Most small-brand briefs describe a website that costs three times the budget. The real failure is not saying so. Everyone politely proceeds, the client gets every page they asked for at the quality the money allows, and the pages that decide the sale end up no better than the pages nobody visits.
Count templates, not pages
A site is not a stack of pages, it is a small set of templates repeated. Forty pages is usually five templates: home, product, collection, editorial, contact. Once you see it that way the budget question changes shape. A home page is seen once per visitor. A product template is seen four hundred times a month. Spending the same on both is a decision, and usually the wrong one.
One template built properly beats five built approximately, because the good one is the room people are actually standing in.
What never gets cut
- Positioning above the fold. One sentence on the home page saying what this is and who it is for. It costs nothing to write and it is the highest leverage element on the site. A visitor who cannot answer "what is this" in three seconds never reaches the beautiful section further down.
- The product or service template. Where money is decided: photography, price, stock, shipping, returns, the maker's note, the reasons to believe. If the budget allows exactly one page to be genuinely good, it is this one, not the home page.
- Contact and how to buy. The number of small-brand sites running a broken form, an unmonitored inbox or no phone number is genuinely startling. Test it monthly. A dead form is a leak that never shows up in analytics.
- The boring layer. Speed, mobile, schema, analytics, a search that works. None of it is visible in a design review and all of it is visible in revenue.
What to cut first
- Bespoke one-off page designs. The About page can use the editorial template. No customer has ever left because a page shared a layout.
- Animation past the useful minimum. A tasteful fade is fine. A scroll-driven sequence is a week of somebody's life that gets switched off on mobile anyway.
- Filtering and faceted search under about fifty products. With twenty products, filters solve a problem you do not have. A well-ordered grid is faster, cheaper and easier to merchandise.
- Second languages on day one. They halve the content budget and double the maintenance before you know which market responds. Build so a language can be added, then wait for evidence.
- The journal, at launch. Build the template, publish later. Three thin posts and a dead date read worse than no journal at all.
Phase it, and mean it
Phase one is the selling core: positioning, product template, checkout, contact, the boring layer. That is the shape of the four to six week build we run, and at the end of it the site is live and taking money rather than waiting for the rest of the plan.
Phase two lands three months later, funded partly by what phase one earned and directed by what the data showed, which is never what the launch meeting predicted. In our experience about a third of the phase two wish list quietly disappears, because the traffic proved nobody wanted it. That third is the money the phased approach saves you.
This only works if phase one is genuinely finished rather than a rehearsal. A stub with placeholder copy and one photograph is not a phase, it is a debt, and it damages the brand for every one of the ninety days it stays up. Cutting scope means shipping fewer things at full quality, not shipping everything at half.
The uncomfortable version of this conversation belongs on the first call, when we tell a founder their budget buys three excellent templates instead of eight adequate ones. It makes for a harder call and a better project. We have yet to meet a client who came back a year later wishing they had spread the money thinner.