Fenwick & Marsh Accountants Group

• You identified five goals for the next six to twelve months: modernising systems and technology, working more efficiently, scaling without adding headcount, improving quality and consistency, and selling more to existing clients. Those goals reinforce each other, which is the good news. The complication is that none of them can be achieved independently while your data and your client records sit across four different practice management systems. Unifying that foundation is the enabling move, not a parallel workstream.

• Key-person dependency is a real commercial risk, not just an operational one. Several partners hold client relationships personally, with limited cover, and you noted that a change of contact is a common reason clients leave. That is a retention problem as much as a people problem. The answer is not just hiring cover; it is making sure client context, service history and relationship notes live somewhere the whole team can access, not only in one person's memory or inbox.

• Your growth model is almost entirely referral and word of mouth, supported by local sponsorship. That is a reasonable base for a regional accountancy practice, and it clearly works well enough to have funded five acquisitions. The problem is that nothing is done to encourage referrals deliberately, to track where they come from, or to protect that channel if a partner leaves and takes a network with them. Formalising how referrals work is a higher priority than adding new marketing channels.

• You have a marketing manager and access to freelancers, which means content capacity exists, but there is no content plan and no formal search engine optimisation work in place. For a professional-services firm where clients search for specific problems ("accountant for contractor", "R&D tax relief specialist", "Xero accountant near me"), not appearing in those searches means the practice is invisible to prospects who are already looking. That is an addressable gap.

• Your paid search activity is described as occasional, with nothing structured behind it. That is worth noting because occasional paid search with no plan is unlikely to produce useful data or reliable enquiries. It is either worth running properly, with a clear target cost per enquiry and a defined service to promote, or pausing until the strategic groundwork is in place. Running it at low intensity in between serves neither purpose.

• The scoring across these ten elements reflects what the discovery confirmed: your capability score for this element sits well below the importance you place on it. You rate this area as highly important (80 out of 100) but your current capability scores are considerably lower. That gap is where the work is concentrated, and it is why the Digital Transformation Audit produces a prioritised roadmap rather than a list of equally-weighted recommendations, so that the effort goes where it changes something meaningful first.

Where time is being lost

• Five acquisitions have left you running four different practice management systems. Each system holds client records, workflow tasks and billing data in a different format, which means staff who move between offices, or who cover for colleagues, have to relearn how things are stored rather than just doing the work. This is not a technology problem you can solve with a new tool on top of the old ones; it is a structural drag that compounds every time you onboard a new starter or open a new office.

• Data entry and re-keying sit at the top of your list of time-consuming processes, which is exactly what you would expect when four systems do not talk to one another. Work that should flow automatically, client details, VAT figures, payroll runs, gets typed in more than once. That is direct cost in staff hours, and it introduces the risk of errors in regulated work.

• Reporting and spreadsheets are the second big time sink. When the underlying data lives in four systems, anyone who needs a practice-wide view has to pull exports, reconcile formats and build the picture by hand. Partners are likely doing some of this themselves, which is an expensive use of their time and produces a view that is already out of date by the time it lands.

• Invoicing and payment chasing appearing on the same list suggests your billing process still has manual steps, likely because client records and billing are managed separately across the different systems. In a fixed-fee monthly model, invoicing should largely run itself. If it does not, that points to a workflow gap rather than a volume problem.

• With 1.5 people covering marketing across nine offices, there is no margin for informal briefing or approval processes. If a campaign idea has to pass through a partner before anything moves, it will move slowly. You do not need a large marketing team to fix this; you need a short written brief template and a clear sign-off chain so that the one and a half people you have can operate without bottlenecks and your freelancers can produce work that hits the mark first time.

Process and structure before the fourth office opens

• Adding a fourth office on the current operational base will replicate the problem, not solve it. Every inefficiency you have today, the re-keying, the manual reporting, the slow onboarding, will appear in the new location too, and you will have a smaller local team to absorb it. The practical first step is to identify which one process causes the most delay right now and fix that specifically, rather than trying to standardise everything before opening day.

• Onboarding new accountants is already slow and competitive. Without a documented onboarding pack that works across all offices and all four systems, each new hire's productivity ramp depends on whoever happened to sit next to them. That inconsistency is a hidden cost in a recruitment market where qualified candidates have choices.

