[Home](https://stoneforgegroup.com/) · The Anatomy of a Well-Run Business The Flagship Assessment # The Anatomy of a Well-Run Business Work through the checklist – eleven areas, six to ten specifics in each – rate what's true of you and what leaving it costs, mark what doesn't apply, and see where to focus next. No email, no sign-up, and nothing you answer is saved or sent. Work through it in one sitting, or save your progress to a file on your device at any point and load it again at the top when you come back. Prefer a 3-minute intro first? Watch it 3 min One question first How does your business mainly win and keep its clients? Businesses lead on one of three things and stay competent at the other two – knowing clients deeply, having the best offering, or being the most efficient way to get it done. The distinction is Treacy and Wiersema's, from The Discipline of Market Leaders. "Not sure" uses the first set – the questions about knowing each client deeply. Six to ten specifics per area. Rate what's true, mark what isn't relevant, and the area scores itself. Explain each statement The file is a plain spreadsheet (CSV) – it opens in Excel, and it's yours. We never see it. ## What a well-run business looks like What does a well-run business look like? Every part working, and every part working together. A business can market brilliantly and still stall because the operations behind it can't keep a promise, or keep flawless books while the pipeline quietly empties. Running well isn't excellence in one loud area – it's the absence of a weak one, across eleven: leadership and direction, how clients are held, the revenue engine, the financials, operations, people, visibility, technology, risk, whether the business actually changes when it learns, and what the whole thing is worth. The assessment on this page works through the specifics under each area – how true each one is of you today, and what leaving it as it is would cost – and the area scores compute from your answers, with the distance between the two readings setting your priority order and an area's worst statements counting as heavily as its average. It's the same two-scale method that importance-performance analysis has used since the 1970s, run on the whole business. Do it alone, or with your leadership team rating it independently and comparing. Nothing you answer leaves the page. **Pipeline.** Every live opportunity the business is working on, with where each one has actually reached. [Full entry →](https://stoneforgegroup.com/glossary/#pipeline) How this works 1 Two ratings per statement. How true it is today, and what a year of leaving it exactly as it is would cost. Neither means anything alone – the distance between them is the finding. 2 Six to ten specifics per area. Rate what's true, mark what isn't relevant, and the area scores itself. The section score is never asked – it is computed from the items and said back. 3 The gap ranks the work. An area that would stop the plan and runs unevenly outranks an area that runs poorly but would cost little to leave. Fix by gap, not by embarrassment. 4 Where you're strong and leaving it would cost little – that's worth a look too. Overinvestment hides there, and those hours belong to the gaps. 5 Run it again in a quarter. The file you save is the baseline; the comparison is the point. The research this is built on The two-scale mechanic is importance-performance analysis, introduced by Martilla and James in the *Journal of Marketing* in 1977 and in continuous use since – rate performance, rate importance, and let the gap set the priority. The three ways of competing that switch the Client Approach section come from Treacy and Wiersema's *The Discipline of Market Leaders*: lead on one of customer intimacy, product leadership or operational efficiency, and stay competent at the rest. The team protocol – owner first, then key people independently, then compare – exists so the first rating each person gives is their own: a rating formed in the room takes the room's shape. No industry benchmarks appear here because honest ones don't exist across industries and sizes – the assessment benchmarks you against your own priorities, which is the comparison you can act on. The guidance under each area is operator judgment – ours, from running and advising owner-led firms – deliberately routed to the profession that owns each terrain rather than dressed in citations. Where a public standard exists we follow it: the security items track standing guidance from CISA and the FTC's small-business program. Everything legal, insurance and tax routes to your own attorney, broker or CPA by design. **CPA (Certified Public Accountant).** A professional designation for accountants who pass the Uniform CPA Examination and meet additional state education and experience requirements. [Full entry →](https://stoneforgegroup.com/glossary/#cpa) Questions owners ask about running the whole business ## What does a well-run business look like? Every important area running soundly, and no weak one dragging the rest. A well-run business isn't the one with the most impressive strength – it's the one without a critical gap: direction people can repeat, clients held on purpose, work arriving deliberately, numbers that steer, operations that hold, people who stay, visibility a buyer can check, technology that's examined, risk that's written down, a habit of changing when it learns, and value that survives without the owner. ## How do I assess my business across all areas? Work through a checklist of specifics – rate how true each one is of you today and what leaving it as it is would cost, and let the area scores compute from your answers rather than from a gut guess. The double rating is what makes it honest: a weak specific that would cost nothing to leave isn't a priority, and a decent one that would stop the plan might be. That's the method on this page. ## What is importance-performance analysis? A prioritization