Most advice on how to build a website for your accounting firm starts with design. That is the wrong end of the process. A firm's site is a piece of commercial infrastructure, and the design is the last layer applied to a set of decisions made earlier: who the site is for, what it needs to produce, how pages map to what people actually type into Google, and what happens after someone fills in a form.

Get those decisions right and the design work becomes straightforward, because every page has a job. Get them wrong and you end up with an attractive brochure that ranks for nothing useful and sends four or five poorly matched enquiries a month to an inbox nobody monitors closely.

This guide sets out the sequence we use when building sites for accounting and CPA practices: commercial brief, positioning, architecture, page structure, proof, enquiry handling, then technical launch. It also covers the failure points we see most often when a firm inherits a site built without that sequence. You can apply the whole thing yourself or use it to brief an agency properly.

Before you write a word of copy, write down what the site is for. Not "a modern online presence". Something measurable.

Common objectives for a practice of 2 to 20 staff include reducing dependence on referrals, making one specialist service easier to sell at a higher fee, giving paid campaigns a destination that converts, or supporting recruitment during a hiring push. Each of those produces a different site. A recruitment-led site needs careers content and team depth. An acquisition-led site needs sector pages and search coverage.

Define the valuable action

Agree what counts as a result. For most firms it is a booked consultation with a business that fits the target profile. For others it might be a completed onboarding form, a downloaded guide from a specific sector, or an enquiry about a defined piece of work such as a first-year company account or a payroll takeover.

Write the number down too. If the practice needs eight new clients a year at a given average fee, and roughly one in four consultations converts, the site needs to produce something in the region of thirty-two qualified consultations a year. That figure tells you how much traffic and how much conversion work is realistic, and it stops the project drifting into an aesthetic exercise.

Agree who owns follow-up

A site that produces enquiries nobody answers within a day is a waste of money. Decide before launch who receives enquiries, how fast they respond, and what happens to the ones that are not a fit. That decision shapes your forms, which we come back to later.

The single biggest determinant of enquiry quality is how specifically the site describes who you serve. A prospect comparing three or four firms in one sitting is scanning for evidence that you already handle situations like theirs. Language such as trusted partner, tailored solutions or proactive advice gives them nothing to compare, so they default to price or proximity.

What specificity looks like in practice

Instead of "we work with small businesses", say which ones and what you do for them. A firm that handles construction subcontractors should talk about CIS deductions, gross payment status and cash retention. A firm that serves owner-managed dental practices should reference associate contracts, NHS pension complications and incorporation timing. That detail is invisible to everyone outside the niche and immediately credible to everyone inside it.

Answer four questions on the homepage

• What work do you do?
• Which businesses do you do it for?
• Why is this firm a safe choice?
• What is the next step?

Everything else on the homepage is secondary. Firm history, values statements and stock imagery can stay, lower down, but they do not answer any of the four.

Positioning is a business decision, not a copy decision

If the partners are not willing to describe a primary client type, the copywriter cannot invent one. Where a firm genuinely serves several distinct sectors, the answer is separate sector pages rather than a homepage that hedges. We cover that structure next.

Search intent mismatch is the quiet reason many accounting firm websites attract no useful traffic. Firms build pages around the words they use internally, such as accounting services, tax compliance or business advisory. Prospects search differently: accountant for a limited company in Norwich, bookkeeper for a construction firm, help with a late corporation tax return, VAT registration for an online seller.

The four page types

• Service pages. One page per service you actually want to sell. Year end accounts, VAT returns, payroll, self assessment, management accounts. Each written for someone who already knows they need that service.
• Sector pages. One page per client type you serve well. These carry the positioning detail and usually convert better than generic service pages because the reader recognises themselves.
• Location pages. Useful where you compete locally and can write something genuinely specific about the area and the businesses in it.
• Guides. Pages answering the questions prospects ask before they contact anyone. These attract earlier-stage traffic and demonstrate competence.

What to avoid

Do not generate forty near-identical pages by swapping a town or industry name. Search engines discount them, and any prospect who clicks two of them sees the same paragraph twice. If you cannot write four hundred genuinely different words about serving hospitality businesses in a given town, that page should not exist yet.

