I’ll research current data on website project scoping, briefs, and digital expectations before writing.I have strong data. Let me get one more angle on requirements gathering and the cost of fixing problems late versus early.I have everything I need. Here’s the article.
The brief says “fresh, modern, easy to navigate.” The expectation is a lead engine.
Read enough website briefs and a pattern jumps out. The document describes appearance. The client wants performance. Someone writes three paragraphs about colour palettes, “clean design,” a homepage hero, and a navigation that “feels intuitive,” then ends the conversation by asking what the new site will do for pipeline. Those two things are not the same request. One describes a brochure. The other expects a machine.
This gap is not a small communication hiccup. It is the single most expensive misunderstanding in the entire project, and it happens before anyone has agreed a budget. You are about to spend a meaningful chunk of money on something you have described as decoration while privately judging it on whether it generates revenue. When the result looks lovely and converts at the same dismal rate as the old site, everybody is surprised. Nobody should be.
The fix is not better adjectives. It is deciding, on paper, before you brief anyone, whether you are commissioning a thing that looks good or a thing that works. Then writing the brief for the second one.
A brochure describes itself. A machine describes what it produces.
The tell is in the verbs. A brochure brief is full of static adjectives: modern, professional, sleek, bold, trustworthy. A machine brief is full of outcomes: a manufacturing buyer should be able to find the right product spec and request a quote in under two minutes; a sales rep should be able to send a prospect to a single page that answers the four objections that kill most deals.
One of those briefs can be evaluated. You can test whether a buyer found the spec sheet. You cannot test whether something is “sleek” because sleek is not a measurable claim, it is a feeling, and feelings cannot be scoped, costed, or held against a vendor at launch.
Here is why this matters commercially. A B2B website earns its keep through conversion, not decoration. According to Unbounce’s 2024 Conversion Benchmark Report, which analysed over 57 million conversions across more than 41,000 landing pages, the median conversion rate across all industries is 6.6%. But for high-intent B2B actions, the kind that actually indicate someone is evaluating you as a vendor, the numbers are far more sober. Even Salesforce, with its stellar reputation, converts less than 5% of its traffic into qualified leads, and most B2B conversion rates aren’t higher than this, particularly in industries with longer sales cycles. When you are operating in that range, the difference between a 1.5% site and a 3% site is the difference between a website that pays for itself and one that quietly drains budget. None of that shows up in a brief that talks about colour.
The brochure brief hides the hard questions
The reason briefs default to appearance is that appearance is easy to write about. You can have an opinion on a hero image in five seconds. Working out which three buyer segments the site must serve, what each one needs to do, and what content has to exist for them to do it takes weeks and several uncomfortable internal meetings. So people skip it. They write the easy brief, send it to vendors, and assume the hard thinking will happen “during the project.”
It rarely does. It happens during arguments, mid-build, when it is most expensive to resolve.
Why the gap is so costly, and why it shows up late
The single most reliable predictor of a failed website project is not a bad designer or a slow developer. It is fuzzy requirements at the start. Unclear requirements account for 39% of project failures, according to PMI data. That is not a rounding error. That is the largest single cause, and it traces directly back to a brief that described a brochure and never specified what the machine had to do.
The cost compounds because the problem surfaces late. There is a well-known curve in software, originally documented by IBM’s Systems Sciences Institute, showing how the cost of fixing a problem grows the longer it stays buried. The cost multiplier progresses: 1x in design, 6.5x during implementation, 15x during testing, and 60 to 100x after release. The exact ratios get argued over, and fairly so. But these figures are directional, meaning exact multipliers vary, but the exponential increase is consistent across research. A requirement you clarify in week one costs a sentence. The same requirement discovered after launch costs a rebuild.
Apply that to a website. The brief said “modern and easy to navigate.” Three months in, sales finally looks at the staging site and points out that there is nowhere for a buyer to compare two product lines, which is the single most common thing their prospects ask for. That is not a tweak. That is information architecture, new templates, new content, and a renegotiated timeline. It is expensive precisely because everyone treated it as a detail to sort out later, when it was actually a core requirement that belonged in the brief.
Who causes this, and it is usually not the agency
It is tempting to blame the vendor for not asking better questions. Sometimes that is fair. But across the projects we see, the brochure brief is almost always an internal failure of alignment that no external party could have prevented.
The mechanism is predictable. Website redesigns rarely involve just one person. Marketing may want more lead-generation pages. The product team may push for new features. Leadership may focus on branding. If these voices are not aligned at the start, conflicting requests will creep in later. So someone in marketing writes the brief alone, in a hurry, and writes the only thing they can write without a fight: a description of how the site should look. The genuinely contested questions, what the site is for and whose job it makes easier, never get resolved, because resolving them means getting leadership, sales, and product in a room and forcing a decision.
Then it gets worse mid-project. Scope creep usually doesn’t start with “we want free work.” It starts with a normal, human thing: new information arrives late. A classic case is when a stakeholder, often SEO, sales, legal, or a late-joining exec, steps in after key pages were already approved and asks to “quickly adjust” messaging, structure, or even whole sections. The late stakeholder is not the villain. The villain is the brief that never gave them a reason to engage early, because it described a brochure they had no opinion about rather than a machine that affected their numbers.