method from Martilla and James (1977): rate each part of something on performance and on importance, then plot the two. Where importance is high and performance low, concentrate. Where both are high, keep it up. Where performance is high and importance low, you may be overinvesting. It's been used on everything from clinics to airlines – here, it's run on the whole business. ## What are the three value disciplines? Treacy and Wiersema's finding: market leaders excel at one of three things – customer intimacy (knowing and shaping around each client), product leadership (having the best offering), or operational efficiency (being the most reliable, lowest-friction way to get it done) – while staying competent at the other two. Which one you lead on changes what good client service looks like, which is why this assessment asks first. ## How often should I reassess the business? Quarterly is the rhythm this was built around – often enough to show movement, far enough apart for movement to exist – and the calendar file below sets exactly that. Adjust to your own pace of change; the constant is comparing against your last answers rather than starting fresh. Treat an area that improved as evidence the method works – and one that didn't as the agenda. ## What is a business health check? A structured review of every part of the business rather than the part currently making noise – leadership, how clients are held, how work arrives, the finances, operations, people, visibility, technology, risk, whether the business changes when it learns, and what the business is worth. The useful version rates each area twice: how true it is today, and what a year of leaving it as it is would cost. A health check that produces only a score tells you less than one that produces an order of work. ## How do I decide what to fix first in my business? Rank by the gap between how true an area is today and what a year of leaving it would cost – not by what went wrong most recently. An area running poorly that would cost you almost nothing to leave is a fact, not a priority; a middling area that would stop the plan outranks it. Done in your head, whatever hurt this week takes the whole answer – on paper, every area gets asked. ## How do I get my leadership team to agree on what matters most? Have everyone rate the business independently before anyone discusses it – the owner first, then two or three key people, without seeing each other's answers. Ratings arrived at together tend to converge on the loudest voice in the room; independent ones surface the real spread. Then lay the results side by side. Where you all agree, believe it. Where you disagree is the agenda for your next leadership meeting. Solo? The protocol still works with anyone who sees the business from beside you. ## What's the difference between a business assessment and a scorecard? Scale and depth. A scorecard is a fast look at one dimension – seven gut answers on how the selling runs, or how the finances run – and takes about two minutes. A whole-business assessment works through the specifics under every area and produces a ranked order of work rather than a position. Use a scorecard to find the area; use the assessment to work the whole business. ## How long does it take, and what do you end up with? One sitting alone, or a working session with your leadership team rating it separately and comparing. You end up with all eleven areas mapped – how well each runs against what leaving it costs – the areas ranked by that gap with what to do about each, where you may be over-invested, and what's carrying you. Plus a full report and your own saved file, so a re-run next quarter compares against today. ## Is this a business audit? No, and the distinction matters. An audit is a formal examination of your accounts against a standard, done by someone qualified to sign it off. This is a self-assessment – a structured evaluation of how the business runs, area by area, scored by you. It produces a diagnostic and a priority order rather than an opinion on your financial statements. Nothing here is verified, and nothing here is filed. ## How do I find the strengths and weaknesses in my business? Rate every area on two scales – how true it is today, and what a year of leaving it as it is would cost – then read the gaps. That gap analysis is what separates a real weakness from something merely imperfect: a weak area that would cost nothing to leave isn't a priority, and it's why this produces an order of work rather than a SWOT-style list. Your strengths fall out of the same arithmetic: the areas that matter and run well are what the fixes get to lean on. The checklist in full ## The eleven areas, and what each one is made of. ### 1. Leadership & Direction Everything else inherits its pace and its clarity from here. **What weak looks like:** Decisions queue on one desk, and the plan lives in someone's head. - The people who matter most here can say where the business is heading – and they'd say the same thing. Ask your two or three most senior people, separately, where the business is heading – if you'd get the same sentence back, that's true. If you'd get three overlapping paragraphs, it isn't. Just you? Then the test is whether your own answer comes out the same two weeks running. - There's a written plan for the year, and it gets looked at more than once a year. A plan that lives in a drawer scores low. True means the written plan gets opened through the year, and something in the coming month changes because of what it says. A plan that never says no – to a client type, a service, a market – is a budget, not a plan. - Decisions move at the pace the work needs – nothing sits waiting for one person's attention. Think of the last three decisions that mattered – pricing, a hire, an offer. Did each move at the pace the work needed? Some decisions deserve weeks; the failure is the one that waited only because it waited on one person. - Work leaves the owner's