing against. They stopped halfway through the decision and went shopping for a design system.

One honest caveat about the instrument, because it changes how you should read those numbers. Not one of the 19 criteria is about how a page looks. No typography score, no layout score, no color score. That's deliberate. The rubric was built to measure whether a stranger can tell who you are, what problem you solve, and why you see it differently, which is the Three Questions Test. The audits can't tell you these sites have bad design. They can tell you that the thing holding them back was decided, or left undecided, long before anyone opened Figma.

What does step one actually produce?

Position produces four decisions, and you'll know you have them because each one can be said in a sentence someone would argue with. Who you're for, stated narrowly enough to exclude people. What those people are currently doing instead of buying from you. What you believe about that current approach that your competitors don't say out loud. And what changes for the buyer when they're on the other side of it.

That last one is where most positioning work quietly fails. The old-way-to-new-way contrast is the spine of the Magnetic Messaging Framework, and it's the piece a design brief can't invent for you. If you hand an agency a brief that says your buyer is mid-market healthcare and your differentiator is ease of use, they'll build you something beautiful and generic, because that's the only thing that brief supports.

Position is also the layer where the difference between narrative identity and visual identity gets real, which we unpack in brand identity vs narrative identity. The visual layer is what your company looks like. The narrative layer is who it is. One of those you can buy in six weeks.

What does step two actually produce?

Message turns the position into reusable language. The hero sentence. How a rep opens a discovery call. The paragraph that goes in every proposal. Whatever your AI tools reach for when a junior marketer asks for a LinkedIn post at 4pm on a Thursday.

This step is where most of the leverage lives and where almost no budget goes. A position nobody can repeat stays on the slide it was written on. The test is whether three people from three departments describe the company the same way without coordinating. If they can't, your brand has an internal problem that no external design work touches.

Step two is also where AI-Parmesan gets sprinkled on, usually with good intentions. Someone adds AI-powered to the hero because the category did. The result is a page that reads exactly like the four competitors doing the same thing, which we've written about in AI-Parmesan and the B2B marketing plague. A strong position makes that sprinkle unnecessary, because the interesting claim is already in the sentence.

When is design finally the right thing to buy?

When you can hand the designer a decision instead of a mood board. The brief says here's who we're for, here's what we're arguing against, here's the sentence that has to land in five seconds, now make that feel inevitable. A good design team is extraordinary at that job and wasted on the other one.

There's a version of design-first that works, and it's worth naming so nobody over-corrects. If your positioning is genuinely settled, your team says the same thing consistently, and the site is simply eight years old and slow, buy the redesign. That's a real situation, and when it's time to hire a website redesign agency walks through how to tell. The failure mode is buying design as a substitute for a decision.

What if we already started at the wrong end?

Most of the work survives. Layout, component system, build, photography, the whole production layer holds up fine through a repositioning. What gets rewritten is copy, and copy is the cheapest thing on the invoice to replace. The expensive part is the second design round nobody planned for, and the quarter of pipeline that ran on the old sentence while you figured it out.

If you're mid-project right now, the move is to stop adding pages and go run the Cover-the-Logo Test on what's already live. Hide the logo, show the homepage to three people outside the company, ask them who it's for and what it's arguing. If the answers come back vague, you've found your actual scope, and the Cover-the-Logo Test is a 10-minute diagnostic rather than a 10-week engagement.

PitchKitchen builds Magnetic Messaging Frameworks for founder-led B2B companies in the $5M-$75M range. We fix broken marketing messages and underperforming websites for CEOs whose sales are stalling because their message isn't doing the work. In practice that almost always means running the order that got skipped.