Very few organisations here are buying web design for a first website. Almost every brief is a replacement: a site that has accumulated pages for a decade, a content system nobody enjoys using, a structure that made sense when the business was smaller. Redesigns are a different purchase from new builds, and the parts that go wrong are almost always decided before any visual work begins.
Most Sheffield projects are redesigns rather than new sites
That matters because a redesign carries inheritance. There is existing content of wildly varying quality, an audience with habits, pages that earn enquiries for reasons nobody has examined, and internal attachments to sections that no longer serve anyone. A web design studio that treats the job as a blank sheet will produce something attractive that loses things you did not know you had.
The first question is therefore not what the new site should look like. It is what the current one actually does: which pages people arrive on, which ones they use before enquiring, which ones exist because somebody insisted in a meeting years ago. That information exists already, and gathering it before the brief costs nothing.
Audit before you brief anyone
Export a list of every page and sit with it for an afternoon. Mark each one keep, merge, rewrite or retire. Most organisations discover that a large share of the site is retirable, and the exercise immediately makes every quote cheaper and more accurate, because a studio quoting against a real inventory is not guessing.
Do the same with the assets. Which photographs do you own outright, which fonts are licensed and for what, where does the logo exist in an editable form. Redesigns stall on exactly these questions at exactly the wrong moment, usually the week before launch when someone needs a high resolution version of something nobody can find.
Deciding what survives, and saying so in the brief
Name the parts that cannot change and the parts that must. An identity in mid life, a booking system that stays, a content system the team refuses to give up, a page that generates enquiries and should not be touched. Equally, name the failures: the service list nobody can navigate, the mobile experience, the forms.
Then write down what happens to retired pages. Old addresses have to be mapped to new ones before the switch rather than after somebody notices the enquiries have stopped. That mapping is a real line of work and it belongs in the proposal as a named item. Its absence is the most common self inflicted injury in a redesign, and it is entirely avoidable.
A brochure site is a legitimate web design brief
Buyers sometimes apologise for wanting straightforward web design that explains what the business does and invites contact. There is nothing to apologise for. For a manufacturer, a consultancy, a practice or a trade supplier, a well structured set of pages with credible photography and a form that works is exactly the right purchase, and it is the brief most reliably ruined by being oversold.
Watch for scope that arrives uninvited: a blog nobody will write, a customer portal nobody asked for, a personalisation layer with no content to personalise. Each of those adds cost to the build and more cost to the years afterwards. A studio that talks you out of features is usually the one to hire.
Staged rollout or a single switch
There are two ways to land new web design across an existing site. One is a single launch where everything changes at once, which is cleaner to explain and riskier to execute because every problem appears on the same morning. The other is staged: the new system applies to key templates first, the rest follow, and the site lives in two states for a while.
Staging suits organisations with a large content estate and limited internal capacity. Its failure mode is drift, where the first phase launches, attention moves on, and the half converted site becomes permanent. If you stage it, put the later phases in the contract with dates rather than intentions. If you switch in one go, insist on a written acceptance checklist walked through with the studio while the final invoice is still open, since testing after final payment turns every finding into a favour.
Web design proposals that can actually be compared
Give every studio the same inventory from your audit, the same answer about who writes the copy, the same named approver and the same position on photography. Ask each for a template list rather than a page count, named stages with exit criteria, the revision policy including what counts as a new round, the redirect mapping as a visible item, and what arrives at handover.
Handover means editable source files, the component library or documented rules behind it, font licences in your name, and administrative access to every account created for you. Ask what the support arrangement covers afterwards, priced separately from the build so you can decline it without losing your site. Collecting all of that from three suppliers takes a while, and you can instead set out the requirement once and receive structured proposals written to the same brief.
If the conversation keeps turning to systems, integrations and ongoing development rather than structure and appearance, the requirement may belong with firms that build and maintain web applications. If what the site really lacks is visibility rather than design, look first at search specialists before commissioning a rebuild that changes nothing about why nobody arrives. For the wider shape of this category, begin at the overview of website design agencies.