desk with real authority attached, not just tasks with check-ins. Authority means the person can decide without checking back, and their decision stands. If work leaves your desk but every call still routes through you, rate this low. - Every senior person owns a number or an outcome, and could tell you what it is. Each senior person can name the number or outcome they own without looking it up – revenue, delivery dates, cash collected, client retention. Shared vague responsibility for everything counts as no. No senior tier yet? Mark it not relevant – the area scores from the rest. - Commitments made in one meeting get checked in the next – slippage gets named and reset. In your regular meeting, are last time's commitments read back and checked? True means slippage gets said out loud and reset. If unfinished things quietly vanish from the agenda, that's the miss. - You know what your own week goes on – and how much of it is work only you can do. Track a typical week, roughly – how much went on work only the owner can do: direction, key relationships, the judgment calls. The rest is work the business hasn't yet learned to do without you, and it's the ceiling on everything else in this area. **Where to start if weakest:** Write the one-page version of where the business is going – one page, not a planning document; if that page already exists and holds, your gap is the second half of this. Put a standing review of it in the calendar before the month ends, and hold the first one. Then pick one kind of decision you'll no longer make, name who does, and let their next call stand. ### 2. Client Approach How you hold clients is a choice – and what good looks like changes with the choice. Customer Intimacy **What weak looks like:** Service runs by mood and memory, and clients hear from you only when something's wrong. - What we deliver is shaped client by client – it reflects what each one is trying to do. Shaped client by client means the work reflects what each client is trying to achieve – not one service poured into every situation. If two very different clients get identical deliverables, be honest here. - Client satisfaction gets measured – asked or counted, not assumed from silence. Measured means asked or counted – a check-in question, a review conversation, a score, a repeat engagement or renewal. Assuming clients are happy because they haven't complained is exactly what this statement is testing – and what happens when something goes wrong belongs in the measure too. - Clients hear from us deliberately – on a rhythm during the work, and still afterwards. During an engagement, contact runs at a cadence the client can rely on – not only invoices, not only problems. For project work the second half matters as much: past clients still hear from you on purpose, because that's usually where the next engagement starts. If contact ends when the invoice is paid, rate low. - We talk to clients in their language – no jargon between us and what they need to know. Their language means the client can repeat back what you told them and what happens next. If your updates need translating – jargon, internal shorthand, technical detail they don't use – that's a no. - Nothing reaches a client unchecked – someone looks before they do, every time. Checked means a second pair of eyes, a review step, a standard – before the client sees it, wherever the team sits. If quality depends on who happened to do the work that week, that's the gap. - Clients would stay if the owner stepped back – the relationship lives with the firm, not one person. The test is whether clients know and trust anyone here beyond the owner. If every renewal, every escalation and every thank-you routes to one person, the client base is a personal asset, not a business one – and both selling and stepping back inherit that. **Where to start if weakest:** List your five most important clients and write down, for each, when they last heard from you without an invoice attached. Set the cadence you can actually keep – then keep it. And put one check between finished work and the client, every time. Product Leadership **What weak looks like:** Improvements ship when someone finds time, and the difference you sell lives in your head – not the client's. - What we sell gets better on a schedule – improvements land because they're planned, not when someone finds time. On a schedule means improvement is planned and delivered on a rhythm – a development plan with dates, whether the offering is a product, a service or a method. If it only improves when someone finds spare time, rate low. - Client feedback reaches whoever shapes the offering – and visibly changes it. Visibly changes means you can point to a recent improvement and name the client input behind it. Feedback that gets collected and filed shapes nothing. - Quality gets checked before anything goes out – every time, not on trust. Every time means a defined check before anything reaches a client – not trust in whoever built it. If quality rests on "our people are careful," this statement is asking you to be honest about it. - We can say plainly what makes our offering better – and clients repeat it back to us. You can say what makes yours better in one plain sentence – and you've heard clients say it back. If the difference lives in your head but not in theirs, it isn't landing. - Clients hear about improvements and what's coming without having to ask. Without having to ask means improvements reach clients on your initiative – a note from you, a walkthrough, a what's-new conversation. If clients discover changes by stumbling on them, rate low. - Clients would stay if the owner stepped back – the relationship lives with the firm, not one person. The test is whether clients know and trust anyone here beyond the owner. If every renewal, every escalation and every thank-you routes to one person, the client base is a personal asset, not a business one – and both selling and stepping back