Plan the internal linking

Sector pages should link to the relevant service pages and to the guides that support them. Guides should link back to the service page that solves the problem being explained. This is how search engines understand which pages matter, and it is how readers move from a question to a conversation.

Once architecture is settled, each page needs a repeatable structure. The pattern that works for accounting firms is simple: state the problem in the reader's language, describe the work you do, show what the process looks like, place proof next to the point of decision, then offer a specific next step.

Put proof where the decision happens

Testimonials stacked on a separate page get very little traffic. A named client comment about a payroll takeover belongs on the payroll page, close to the enquiry form. Mini case notes work well: the situation, what you did, what changed. Numbers help where you have them and are permitted to publish them.

Process transparency reduces hesitation

Cautious buyers want to know what happens after they get in touch. A short outline of the first two weeks, covering the introductory call, information gathering, professional clearance, engagement letter and software setup, removes a lot of unspoken friction. It also filters out people who are not ready to move.

Specific enquiry routes beat one contact form

A single Contact Us button collects everything, including the enquiries you do not want. Separate routes work better: book a company year end review, discuss moving payroll across, ask about a late filing, request a quote for bookkeeping. The route someone chooses tells you what the conversation is about before you pick up the phone.

Ask only what you will use

Long forms suppress enquiries and short forms flood you with unqualified ones. Ask for the three or four fields that let you triage properly: business type, turnover band, current software, what prompted the enquiry. Anything you will not read at triage should not be on the form.

The technical layer decides whether the site loads quickly, whether search engines can read it, and whether you keep the rankings you already have.

Migration comes first

If you are replacing an existing site, inventory it before anything is deleted. List every page, its traffic, its rankings and any inbound links pointing at it. Map each old URL to a new one and set permanent redirects. Firms that skip this step routinely lose a third or more of their organic traffic overnight and spend months recovering it.

Build standards to insist on

• Fast loading on mobile connections, with images compressed and sized properly.
• Correct heading order on every page, one H1, then H2s and H3s in sequence.
• Labelled form fields, visible focus states and keyboard access.
• Colour contrast that passes accessibility checks, particularly on buttons.
• Structured data for the organisation, services and any FAQ blocks.
• An accurate sitemap and clean metadata on every page.

Test the plumbing before launch

Submit every form yourself. Confirm the enquiry lands in the right inbox or CRM, that the confirmation email sends, and that the analytics event fires. Check the site on a phone, a tablet and a desktop. Proofread anything regulatory, including your registered office, company number and professional body references, because errors there undermine the credibility everything else is trying to build.

Instrument it so you can improve it

Track form submissions, calls, and which pages preceded them. Without that you cannot tell whether a page is doing its job, and every future change becomes guesswork.

A well-built site converts the demand that reaches it. It does not create demand on its own. Firms that treat launch day as the finish line usually see a modest bump from existing referral traffic converting better, then a flat line.

Three things need to run alongside it

Search coverage takes months to build. Guides, sector pages and local pages accumulate rankings gradually, which is why the content plan should start at launch rather than after the first quiet quarter. One firm we work with, Wings Online Filings, moved from five to seven enquiries a month to fifteen to sixteen over roughly two and a half months once an SEO architecture and content system was running behind the site, and converted nine of those into clients.

Paid acquisition fills the gap while search builds. It also tests messaging quickly. If a sector page cannot convert paid traffic, it will not convert organic traffic either.

Follow-up decides how much of the demand you keep. Enquiries that sit unanswered for two days convert far worse than those answered in an hour. A CRM with automated acknowledgement, reminders and a defined sequence for people who are not ready yet recovers a meaningful share of enquiries that would otherwise disappear.

Sequence it sensibly

Build the site properly, launch it with measurement in place, then add acquisition and follow-up on top. Doing it in that order means every pound spent on traffic lands somewhere that can convert.

This is the order we work in when building a website for an accounting practice. Each step depends on the one before it.

Five failure points account for most of the accounting firm websites that produce nothing.