Learning how to build an online marketing system for accountants is mostly a sequencing problem. Almost every firm owner already knows the components exist. The website, the Google listing, the LinkedIn posts, the occasional ad campaign, the CRM nobody quite finished configuring. What is usually missing is the connective tissue that turns those separate assets into one chain where each part hands work to the next.

A system is different from a set of activities. Activities produce output when you push them. A system produces output on a schedule because the inputs, the routing, and the follow-up are defined in advance. When a firm tells us marketing has not worked, we usually find four of the five components present and one absent. The absent one is normally follow-up, and the whole chain fails at that point regardless of how good the website looks.

This guide sets out the five components, the order they should be built in, the sequencing over a first ninety days, the mistakes that appear most often in practice, and the measurements that tell you whether what you have built is working. It is written for firms doing compliance and advisory work for UK and US small businesses, with a team small enough that the owner still sees most new enquiries personally.

Every functioning acquisition system in a small accounting firm contains the same five parts. You can build them in different technologies and at different budget levels, but if one is missing the others carry a load they were never designed for.

Positioning

The decision about who the firm is for and what problem it solves for that group. This is the input to everything downstream. Without it, your pages describe services rather than situations, and your ads compete against every other firm in the country on the phrase "accountant near me".

Search architecture

The set of pages that capture people already looking. Service pages, niche pages, location pages if you serve a defined area, and resource content that answers the questions your target client types into Google before they are ready to speak to anyone.

Paid acquisition

Google and Meta campaigns that buy attention while search rankings build. Paid produces enquiries in weeks. Search produces them in months. Running both means you have cash flow from acquisition during the period when the slower asset is maturing.

Follow-up automation

The CRM layer that catches every enquiry, responds within minutes, chases the ones who do not book, and stops leads sitting in an inbox during a filing week. This is the component most commonly missing and the one with the largest effect on conversion.

Client onboarding

Digital engagement letters, AML checks, data collection, and the handoff into your practice software. Growth that increases admin load faster than it increases fee income is not growth you can sustain, so this component determines your ceiling.

The chain runs left to right. Traffic without follow-up leaks. Follow-up without onboarding creates a bottleneck at the point where you actually get paid.

Positioning is the cheapest decision on this list and the one that changes the economics of everything else. A generalist firm bidding on general accountancy terms competes with national brands, franchise networks, and every local practice within twenty miles. A firm positioned around a defined client type competes with a much smaller field, converts at a higher rate, and can charge on the basis of specific knowledge rather than hourly effort.

How to choose the segment

Start with your existing portfolio rather than with market research. Pull the last two years of client data and sort it four ways:

• Which sectors produced the highest realised fee per hour of work
• Which clients generated advisory work beyond the compliance engagement
• Which clients renewed without fee negotiation
• Which work your team completed without the partner having to intervene

The overlap between those four lists is usually two or three sectors. That is your candidate set.

Then check demand

A segment you enjoy serving is worthless if nobody in it is searching. Cross-check your candidate sectors against two data sources. Search volume tells you how many people look for sector-specific accounting help each month. Business registry data, Companies House in the UK, tells you how many entities of that type exist in your target geography and how many were incorporated recently. New incorporations matter because those businesses are actively choosing an accountant rather than staying with an incumbent.

A segment with meaningful search volume and a steady incorporation rate is a viable position. A segment with strong margins and no search demand still works, but it needs an outbound approach rather than a search approach, which is a different build.

Most accounting firm websites are organised around the firm. Home, About, Services, Blog, Contact. That structure reflects the internal org chart and captures almost no search intent, because nobody types "about our team" into Google.

Organise around what people actually search instead. Three page types do the work.

Service pages

One page per service you want to sell, written to the search terms that service attracts. A page targeting "limited company accountant" needs to answer what the service includes, who it suits, how the process runs, and what happens after you engage. One page trying to cover eight services ranks for none of them.

Niche pages

One page per client type in your chosen position. "Accountants for dental practices" is a different page from "accountants for construction firms", with different examples, different compliance concerns, and different case evidence. These pages convert far better than general service pages because the reader recognises their own situation in the first paragraph.

Resource pages

Guides that answer the questions your target segment asks before they are shopping. Someone searching how the Making Tax Digital for Income Tax rules affect them is not ready to switch accountants today, but they will remember where they got the clearest answer. Resource content also earns links and gives your service pages something internal to point at.