inherit that. **Where to start if weakest:** List your five most important clients and write down, for each, when they last heard from you without an invoice attached. Set the cadence you can actually keep – then keep it. And put one check between finished work and the client, every time. Operational Efficiency **What weak looks like:** Every job runs its own way, and clients call to ask where things are. - The service runs the same way every time – documented steps, not personal styles. Same way every time means documented steps that don't bend to whoever's running them. If two people would deliver the same job differently, the process isn't holding. - We measure where time and cost leak in delivery, and fix the biggest leak first. You measure where delivery loses time and money – rework, waiting, handoffs – and the biggest leak gets fixed first. Fixing whatever annoyed someone most recently doesn't count. - Client feedback gets read for where the process fails, not just for how people feel. Read for process failure means feedback maps to a step that gets changed – including when the firm got it wrong: a failure that changes the process is recovery; one that doesn't will repeat. If feedback is collected for sentiment and filed, the process never learns. - Clients know what happens next without calling – communication is standard and clear. Clients know what happens next without calling: stages, timing and what's needed from them are visible or told in advance. If clients ring to ask where things are, that's the test failing. - Updates reach clients automatically wherever possible – transparency without manual effort. Automatic means the system tells the client – status changes, confirmations, schedules – without someone remembering to. Manual updates get skipped in busy weeks, which is when clients most want them. - Clients would stay if the owner stepped back – the relationship lives with the firm, not one person. The test is whether clients know and trust anyone here beyond the owner. If every renewal, every escalation and every thank-you routes to one person, the client base is a personal asset, not a business one – and both selling and stepping back inherit that. **Where to start if weakest:** List your five most important clients and write down, for each, when they last heard from you without an invoice attached. Set the cadence you can actually keep – then keep it. And put one check between finished work and the client, every time. ### 3. Revenue Engine Work has to arrive on purpose – an engine, not a run of luck. **What weak looks like:** Warm introductions plus the owner are the whole engine, and the pipeline empties whenever delivery gets busy. - New work would keep arriving for a while if the owner stopped selling tomorrow. Imagine you stopped selling today. Existing momentum – referrals arriving, campaigns running, someone else closing – would keep work coming for a while, or it wouldn't. If the pipeline is you, this is a no. - The growth target splits between named client expansion and new business – a number on each, not a hope. A real target splits in two: growth from clients you can name, and new business you can't name yet – with a number on each. The named half you can test today; the new-business half is what the rest of this engine has to produce. One number with no split is a hope. - There's a way of winning work you could deliberately do more of – and more work would come. The test is control: you could decide to increase a channel – more outreach, more of a campaign, more of a partnership – and more work would come, on whatever cycle your market runs. Referrals you wait for can't be increased by deciding. - Someone owns winning work – with the time and tools to keep at it even when delivery gets busy. Owns by name: one person is responsible for the activity that wins work – outreach, proposals, follow-through on what comes in – with time protected for it in busy months. If selling is everyone's job when there's time, it's no one's job. - You know what's coming – where each pursuit stands, and the numbers that move before revenue. For deal-by-deal work: stage, next step, what's actually true of each. For volume work: the flow – inquiries in, proposals out, conversion. Either way you watch numbers that come before revenue, because revenue alone tells you too late. - Losses get examined for whether they were ever real – lost to a competitor, or lost to no decision. Go through your last few losses one at a time: did a competitor actually win it, or did the client decide to do nothing? The two need opposite responses – one is a competitiveness problem, the other means the deal was never real – and if nobody separates them, the pipeline keeps filling with conversations that only felt like deals. - Prices hold under pressure – when a discount goes out it's a decision with a reason, not a reflex to get to yes. Look at the last few things you won: how many closed at the price you first put on them? A discount given to close a deal that was already closing is margin gone for nothing, and it teaches the buyer what your price really is. The test isn't whether you ever move – it's whether you can say why each time. - Where clients can buy again, they do – and you can say which grew and which stayed flat. Winning more from a client who already trusts you skips the cost of being found and proven all over again. If you can't say which accounts grew, the question is whether anyone owns that. Genuinely one-off work – a build, a transaction, a case? Mark it not relevant; for you the referral question later carries this weight. - The people who win work here have been taught how – not left to work it out because they're good at the job. In many owner-led firms the people winning work are engineers, lawyers, estimators – experts whose training was in the work, not the selling. Being excellent at the craft doesn't teach you how to open a conversation or test whether a need is