• Recruiting and retaining qualified accountants is your stated pressure point, not finding new clients. That distinction matters for how you spend marketing resource. A proportion of your digital presence should be pointed at prospective employees, not just prospective clients. Your website and any content you produce should reflect what it is like to work at Fenwick & Marsh, not just what services you offer.

• Armadello would address the reporting problem directly: rather than building a spreadsheet view of the practice each month by stitching together exports from four systems, it would bring your channel and commercial performance into one place so partners have a single, current picture. That is not a luxury for a nine-office firm; it is what makes the difference between decisions taken on real data and decisions taken on instinct.

The problem is that your current approach does not reflect that. New business still arrives mainly through word of mouth and referrals, but there is no formal mechanism behind either, and your cross-sell activity is largely left to chance.

Who you are winning and why that matters

• Your ideal client profile is well defined, which puts you ahead of many firms: owner-managed businesses turning over £250,000 to £10 million, within half an hour of an office, on a fixed monthly fee, wanting a named partner they can call. That specificity is useful because it tells you exactly who to protect, who to pursue more of, and who probably does not belong in the portfolio at all.

• The 70/20/10 revenue split (recurring compliance, personal tax, advisory and transaction work) is not unusual for a firm at your stage, but it does mean your most profitable and differentiated work, the advisory and transaction piece, is currently the smallest slice. That is worth noting because it is also the service line clients are least likely to find through a competitor without being prompted.

• A monthly cancellation rate of roughly 0.7% is broadly what you would expect once you strip out business closures and retirements, but it is worth tracking separately so you can distinguish genuine dissatisfaction from the unavoidable. At the moment you cannot easily do that because you have no structured measure of client satisfaction in place.

The referral and review gap

• Most of your new business comes from word of mouth and referrals. That is not a weakness in itself; in professional services it is how the best firms grow. The problem is that you have no formal structure around it, which means it relies entirely on clients spontaneously recommending you rather than being given a reason and a prompt to do so.

• Your Google Reviews are reasonably strong, but they arrive informally. A firm with clients staying nine years or more has a large pool of satisfied people who would very likely leave a review or make a referral if someone simply asked them at the right moment, after a tax saving, at year-end, after a transaction completes. Putting a simple, repeatable process behind that ask would compound over time.

• A structured referral programme does not need to be complicated. The starting point is agreeing internally when you ask, how you ask, and who is responsible for following up. Once that is consistent, you can consider whether a light incentive structure adds anything, though many professional services firms find that a timely, personal request outperforms anything more formal.

Cross-sell and upsell

• This is the most commercially significant gap in your current approach. Clients using only compliance services are almost certainly not aware of what else you could do for them, and there is no systematic process for telling them. At an average lifetime value of £38,000, moving even a proportion of your compliance-only clients onto an additional service line would represent a meaningful revenue increase without any new client acquisition cost.

• The opportunity is particularly visible in the advisory and transaction line. A business owner going through a growth phase, a refinancing, or a potential sale is often already your compliance client. That conversation should not happen by accident; it should be triggered by something in their account or annual review cycle.

• A simple trigger-based model, where account managers or partners are prompted to have a specific conversation based on what is happening in a client's business, would be a practical first step. This does not require sophisticated technology; it requires a shared view of each client's current services against what they are eligible for, and an agreed point in the relationship where that conversation happens.

• Armadello would give you a consolidated view of client service uptake alongside revenue, so you could see at a glance which clients hold only one service line and what the aggregate opportunity looks like. That is the kind of picture that currently requires someone to pull data from multiple systems by hand.

Retention and satisfaction measurement

• Your account management, email newsletters and check-in calls are all sensible retention tools, but without any structured measure of client satisfaction you have no early warning when a relationship is deteriorating. Clients in professional services rarely leave suddenly; they drift, often over months, and the signals are there if you are looking for them.

• You are already considering Net Promoter Score. That is worth doing, not because the number itself is the point, but because it gives you a consistent, comparable signal across your client base and a prompt to have a real conversation with anyone who scores you poorly. Starting with a short annual survey to your compliance clients, timed around year-end, would give you data without adding significant overhead.

• The nine-year average relationship is a strong retention story, but it is a retrospective measure. What you do not yet have is a forward-looking one: which clients are at risk this year, and why. Tracking satisfaction formally, even in a lightweight way, starts to give you that.