What to check before writing anything

• Does each page target one primary search term and answer it fully
• Does the page load quickly on a phone, since most first visits are mobile
• Is there one obvious action on every page, and is it the same action
• Does the enquiry form ask enough to qualify without asking so much that people abandon it

Prads at Wings Online Filings went from five to seven enquiries a month to fifteen to sixteen in the first two and a half months after his site was rebuilt around this structure, and converted nine of them into clients.

Search architecture is the better long-term asset and the slower one. Expect three to six months before a new page structure produces reliable organic enquiries, longer in competitive urban markets. Paid acquisition covers that gap.

Google versus Meta

The two platforms do different jobs and should be budgeted separately.

Getting Google campaigns right

Bid on service and niche terms, not on your own brand name and not on broad terms like "accountant". Send every ad group to a matching landing page rather than to the homepage. Add negative keywords for job seekers, students, software queries, and free-tool searches, or you will pay for clicks from people who will never become clients.

Getting Meta campaigns right

Meta needs a reason for someone to stop scrolling, and "we are a friendly local accountant" is not one. A defined entry offer works better. A fixed-scope review, a compliance readiness check for a specific sector, or a switching service with a clear process. The offer needs a defined scope and a defined outcome so the prospect knows exactly what they are agreeing to.

Chris at Thomas Emlyn Ltd runs an outbound engine built on registry data alongside his paid work, producing thirty to fifty cold leads a month and five to ten booked meetings from them. The principle holds across channels: define who you are targeting narrowly, then give them a specific reason to respond.

This is where most firms lose the money they have already spent. Enquiries arrive, land in a shared inbox, and sit there while the team works through a filing deadline. By the time someone replies, the prospect has spoken to two other firms.

What the follow-up layer has to do

• Capture every enquiry from every source into one place, including form fills, phone calls, and chat
• Send an acknowledgement within minutes, not hours
• Offer a booking link immediately so the prospect can self-schedule
• Chase non-responders on a defined sequence over the following two weeks
• Alert a human when a high-value enquiry arrives
• Record source data so you know which channel produced which client

Qualification before it reaches your calendar

Volume is only useful if the meetings are worth taking. A qualification layer on the enquiry form, asking about turnover band, entity type, current arrangement, and reason for looking, lets you route enquiries rather than treat them all identically. Annabel uses an AI qualification step on her form so her calendar only carries opportunities worth the partner's time. Everything below the threshold gets a different path, usually a self-serve option or a junior team member.

Then close the loop into onboarding

When a prospect says yes, the system should produce an engagement letter, run identity and AML checks, collect the information you need, and open the client record in your practice software without anyone rekeying data. Firms that skip this find that every ten new clients adds a day a week of administration, which caps growth at whatever the owner can personally absorb.

Traffic and impressions tell you very little. Six numbers tell you whether the system is functioning and where it is failing.

The six numbers

1. Enquiries by source. Split organic, paid, referral, and direct. Without this you cannot tell which spend to increase.
2. Median response time. Measured from enquiry received to first human or automated contact. Anything over an hour is costing you conversions.
3. Enquiry to booked meeting rate. If this is low, the problem is follow-up or qualification, not traffic.
4. Meeting to client rate. If this is low, either the wrong people are booking or the sales conversation needs work.
5. Average annual fee of new clients. Volume without value means you have positioned too broadly.
6. Time from acceptance to first work delivered. The onboarding measure. Rising numbers here mean the admin bottleneck is forming.

Reviewing them

Look at these monthly, not weekly. Marketing data at small firm volumes is noisy, and weekly review encourages changes that undo work before it has had a chance to show effect. Look at the trend across three months and change one variable at a time.

One caveat on attribution. People find you through several touchpoints before they enquire. Someone might read a guide, see an ad three weeks later, then search your firm by name and arrive direct. Last-click attribution will credit that to direct traffic and undervalue both the guide and the ad. Ask on the enquiry form how they heard about you, and treat the self-reported answer as a useful cross-check against the platform data.

A realistic sequence for a firm starting from a general website and no structured acquisition. Each phase depends on the one before it.

These are the failures we see most often when reviewing an existing setup.