notes on
performance

How to build a site that sells, and keep it selling.

Your website is sales infrastructure.

Your website is infrastructure, and infrastructure either carries load or creates drag.

Your buyers have a problem, a shortlist, and a few minutes. They are checking whether you handle situations like theirs.

The site's job is to do what a salesperson would in the room: confirm the buyer is in the right place, pre-empt objections, and make the next step obvious. Every page moves someone toward a decision or leaves them to figure it out alone.

Most sites do the latter, and not by accident. One of the most common patterns we see in B2B is a services page organized around the company's org chart. That structure decides who does the translating, and it assigns the job to the buyer. They have to reverse-engineer what you do from a menu of departments. Most will not bother. A few email to ask, and a salesperson spends twenty minutes answering what the page should have said in five seconds.

We call this org-chart architecture: a site organized around how the company is built rather than how the buyer decides. Its cost lands on the people around the site. That is what drag looks like: work the structure should have absorbed, redistributed to your team.

Because the problem lives in the structure, it survives every copy pass and every redesign.

You cannot restyle your way out of confusion.

The fix is buyer architecture, where the site does the translating. Everyone agrees a site should follow the buyer's journey. Fewer teams treat that as a constraint on the order of the build. Buying decision first: who is evaluating you, what they need to see, in what order. Structure follows. Design follows structure.

On our rebuild for Uassist.ME, we restructured seventeen years of accumulated services into five buyer paths. The company did not change. Its buyers no longer had to understand its structure to know whether it solved their problem.

Getting that order right at launch is necessary. It is not sufficient, for two reasons.

The site has to be able to change.  Most are built to launch, not to last: layouts harden into constraints, content models go rigid, new sections need workarounds. When small edits feel structural, teams stop making them, and improvement becomes optional. So build for the second year: components that get reused, a content model that fits how the site will actually be run, and enough structural slack that a new service does not require a redesign.

The buying decision moves. Positioning sharpens, services ship, the ICP changes shape. What the site said at launch was accurate at launch. The gap between a well-built site and a well-aligned one is hard to see from inside, and the signal is rarely a broken page. It is a sales process doing work the site should be doing. So review the site when the business changes: new positioning, new service, new buyer, new review. Caught early, drift is a calibration. Left alone, it is a rebuild.

Currency is the precondition for all of it. You cannot build for a buying decision without knowing what it currently is, and that answer has a shelf life.