Customer communications

• Every office sending different letters, using different portals and running different chase timelines means a client who moves, refers a colleague, or simply compares notes with another Fenwick and Marsh client gets a different experience depending on which door they walked in through. The rebrand made this worse by changing something long-standing clients recognised without explaining to them why or what had changed. The first thing to fix is agreeing one set of templates, one tone and one client-journey sequence across all offices, not because consistency is a branding ideal, but because inconsistency costs you client confidence.

• Your email list sits at a basic customer-versus-prospect split, which is a reasonable starting point but leaves a lot of relevance on the table. A firm of your complexity has clients with very different needs: owner-managed businesses worrying about exit planning, sole traders approaching year-end, SMEs navigating corporation tax changes. The tax-year-end guides and budget-day summaries you already produce are genuinely useful content, but they land with the same weight on every inbox. Segmenting by client type, service line or business stage before the next tax season would let you send the right guide to the right client rather than broadcasting everything to everyone.

• The one partner writing well on LinkedIn about exit planning is the clearest signal that you have expert opinion worth sharing. The problem is that it sits outside any coordinated effort, so it does not build on your other communications, is not promoted to the email list, and stops entirely when that partner is busy. A simple content calendar shared between your marketing manager and that partner, with a monthly review rather than a standing meeting, would give the activity enough structure to be reliable without demanding time nobody has.

• Most client contact being you chasing records at year-end is understandable operationally, but it means the relationship is largely transactional and reactive on your side. There is an opportunity to shift that balance slightly by building two or three planned touchpoints through the year that are useful rather than administrative: a mid-year check-in prompt, a relevant regulatory summary, a short tax-planning note ahead of year-end. None of these require significant production effort, but they change the character of the relationship and reduce the surprise clients feel when the year-end chase arrives.

• There is no central visibility of how quickly your team responds to client enquiries, whether by email, phone or portal message. Portal messages going straight to the job holder sounds efficient but means no one can see whether they have been answered, how long they took, or whether they fell into a gap when someone was on leave. Without that visibility you cannot manage response quality, and clients notice the difference between offices. Even a shared inbox log or a simple helpdesk tool, reviewed weekly by office administrators, would give you the data you need to set a consistent standard.

• Anicca's content team can help turn the guides and expert opinion you already produce into a more regular and better-targeted programme, working within your marketing manager's existing capacity rather than replacing it. The goal is a structured pipeline that does not depend on whoever has a spare hour that week.

Internal and team communications

• The sign-off process you described, where the managing partner must approve anything above roughly £1,000 or touching specific services or regulation, is appropriate for a regulated professional-services firm, but it only works if the people preparing content for approval understand the boundary clearly. Routine social posts not needing sign-off is sensible, but with five offices and no coordinated tone-of-voice guide, the definition of "routine" probably varies. A one-page internal guide covering what requires sign-off, what does not, and what the tone of the firm looks like in practice would reduce unnecessary back-and-forth and protect the managing partner's time.

• The preference for a named contact and a monthly written summary rather than calls is noted and should shape how any external marketing work with Anicca is structured. Written summaries work well when they are built around the metrics that matter to you, and Armadello would make that straightforward by bringing your channel activity, email performance and enquiry data into one place so the monthly summary reflects real commercial output rather than a collection of platform statistics assembled by hand.

• With five acquisitions on four different practice management systems, internal communication about client work is fragmented by design rather than by accident. That is primarily a systems-integration challenge rather than a communications one, and it sits outside the scope of this roadmap, but it affects communications directly: if staff in one office cannot see the history of a client managed from another, they cannot communicate consistently with that client. Any work you do on client-facing communications consistency needs to acknowledge that constraint and plan around it until the underlying systems problem is resolved.

What you have to work with

• You have a marketing manager and access to freelancers, which is a real foundation. The problem is not resource, it is direction. Without a content plan and a house style, that resource gets spent reactively, filling gaps when someone notices them rather than building something that compounds over time.

• Canva and ChatGPT are already in use, which tells you the appetite for faster, more designed content is there. The difficulty is that without guidelines, Canva becomes a different tool in each office, and the output drifts further from a unified identity every month.

• A logo and a letterhead template are not brand guidelines. They tell someone what the firm looks like on headed paper, but they do not tell a fee earner in the Bristol office how to write a proposal introduction, what tone a LinkedIn post should take, or whether to use the word "client" or "customer". That gap is why nine offices have produced nine slightly different versions of the same firm.

• Blog output tied to legislation changes is not a content strategy, even though the instinct behind it is sound. Legislation-triggered posts are useful, but they arrive in clusters when the rules change and then disappear for months. A competitor who publishes on the same topics consistently, not just reactively, will own more of the search landscape for those terms over a longer period.