real. The question is whether anyone ever showed them. - Opportunities get qualified before they get effort – you decide early which are real and which to decline. The discipline is the early no – budget, timing, a real need, a reachable decision-maker, tested before the proposal gets written rather than discovered after the loss. The effort saved goes to the deals that deserve it. **Where to start if weakest:** Run the two-minute sales scorecard – it splits this area into seven parts and shows which is thinnest. Name one channel you could deliberately do more of, and do more of it for a quarter. And give the winning of work an owner – a name, with hours attached. ### 4. Financial Management The numbers either steer the business or report on it afterwards. **What weak looks like:** The bank balance is the dashboard, and the squeeze arrives unseen. - There's a budget for the year, and actuals get compared – variances understood, acted on where it matters. A budget with actuals compared means someone looks at the difference on a rhythm, understands what moved, and acts where it matters. A budget written once and revisited at year-end scores low. - Cash is forecast far enough ahead that a squeeze is seen coming, not discovered. Forecast far enough ahead means you'd see a squeeze coming with time to do something about it – not the morning it arrives. If the bank balance is the early-warning system, rate this low. **Forecast.** A forward view of what the numbers will do, built from what the business knows now – work already won, work likely to be won, costs already committed, and the timing of both. [Full entry →](https://stoneforgegroup.com/glossary/#forecast) - You know what each service, product or client actually earns – margin, not just revenue. Margin, not revenue: you know what each service, product or client actually leaves behind after the cost of delivering it. At volume, client-by-client isn't the ask – service-level margin, honestly held, is. Busy-but-poor lives in the gap this statement is testing. - Invoices go out on time, and what's owed gets managed on a rhythm. On time and managed on a rhythm – invoices go out when the work does, and what's owed gets followed at a standing cadence, not when cash feels tight. - Big commitments get modeled before they're made – a hire, a lease, a price change. Modeled before committed means the hire, lease or price change was run through the numbers first – what it costs, what it must return, by when. Deciding first and checking later is the miss. - The same few measures get watched from one period to the next – results, and the actions that drive them. The same few measures, on a rhythm, and of two kinds: the results you're after – revenue, margin, cash – and the actions that produce them, like proposals out or work delivered. A different dashboard every month, or numbers pulled only when something feels wrong, doesn't steer. - Prices are set deliberately – weighing value, cost and the market – and reviewed on a rhythm. Value-based, cost-plus and market-rate pricing all have their place – the failure is pricing nobody has examined: last year plus a bit, or a number set once and never revisited. The test is whether you can say, for your main services, why the price is what it is – and when you last looked. - The books are accurate and closed on time – you'd commit money on the last close as it stands. Numbers that arrive long after the period ends, or get corrected after you've read them, are worse than no numbers, because you'll act on them anyway. Whatever your close rhythm, the test is simple: would you commit real money on the last set as they stand? - You know how long finished work sits unbilled, and how long invoices sit unpaid. For project work, the gap between finished and invoiced is where cash quietly disappears – finished in March, invoiced in May, paid in July. For retainer work the first gap barely exists, so the second is the whole question: how long invoices sit unpaid. Both are measurable in days, and both are fixable. - When there's money to invest, competing options get weighed against each other – not funded in the order they were asked for. Growth money is finite and every request arrives urgent. The test is whether a hire, a piece of equipment and a marketing push ever get compared on the same terms – what each returns, and by when – or whether whoever pushed hardest got the yes. **Where to start if weakest:** Get the books to where you'd commit money on them – if they're already there, start at the cash forecast: simple, and far enough out to act on what it shows. Then pick the same few measures and look at them on a rhythm, so change is visible. ### 5. Operations & Systems Consistency is what lets the business grow without breaking. **What weak looks like:** Quality depends on who did the work, and every fix is a one-off. - The key processes are written down well enough that someone new could follow them. The test is a new person: could they follow the written process for your key work without a veteran interpreting? If the real process lives in people's heads, the documentation isn't holding. - Repetitive work is automated where it can be – people don't retype what systems can move. Automated where it can be means people don't retype, re-copy or re-enter what systems could move. Every manual bridge between tools is hours and errors this statement is counting. - Your tools fit how you actually work – one source of truth, not five overlapping systems. One source of truth: you know where the current version of anything lives. If the same information lives in five tools and nobody's sure which is right, rate low. - Templates and checklists carry the repeat work – quality doesn't depend on memory. Templates and checklists carry the repeat work so quality doesn't depend on memory or mood. If every proposal, onboarding or delivery starts from a blank page, that's the gap. - When delivery breaks, the fix changes the process – not just that one instance. When something breaks, does the fix change the process – or just that one instance? Fixing the instance feels faster, and leaves the cause in place. - What was sold is what gets delivered – scope, price and promise survive the handoff. Put the last proposal beside what delivery actually did. Where they differ – extra scope, softer deadlines, quiet discounts – margin leaks in the gap. The test is whether selling and delivering read from the same page, literally. - You know how much work the business can take on – and how full it is right now. Capacity as a number: what the team can deliver in a month, and how much of it is committed. Without it, every yes is a guess, and the cost lands on delivery or on the client. Fixed-fee or hourly, the question is the same. - Deadlines and promises to clients are met – and you know how often, not just how it feels. The test is a rough count, not a feeling: of recent delivery commitments, how many landed when and as promised? Clients keep this count whether or not you do. **Where to start if weakest:** Write down the process you repeat most – one page, steps in order, by whoever runs it best. When the next thing breaks, change the process, not just the outcome. Then pick the worst double-entry between your tools and close it. ### 6. People & Talent Your goals for the business are only as real as the people who can carry them. **What weak looks like:** Skill gaps get worked around instead of named, and exits come as surprises. - Roles are defined before the search starts – hiring has a bar and a process. Defined before the search means the role – outcomes, skills, what it pays – exists on paper before candidates do, and the same bar applies to associates and contractors you engage. Hiring around a person you happened to meet is the common miss. - The skills the plan needs exist on the team – gaps are named, not worked around. Gaps are named: you can say which skills the plan needs that the team doesn't have, and what you're doing about each. Working around a gap quietly for a year counts as unnamed. - People learn on purpose – training is planned and happens, not fitted in when things go quiet. Planned means the calendar says so, because things never go quiet on their own. For a team of contractors, the version of this is keeping their bar current – the odd conference trip doesn't make either true. - Good people can see what's next for them here – and stay because of it. Good people can see what's next for them here – a next role, next responsibility, next year. If the honest answer is "more of the same," your best people are answering it too. This is also where a second line gets built: someone growing toward each senior role. - People stay – and when someone leaves, you know why. People stay – and when someone leaves, you know the real reason. If departures surprise you, or the reasons never get asked, this is testing exactly that. - People know how they're doing – told plainly, on a rhythm, including when it's not good. Everyone can answer "how am I doing here?" without guessing, because someone tells them – and underperformance gets named and worked, not worked around. If the first honest conversation happens at the exit interview, it happened years too late. **Where to start if weakest:** Write the role description for the position that's hurting most – before looking at anyone. Name the gaps out loud: which skills the plan needs that don't exist here yet. Then ask your best person what would make them stay – before the day they tell you. ### 7. Marketing & Visibility Being findable and provable is the door the revenue engine walks through. **What weak looks like:** Activity without return, and proof that's gone quietly stale. - Someone searching for what you do would find you – and what they'd find is current. Search for what you do the way a buyer would. You show up, and what they find is current – site, listings, proof. Rate what a stranger would find, not what you know is there. - Proof of your expertise gets published on a rhythm – work, results and answers a buyer can check. Proof a buyer can check means published, on a rhythm – work, results, answers to real questions – and ideally somewhere you own: your site, your list, not only rented feeds. A site that hasn't changed in a year tells buyers something too. - Referrals are asked for, tracked and thanked – not left to luck. Asked, tracked and thanked: referrals are requested deliberately, recorded, and the referrer hears back. A referral arrives pre-trusted – which is exactly why it shouldn't be left to happen by accident. - Client proof is recent and usable – names, results and words you can point to. Recent means recent for your market – proof ages at the pace the work changes. Names, numbers and words you could put in front of a prospect tomorrow, that a buyer would take as current rather than historic. - Marketing spend is judged by what it returned, not by whether it felt active. Judged by return means you can say what came back from what you spent – inquiries, meetings, work. Activity that only feels productive is what this is testing. - Where local standing matters to your market, the business shows up in its community on purpose. Only relevant if local standing matters in your market. If it does: the business shows up on purpose – chosen commitments, kept – rather than scattered sponsorships nobody remembers. - You know which clients are your best – by margin and by fit – and you go looking for more like them. Best doesn't mean biggest. It means the ones that earn well, are good to work with, and come back. If you can't name the three you'd like ten more of, marketing is aiming at everyone, which is the same as aiming at nobody. - When AI assistants answer your buyers' questions, the business is part of the answer – and it's accurate. Some buyers now ask an AI assistant before they ever type a search. The test mirrors search: ask the questions