The board member chasing a competitor’s site
There is a specific version of this worth naming. An executive sees a competitor’s website, likes it, and the brief becomes “make ours like theirs.” This is the purest form of the brochure trap. You are now scoping your business-critical asset around the surface of a company whose strategy, buyers, and sales process you do not understand. You have copied the paint job of a machine without knowing what it was engineered to do. Resist it. Admiring a competitor’s homepage tells you nothing about whether it generates revenue, and quite often it does not.
What a machine brief actually contains
A brief that scopes a machine starts from outcomes and works backwards. It is harder to write and far cheaper to build against. The shift is well understood by the better practitioners: before locking in budget or scope, be clear on what this redesign needs to accomplish for the business. A visual refresh alone rarely justifies the investment. Define the primary goals driving the redesign, whether that’s improving lead quality, supporting a new go-to-market strategy, repositioning the brand, or enabling faster content updates.
In practice, a machine brief nails down a handful of things that a brochure brief leaves vague:
- The primary jobs the site must do. Not “showcase our services,” but “let a procurement lead shortlist us without talking to sales” or “give a returning buyer a one-click path to reorder.”
- Who the site is for, named specifically. The two or three buyer types that matter, what they arrive wanting, and what evidence they need to act.
- How success will be measured after launch. A target conversion action and a realistic benchmark for it, grounded in your actual deal size and traffic, not a vanity figure.
- What is explicitly out of scope. This one is non-negotiable. If what’s not included isn’t written, clients will assume it is included.
That last point does more work than any other. Vague scope is the open door. As one analysis puts it plainly, when a project scope says “build a website” instead of “build a 5-page marketing site with contact form, blog with CMS, and responsive design,” every additional page or feature request feels reasonable to the client, because the scope never said otherwise. Naming what you are not building is how you protect the budget for what you are.
Replace adjectives with measurable buyer behaviour
The discipline that turns a brochure brief into a machine brief is translating every aesthetic instinct into a behaviour you can observe. “Trustworthy” becomes “a first-time visitor can find a relevant case study and proof of results within one click of the homepage.” “Easy to navigate” becomes “a buyer can reach any product detail page in two clicks from any entry point.” Now your designer has a target, your developer has a spec, and you have something you can actually test at launch instead of arguing about taste.
Form design is a good, concrete example of why behaviour beats opinion. There is a measurable relationship between how much you ask for and how many people finish. Analysis of over 40,000 landing pages found a clear negative correlation between the number of form fields and conversion, and as form fields increased from 3 to 7+, conversion rates dropped consistently; for high-intent B2B conversions, stick to 3-5 fields. A brochure brief never mentions the form. A machine brief specifies it, because the form is where the machine either produces a lead or does not.
The discovery work has to happen. The only question is when.
Here is the thing the brochure brief is really avoiding. The hard scoping work does not disappear when you skip it. It just relocates to the most expensive possible moment. You can do the thinking up front, when changing your mind costs a conversation, or you can do it mid-build, when changing your mind costs a change order and a slipped launch.
In our experience, most project failures trace back to this exact decision. The teams that get burned are not the ones who picked the wrong vendor. They are the ones who briefed a brochure, signed a fixed price against it, and then discovered the machine they actually needed once the work was underway. The proposal looked competitive because it was priced against the easy brief. The real project was always going to cost more, and the gap got paid for in overruns and frustration rather than honest scoping.
This is the entire logic behind a blueprint-first approach. Validate what the site must actually do, with the real stakeholders in the room, before anyone commits budget to building it. What we recommend to clients is to treat discovery as a distinct, funded phase that produces a real specification: named buyer jobs, agreed scope, explicit exclusions, and a measurable definition of success. Not a mood board. A blueprint you could hand to three different vendors and get three comparable quotes against, because they are all pricing the same machine.
The data backs the discipline. The reason so many projects run over is structural: the widely cited Standish Group CHAOS research found that only 29% of projects are completed on time and on budget, which means 71% either overshoot their timeline, exceed their budget, or both. Front-loading the requirements work is the cheapest intervention available against that pattern, because it moves the expensive discovery to the cheap end of the cost curve.
What to do this week
Pull up your current brief, or the draft you are about to send out, and run a simple test on it. Highlight every sentence. If a sentence describes how the site looks or feels, mark it one colour. If it describes a specific thing a named type of visitor will be able to do, mark it another. Then look at the balance.
If the page is mostly the first colour, you have written a brochure brief and you are quietly expecting a machine. Before you send it anywhere, do three concrete things. Get your sales team to write down the four questions prospects ask most often that the website fails to answer; those are your most important pages. Get leadership to commit, in one sentence, to the single business outcome that justifies the spend. And write a short “not in this project” list, because the items you exclude on purpose are the budget you protect for the items that matter.
If you want a second pair of eyes on whether your brief is scoping a machine or a brochure, you can book your free discovery call and we will pressure-test it against your actual goals before you spend a penny on the build.
A website is bought once and lived with for years. Spend the awkward weeks now, arguing about what it has to do, while the argument is still free. The alternative is to find out during the build, when the machine you forgot to specify turns out to be the only part that mattered.
Ready to get started?