• The LinkedIn activity from one partner is the most visible organic channel the firm currently has. One person carrying that load is fragile. If that partner has a busy month, the channel goes quiet. The opportunity is to use what that partner has already built, the audience, the posting habit, the credibility, as proof of concept for a wider programme involving two or three more voices across the group.

• You said you are somewhat comfortable providing spokespeople, testimonials and case-study content. That is the right instinct to act on. A professional-services firm's most credible content is a client describing a real problem that was solved. Even a handful of properly written case studies, cleared through your professional indemnity insurer and written without disclosing anything confidential, would do more for new-business conversations than a year of generic blog posts.

What to build next

• The first practical step is a content plan that names five or six topics the firm genuinely wants to be known for. These are not broad themes like "tax" or "accounts". They are the specific questions your best prospective clients are already searching for and asking at first meetings. A structured content plan built around those topics gives your marketing manager and your freelancers a brief to work from rather than a blank page.

• Behind that plan you need one short style guide: the words you use and the words you avoid, the tone in a social post versus the tone in a newsletter, who approves what before it goes out. It does not need to be a fifty-page document. A clear two-page reference that anyone in any office can open in thirty seconds is enough to stop the drift.

• The monthly Mailchimp newsletter is an underused asset. A monthly send is a reasonable frequency for a professional audience, but the content of those newsletters should map to the same plan that drives the blog and social output. At the moment the newsletter, the blog and LinkedIn are three separate efforts. Aligning them around the same monthly themes means one piece of research or one idea produces content across all three, which is a much better return on the time your marketing manager spends.

• On the AI drafting question: several staff are already using ChatGPT to draft letters and content without a sanctioned policy in place. From a content production standpoint, the practical step right now is to keep a simple log of what is AI-assisted and what is written from scratch, so that when disclosure expectations firm up, you are not working backwards through months of published material. The EU AI Act's transparency duty under Article 50 requires AI-generated content to be clearly disclosed, with that obligation applying from August 2026. That applies directly where your content reaches EU-based readers; if your audience is entirely UK-based, treat it as the direction UK expectations are heading. Either way, a simple record now is far less effort than a compliance review later. The broader question of what your firm should and should not do with AI tools, and how to govern that across nine offices, is work that sits squarely within Anicca's AI Adoption Roadmap rather than here.

• The proposals and engagement letters flowing constantly across the group are content too, and they are arguably the most read content the firm produces. They carry the firm's name into every client relationship. Bringing those into the same style framework as the external marketing content is worth including in scope when you build the guidelines, even if the legal and compliance review sits with your partners rather than the marketing function.

Where your new business actually comes from

• Referrals and acquisition-driven client books account for roughly three quarters of new business between them. That is entirely normal for a regional accountancy group, and it is not a problem in itself. The problem is that neither channel has any formal structure around it: no agreed referral process, no tracking of which relationships or acquisition sources produce the best long-term clients, and no way to see whether that 55% referral share is growing, shrinking or simply staying flat by habit.

• Organic and local search delivers around 15% of new business by your own estimate, and the Armadello audit data supports that: organic search accounts for 41% of website sessions over the last twelve months. That is your most measurable acquisition channel and the one where you have already made some investment, yet it generates only 340 enquiries from 84,000 sessions, a conversion rate of 0.4%. That gap between traffic and enquiries is the most immediate commercial problem in your channel picture, and it sits squarely in the website and conversion layer rather than in search visibility itself.

• Your estimated cost of winning a new business client through organic routes is around £350. Without formal tracking, that number is almost certainly approximate, and it may be masking real variation: some service lines or locations will be acquiring clients more cheaply than others, and you have no way to see which. Agreeing a cost-per-enquiry target and measuring against it consistently would turn that estimate into something you can actually manage.

LinkedIn and Facebook

• You post on both LinkedIn and Facebook with one marketing manager and occasional design support, and no paid social spend at all. That is a thin resource for two platforms. LinkedIn is the correct priority for an accountancy firm targeting business owners and finance directors in the Midlands: it is where your prospective clients are making professional decisions. Facebook has some value for personal tax and smaller business audiences, but spreading effort across both with limited resource means neither is done particularly well.