a buyer would ask, and check whether you're part of the answer – and whether what it says is accurate. What AI says about you is assembled from what's publicly checkable, which is the same proof the rest of this area builds. - What you do and who it's for fits one sentence – and reads the same everywhere a buyer looks. Write the sentence, then check it against your site, your proposals and what your people say when asked. If a buyer would meet three versions of the firm in three places, the message isn't holding. **Where to start if weakest:** Search for what you do the way a buyer would – then ask an AI assistant the same question, and fix the first wrong thing either shows you. Get one piece of current proof live: a result, a name, words a buyer can check. Then stop any spend you can't trace to a return. ### 8. Technology, Data & AI Quiet risk and quiet leverage both live here – and both compound. **What weak looks like:** Backups assumed rather than tested, and AI tools in use that nobody has examined. - Backups exist and a real restore has been tested – recently, not assumed. Tested means someone actually restored a file or system from backup, recently enough that what it proved still holds for the systems you run today. An untested backup is a hope, not a backup. - People can reach what their job needs and no more – and leavers lose access the day they go. Access matches the job, and people who leave lose it the day they go. If a departed contractor could still open your drive tonight, rate this low. - The team has had recent training on phishing and scams – not once, years ago. The test is whether your team would recognize this week's version of a scam email – the ones that get through are the ones that look like ordinary business. Training old enough that it describes last year's scams has already expired. - You know which AI tools the business uses – and what happens to the data that goes in. You know which AI tools are in use – including the ones individuals adopted quietly – and what happens to the data typed into them. Client information in an unexamined tool is the risk this is naming. - AI output gets a human check before it reaches a client or a decision. A human check means someone accountable reads AI output before it reaches a client or a decision – verifies it, owns it. Convenient is not the same as checked. - The numbers you steer by don't depend on one person – someone else could produce and explain them. A system here doesn't have to mean software – a shared, documented spreadsheet more than one person can open and understand qualifies. The risk being named is the private one: numbers only their keeper can produce or explain. If that person is out for a month, the steering goes with them. - Sign-ins are protected by a second factor, and passwords are managed – not remembered, reused or shared. Multi-factor authentication on email, banking and anything holding client data means a stolen password isn't enough on its own. A password manager does the rest – unique passwords nobody has to remember. If one shared spreadsheet of passwords is holding the doors, rate this low. **Where to start if weakest:** Test one restore from backup this week – if it fails, you've found the priority. Turn on multi-factor authentication for email, banking and anything holding client data. Then list who can reach what, and close what former staff and lapsed contractors still hold. ### 9. Risk & Compliance A risk nobody has written down is a risk nobody is watching. **What weak looks like:** Insurance from two business changes ago, and agreements that don't describe the work anymore. - Insurance matches what the business actually does now – reviewed since the business last changed. Reviewed since the business last changed – new services, new size, new geography all change what coverage you need. Insurance bought for the business you used to run insures that business, not this one. - Someone tracks the rules of your industry – changes reach you before they bite. Someone tracks the rules of your industry – licensing, regulation, employment and contractor classification, compliance dates – and changes reach you before they bite. Finding out from a client or a fine is the failure mode. - Client agreements are current and signed – the work you're doing is the work on paper. Current and signed means the paper matches the work you're actually doing now. Work that has drifted beyond the agreement is unprotected exactly when something goes wrong. - The risks that could genuinely hurt you are written down, with what you'd do about them. Written down, with what you'd do: the handful of risks that could genuinely hurt – key client lost, key person out, premises or systems unavailable, a critical supplier gone – each with a first response. In heads doesn't count under pressure. - Key-person risk is named – you know what stops if one person is suddenly gone for a stretch. Name your key person – often you. Pick the absence that would actually hurt at your size – a week for some firms, a season for others – and know specifically what stops, and what you've done about the worst of it. "We'd cope" is not a plan. - You know what client information you hold, where it lives, and what you've promised about protecting it. Three answers: what confidential client information the business holds, which systems it sits in, and what your agreements promise about it. A breach is survivable; discovering mid-breach that you'd promised more than you built is the compounding failure. **Where to start if weakest:** Read your insurance schedule against what the business does now – this quarter's version. Put the agreement your biggest engagement actually runs on next to the work being done. Then write the five risks that could genuinely hurt, each with a first response. ### 10. Continuous Improvement The difference between a good year and a good business is whether looking causes changing. **What weak looks like:** Reviews