• Organic social accounts for 9% of sessions in the audit data, which is a reasonable return for a purely organic effort, but social traffic is the least likely of all your channels to convert directly into an enquiry. Its real job for a firm like yours is visibility and credibility with people who are already in your orbit, not cold acquisition. Measuring it by enquiry volume alone will always make it look weak; measuring it by whether it reinforces the referral relationships that drive 55% of your business is a more honest test.

• There is a specific opportunity here tied to AI search. You mentioned that clients are increasingly arriving after asking ChatGPT a tax question first, sometimes with a confident but partially wrong answer. Publishing clear, authoritative content on Midlands-relevant tax and accounting questions, and doing it systematically rather than ad hoc, is the way to get your firm cited by those tools and to position your partners as the people who give the correct answer. That is primarily an SEO and content programme, but LinkedIn amplification of the same content extends its reach to exactly the professional audience who will share it onward. Anicca's content and SEO work is built around this kind of structured, compliance-aware publishing for regulated professional services firms.

The measurement gap

• There are no formal channel targets. Enquiry volume is tracked loosely, but there is no agreed cost-per-enquiry figure, no channel-level comparison, and no single view that brings traffic, enquiries and the cost of generating them together. That means when your marketing manager makes a decision about where to spend time this week, it is based on instinct or availability rather than evidence.

• Armadello would resolve this directly. It would pull your GA4 data, channel traffic and enquiry numbers into one place so you can see, without manual assembly, which channel is sending traffic that actually converts and which is sending volume that goes nowhere. For a firm at your stage, that single view is worth more than any new campaign, because it tells you where the next unit of effort belongs before you commit it.

• A small, tightly scoped paid search test targeting high-intent local queries, for example "accountant Midlands" or specific service terms, is worth considering once your conversion tracking is properly in place and you have a reliable cost-per-enquiry baseline to measure against. Without that baseline, any paid spend is guesswork. This is not a paid media strategy; it is one optional test to establish whether the channel earns its place, and it should only be considered after the measurement foundations are solid.

• Four practice management systems, each acquired through a different deal, hold your client records, billing history, workflow and capacity data in separate silos. Nothing flows between them automatically. When a partner needs a picture of capacity across the group, someone has to open four systems and reconcile the numbers by hand. When a client moves between offices or entities, their record has to be re-keyed rather than transferred. This is where your integration problem is most acute, and it has no quick fix, but it does have a logical starting point: map exactly which data each system holds and identify the one or two fields - client status, billing stage, contact detail - that are re-entered most often. Joining those specific data points first, even through a lightweight connector tool rather than a full migration, removes a disproportionate amount of the daily manual work.

• Xero is the one part of your estate that already connects cleanly to other tools, and that is genuinely useful as a foundation. The practice finance data flowing in and out of Xero is more reliable than the data sitting in any of the four practice management systems, precisely because it has working connections. Any integration plan should treat Xero as the anchor and ask which practice management system connects to it most naturally. That system becomes the logical starting point for a phased consolidation, rather than trying to converge all four at once.

• Your website, marketing email platform and whatever you use to capture and hold enquiries are most likely not connected to any of the practice management systems. That means a prospect who fills in a contact form, books a call, or replies to a newsletter has to be manually added to the relevant practice system by whoever picks up the enquiry. Beyond the time this takes, it means enquiry data is almost certainly incomplete, so you cannot see clearly how many leads came in, which services they asked about, or how quickly they were followed up. Connecting your website enquiry forms to a central contact record, even a simple shared CRM, would close that gap before any bigger integration work begins.

• You told us you already use ChatGPT, Microsoft Copilot and AI built into tools you already pay for, including Dext and Xero. Those embedded AI features in Dext and Xero are low-risk and worth getting staff comfortable with, because they operate on data you already hold in those platforms and carry no additional confidentiality exposure. The broader question of how the firm should approach AI tools, especially given the partners' concerns about ICAEW guidance and the quiet informal use of ChatGPT that is already happening, sits outside the scope of this roadmap. That is the territory of Anicca's separate AI Adoption Roadmap, which is the right place to set a clear usage position and close that gap properly.

• The absence of a single reporting view across your digital activity is a practical consequence of the disconnected systems. At the moment, any attempt to understand how many enquiries came from which channel, what happened to them, and whether they became clients requires pulling separate exports from your website, your email platform, and whichever practice system holds client records. Armadello would bring those data flows into one place - traffic, enquiry volume, channel source and, once the practice system connection is in place, client outcome - so leadership has one accurate picture rather than a reconciliation exercise at the end of every month.

• The realistic first move is not a system replacement. It is a connections audit: a structured session mapping every system in use across the group, what data each one holds, which ones expose any kind of connection point to other tools, and which manual hand-offs consume the most time. From that you can sequence the integration work by commercial impact rather than technical preference, starting with the connection that removes the most friction for the fewest pounds of effort.

• The four practice management systems each hold their own version of revenue, work in progress and lockup data. None of them talk to each other automatically, so every month someone has to extract figures from each, reconcile the differences, and reassemble them in a spreadsheet. A week of senior finance time going into that process every month is not a reporting problem, it is a capacity problem with a reporting cause.

• The five KPIs leadership tracks (revenue, cash flow, utilisation, and retention or churn) are all commercially important, but tracking them monthly rather than weekly means that a deteriorating utilisation figure or a client about to churn may already be two or three weeks old by the time it is visible. For a group that has grown through acquisition and still has integration work ahead, that lag creates real exposure.

• Data quality is described as "distributed across the team", which in practice means nobody owns it. When a figure looks wrong in the Excel consolidation, there is no single person whose job it is to investigate and correct it. That ambiguity slows decisions and quietly erodes trust in the numbers themselves, so partners end up relying on their own local view rather than the group picture.

• The digital marketing picture has a parallel version of the same problem. Traffic, enquiry and channel data sit in native platform dashboards (GA4 and whatever each tool reports natively), with no consolidated view. The 340 website enquiries recorded over the last 12 months are the most commercially significant digital outcome the business generates, and right now there is no single place that shows how those enquiries connect to channel, office, service line or eventual revenue.

• Armadello, Anicca's reporting product, is built precisely for this kind of situation. It pulls channel and commercial performance into one place automatically, so leadership sees traffic, enquiries, spend (if and when any is introduced) and return in a single view rather than stitching platform exports together. For Fenwick and Marsh, the immediate value would be connecting the GA4 data the business already generates to the commercial outcomes the finance director is trying to track, so the digital and financial pictures sit alongside each other rather than in separate reports compiled at different times by different people.

• The practice management integration problem sits above what any marketing reporting tool can solve on its own. Getting the four systems to feed a single source of truth for revenue and utilisation is primarily a systems consolidation question, and it will not be resolved until the post-acquisition integration work is complete. That said, the reporting discipline introduced through the digital side of the business, including agreed KPI definitions, a named data owner, and a weekly rather than monthly review frequency for the metrics that can move quickly, creates a model the finance function can extend once the systems are rationalised.

• One practical step that costs nothing immediately: agree a definition for each of the five leadership KPIs and write it down. Revenue, utilisation, retention and churn can all be calculated differently depending on which system or which partner you ask. Without a shared definition, the consolidation exercise produces a number that everyone interprets differently, which is part of why it takes so long to produce and is only partially trusted when it arrives.

Software and subscription costs

• The four practice management systems are the most visible symptom of the acquisition history. Each system was probably the right choice for the firm that ran it, but running them in parallel means you are paying for overlapping features, training staff on different workflows, and making it structurally impossible to get a single view of client profitability across the group. Consolidating onto one platform would reduce licence costs and, more importantly, would make the margin analysis described below actually possible.

• At £14,000 a month, software is already a material cost line. The risk of leaving it unreviewed is not just the waste on duplicate tools; it is that the consolidation work becomes harder and more expensive the longer each system is left to accumulate more client data and deeper staff habits.

• The AI-capable features inside Xero and Dext are tools you are already paying for. Before any separate AI spend is considered, it is worth understanding what those built-in features can actually do for data entry, bank reconciliation and document processing, since you have already funded them.

Marketing spend and what it is returning

• Your total marketing spend is somewhere between £5,000 and £10,000 a month, split between local paid search and trade or community event sponsorship. That is a meaningful commitment for a professional services firm, but there is no current way to connect that spend to the 340 enquiries recorded in the last twelve months, let alone to what those enquiries were worth in fees or gross margin once won.

• The paid search spend is the more measurable of the two. Cost per enquiry from that channel could be calculated relatively quickly if the campaign data were joined to the enquiry data from your website. At an average recurring client value of around £4,200 a year and a gross margin of roughly 58% on compliance work, even a rough cost-per-enquiry figure would tell you whether paid search is generating margin or just activity.