happen, minutes get taken, and nothing runs differently afterwards. - The team hears how the business is doing on a rhythm – and questions get straight answers. On a rhythm means a standing cadence – numbers, wins, what's next – not silence between crises. The second half is the culture test: when someone asks about a number, a direction or a decision, they get a straight answer or an honest "not yet" – asking isn't a risk. - There's a working path from the people doing the work to the people changing how it's done. A working path means someone doing the work can raise a better way and see it reach whoever can change it – and hear back. Suggestion boxes that eat ideas count as no path. - What clients say gets collected, read and acted on – not just collected. Collected, read and acted on – you can name a recent change made because of client feedback, on whatever cycle your clients actually speak. Collecting it is the easy part. - The way you work gets reviewed on a rhythm – and something changes because you looked. Something changes because you looked: reviews produce decisions, owners and dates – and the next review checks whether the change did what it was meant to. If the same issues appear in every review, looking isn't causing changing. - Everyone can connect what they do in a week to what the business is trying to do. Ask the person furthest from the owner what the business is trying to do this year and how their week serves it. If the honest answer is a shrug, the connection this tests isn't there. - Someone watches the world your clients buy in – competitors, changing needs, changing ways of buying. Every other statement here looks inward; this one looks out the window. The test is whether anything you did this year traces to something you noticed out there – a competitor's move, a shift in what clients ask for, a change in how buying happens. If the market has to arrive as lost revenue to be noticed, it will. **Where to start if weakest:** This assessment is the loop – put the quarterly re-run in the calendar below. Bring one change to the first re-run, made visibly because of what you found. Then give the team one working path to raise a better way – and answer everything that comes through it. ### 11. Value & Ownership What the business is worth, and whether that value survives without you. **What weak looks like:** The business is worth what it earns while you're in it, and nobody has checked what that means. - No single client is a large enough share of revenue that losing them would be a crisis. Worth knowing your largest client's share of revenue as a number rather than a feeling – lenders and acquirers read concentration as risk and price it that way, whatever the exact share. The same number tells you how hard the ninety days after that client left would be. - You know roughly what the business is worth, and what a buyer would actually be paying for. You don't need a valuation to answer this – you need to know whether the value sits in contracts, systems, a team and repeatable demand, or in you. It matters long before any sale, because the business that runs without you is easier to live with and easier to sell for the same reason. - You know what credit the business could reach – and on what terms – before you need it. Borrowing capacity is easiest to arrange when you don't need it and hardest in the month you do. Debt-free on purpose still passes – the test is knowing what you could reach if you chose, not using it. If a bad quarter would start that conversation from scratch, rate low. - What the owners take out is deliberate and separate – business cash and personal cash don't blur. In owner-led firms business cash and personal cash run together easily, and it hides the truth in both directions – the business looks poorer than it is, or richer. The test is whether what the owners take is set, predictable and visible in the numbers, rather than taken as cash allows. - You know how today's tax decisions affect what the business would sell for. Minimizing profit for tax is rational year to year and can be expensive at exit – where a sale price is built from earnings, years of suppressed profit read as a smaller business. Worth being on that trade deliberately rather than by default; your CPA can size it. - With more than one owner: what happens on exit, deadlock or death is agreed, signed and current. Sole owner? Mark it not relevant. Otherwise this is the largest unwritten risk in the business: the buy-sell terms, valuation method and decision rules for the day an owner leaves, disagrees or dies. Easiest to agree while nobody needs it – your attorney has the template. **Where to start if weakest:** Write down your largest client's share of revenue, and what you'd do in the ninety days after they left. Separate the money: set what the owners take, and make it visible in the numbers. Then ask what a buyer would actually be paying for – and start building whichever answer you didn't like. ### This is yours to run, whether or not we ever speak. The assessment is free and complete on its own – run it quarterly with your team and it will keep telling you where to look. A fractional seat is the further step: a senior operator in the business with you, taking the two or three areas that keep coming up weakest and building them so they stop coming up. Where those areas are revenue or financial, that's precisely what we do. The first conversation is free – bring your report. [Talk to Us →](https://stoneforgegroup.com/contact/) Fifty minutes, about your business, with the person who'd hold the seat. What this produces is information, not advice – general by design, built from nothing but your own answers, and no substitute for guidance from someone qualified who knows your situation. Using it creates no client relationship or contract with Stoneforge LLC; results are provided as-is, errors and omissions excepted – see [our full terms](https://stoneforgegroup.com/terms/).