• Sponsorship is harder to attribute, and that is partly why it tends to survive budget reviews: you cannot prove it works, but you also cannot prove it does not. It may be doing important relationship work that never shows up in a conversion report. The honest answer is that you do not currently know, and without even a basic tracking approach, that will remain true.

• There is no content budget at all within the current spend, which means all organic content production falls to your marketing manager and freelancers working within an already stretched workload. Organic search currently brings in 41% of your website sessions, which makes it your single largest traffic source. Protecting and growing that without any dedicated budget is fragile.

The margin visibility problem

• You know your blended margins at product level: compliance at 55 to 60%, payroll bureau at around 35%, advisory at around 70%. What you do not know is the cost of winning clients into each of those lines. If most of your marketing spend is attracting compliance enquiries, that is defensible but not the highest-margin work you could be targeting. If it is attracting payroll bureau work, the return on that spend looks significantly weaker once the lower gross margin is factored in.

• The absence of a joined-up commercial view means budget decisions are made on instinct and habit rather than evidence. Spend on a channel that is generating low-margin work can continue for months before the pattern becomes obvious at a financial review. A channel that happens to be winning advisory clients could be bringing in your highest-margin work with no one realising, and no case to put more behind it.

• Bringing cost per enquiry, fee won, and gross margin by service line into one place is the practical step that changes this. Armadello would connect your channel and enquiry data to the commercial outcomes behind them, giving you one view of what each marketing pound is returning in margin rather than in clicks. That is the difference between knowing your spend and knowing whether your spend is working.

• The longer this picture stays fragmented across platform exports, the accountant's software and separate partner judgement, the harder it is to make a confident case for increasing the marketing budget even when the numbers would justify it.

What you do not have is any consolidated view of what the acquired firms brought with them. That is the exposure.

Marketing and data compliance

• The acquired firms' vendor agreements have never been reviewed as a group. That means you may have third-party processors handling client or prospect data under agreements that are out of date, inconsistent with your group standards, or simply unknown to your current leadership. Under data protection law (GDPR), the responsibility for knowing who processes your data and on what terms sits with you as the controller, not with the vendor.

• The marketing-consent rules (PECR) apply to any email or direct marketing you send. If those acquired firms ran their own marketing lists under different consent practices, and those contacts have been folded into your current activity, the consent basis for reaching them may not meet the standard required. This needs checking before your marketing output scales, not after.

• Your cookie and tracking setup across the group's websites almost certainly reflects the fact that sites were built at different times by different people. Each tool added over the years, whether that is analytics, chat, forms or booking widgets, potentially collects data without a corresponding consent mechanism. A light review of what is being collected, by whom, and whether your cookie notice accurately describes it, is a practical and relatively quick exercise that removes a category of risk entirely.

• AI-generated content and AI-assisted client interactions are coming, whether or not you sanction them today. The EU AI Act's transparency duty, specifically Article 50, requires clear disclosure when a person is interacting with AI or being shown AI-generated content, applying from August 2026 to EU-facing businesses. Fenwick & Marsh operates in the UK only, so Article 50 does not apply to you directly, but it signals exactly where UK regulatory expectations are heading, and folding an AI-content disclosure position into your next consent and cookie review costs almost nothing compared to retrofitting it later.

• Making Tax Digital for income tax will generate a sustained increase in client communication volume, which means more data moving through your systems, more automations, more integrations and more potential for something to fall outside your documented processes. Getting your data-processing agreements consolidated and your consent framework documented now means you are not also doing that work under time pressure when the volume arrives.

• The practical first move is to appoint one named owner for compliance across the group, not a committee, one person with authority, and commission a structured review of consent, cookies and all data-processing agreements across every entity and website. This does not need to be a lengthy project. It needs to be thorough, documented and done once properly. Anicca can support the marketing-data side of that review, specifically what your website tools collect, how consent is captured and presented, and whether your current setup can be relied upon when marketing activity increases.

AI usage

Several members of your team are already using ChatGPT for drafting letters, unsanctioned and without any agreed policy. Your AI usage policy is in draft.

This is a real risk in an accountancy practice given the sensitivity of client information, and your partners are right to be concerned, particularly ahead of formal ICAEW guidance. Closing that gap, deciding what is and is not permitted, building a usage policy and training your people on it, is precisely what Anicca's AI Adoption Roadmap covers.

That is a separate engagement from this report, and we would recommend treating it as a parallel workstream rather than deferring it.