Setting up a new item takes a few minutes. A box arrives, a customer is waiting, someone fills in enough fields to sell it, and the day moves on.

The trouble is that nobody comes back. The item goes on selling for years with the cost, the price rule, the department and the reorder settings it was given in those few minutes. If any of them was wrong, it stays wrong.

What item setup touches

Each field on a new item feeds something else in the business. That’s what makes a small mistake expensive.

A part set up in the wrong department makes that department’s margin look wrong for as long as it sells. A price typed in by hand stays put when the cost goes up. A missing reorder point means the item sells out once and is never ordered again.

Decide who owns item setup

When whoever’s at the counter sets up the item as the box arrives, it’s fast, but the rules end up depending on who was working that day.

Item setup works better with an owner: one person, or a small group, who sets up new items, or at least reviews them before they’re final. They know the rules, and they check what others have set up.

A short checklist for every new item

Write your own rules down: how departments are split, which pricing rule applies to which kind of item, and when a manual price is allowed. The checklist only works if everyone is following the same rules.

Review new items regularly

Even with a checklist, mistakes get through. A short review of the items set up in the last month catches most of them while they’re still new: anything with a manual price, no reorder settings, an empty bin or a department that doesn’t fit.

It’s also worth looking back at the oldest items once in a while. Costs, vendors and rules change, and items set up years ago were set up under rules that may no longer apply. That’s often where the prices that are costing you money turn up, and it’s the kind of question the system you already have can usually answer.

Equipment dealers and seasonal retailers know the pattern well. A few months bring in most of the year’s revenue. When they arrive, there’s no time to think: the phones ring, the service bays fill and the counter is three deep. Whatever was planned before then is what you’ve got.

That’s why the slow months matter so much. They’re the only time there is to look back at last season, decide what to change, and set the business up for the next one.

Start with last season

Planning starts with an honest look at what happened, while it’s still fresh.

Write it down. A few lines under each heading are enough. Without them, next year’s plan is built on memory, and memory flatters.

Four plans, made together

Preseason planning covers four things, and they depend on each other. A buying plan that ignores staffing, or a service plan that ignores parts, just moves the problem.

Buying with the season in mind

Early-order programs can be worth having. They can also fill the floor with whatever earned the best discount. The test is the same one that applies to any purchase order: is this need, want, or vendor pressure?

Start from last season’s sell-through and what you know about the coming one, then decide what the program is worth to you. Plan when stock should arrive as carefully as how much of it to order.

Level the service workload

A service department can only do so many hours in a week. When the season starts, everyone wants their machine at once. Preseason service work, booked into the slow weeks, moves some of that demand to a time the shop can handle it.

That only works if it’s planned: the offer, the booking schedule, and the parts on the shelf to do the work. It also gives technicians steady hours in a month that would otherwise be thin.

Hire and train before you need them

Seasonal staff hired in the first busy week learn on the job, at the counter, in front of customers. The same people hired a few weeks earlier can learn the routines while there’s time to teach them.

A short one-page guide for each routine makes this far easier, and it’s work that can only be done in the slow months.

Put it on a calendar

The plan can be short. It needs to exist, have dates on it, and have a name next to each part.

Warranty work is easy to think of as a service to the customer. The machine broke, it was covered, it got fixed, everyone’s happy. The part that’s easy to lose track of comes afterwards: your shop did the work, used the parts and paid the technician, and the business only gets that money back if a claim goes in correctly and on time.

Until it’s paid, a warranty claim is a receivable, just like a customer on account. The difference is that nobody chases it unless someone is assigned to.

How warranty money slips away

When a claim comes up short, it’s usually at one of a handful of ordinary steps.

Each one is a small slip on the day. Added up over a season, they can be a real amount of money that’s left on the table, with no line on any report that shows it.

Treat claims like receivables

The fix starts with how the business thinks about a claim. A claim is money owed to the business, with a due date, an owner and a status.

If those four questions can’t be answered for last month’s warranty work, that’s the first thing to fix.

Build the habit at the work order

Most claim problems start before the claim. The work order is where the information gets captured, or doesn’t.

Every manufacturer has its own rules, deadlines and labor allowances. Whoever files claims should have each one’s current requirements to hand, and they change more often than people expect.

Review the open list every month

A short monthly review catches most of it. List every warranty work order from the month, every claim filed, every payment received and every rejection. Anything without a claim, or without a payment after a reasonable wait, gets a name next to it.

Over time the same review shows which manufacturers pay slowly, which kinds of repair get cut, and whether the labor allowances cover the time your technicians actually spend. That last one feeds straight into your effective labor rate.

Your shop rate is easy to find. It’s on the wall, it’s on every quote, and it was set with some care. What the shop actually brings in for each hour a technician is on the clock is much harder to find.

The two numbers can drift a long way apart without anyone noticing, because nothing on a normal service report puts them side by side. The work orders look fine one at a time. The gap only shows when you add them up.

Start with what an hour costs you

Before the rate, there’s the cost. Many shops know exactly what they charge for an hour of labor and have never worked out what that hour costs them.

If a technician is paid hourly, the cost is the wage plus payroll taxes, benefits and paid time off, spread over the hours they work. If they’re on salary, it’s the full year’s cost divided by the hours they’re actually in the building. Either way, it comes to more than the number on the paycheck.

That cost is paid for every hour in the building, whether the hour gets billed or not. That’s why the next three numbers matter.

Three numbers to know

Service labor comes down to three figures, and each one answers a different question.

The last one is the number that matters. It’s what an hour of technician time is worth to the business once everything that happens between the clock and the invoice has happened.

Where the hours go

When the effective labor rate comes in well under the posted rate, the difference is usually a lot of small things, each of them reasonable on the day.

Some of these are a cost of doing good work. The trouble starts when nobody knows how big they are, so nobody can decide which ones are worth the money.

How to measure it

A first answer takes one month of numbers you probably already have.

Do it for a busy month and a slow one if you can. The pattern is usually different, and both matter.

If labor dollars aren’t separated from parts on your invoices, or internal work orders are closed without hours, that’s the first thing to fix. You can’t manage a rate you can’t see.

What to do with the answer

Once the gap has a size, each leak becomes something you can decide about.

Diagnosis time can be priced and quoted up front. Waiting time usually traces back to parts or scheduling, and it shows up in where work orders stall. Internal work on new units is a real cost of selling equipment, and it belongs in the margin on the unit, where it can be seen. Warranty time can be tracked against the allowance so you know which claims are costing you.

Some of the gap will stay, and that’s fine. What matters is knowing what the shop earns per hour, so the posted rate, the flat-rate guide and the staffing plan are built on the real number.

Make it a monthly number

The effective labor rate is worth reviewing every month alongside the service department’s sales. It moves with the season, with staffing and with the mix of work, and a change in it is often the first sign that something in the shop has shifted.

Put it on the same page as billed and paid hours. Three numbers, one line each. Over a few months, the owner and the service manager will start to see the same thing, which is most of the battle.

It doesn’t much matter whether you bought the business, took it over from a parent, or moved up from the parts counter. The job in the first year is the same: work out what the business actually does, as opposed to what everyone says it does, before you start changing it.

That’s harder than it sounds. A business that’s been running for twenty years doesn’t look like a set of choices. It looks like the way things are.

What you actually inherited

Every established store runs on decisions made a long time ago, for reasons that were good at the time. A markup rule. A stocking list. A labor rate. A vendor everyone uses because the last owner liked the rep. Who’s allowed to approve a return. How the month gets closed.

Some of those are still exactly right. Some stopped being right years ago and nobody noticed, because nothing ever forced the question. From the outside, the two look identical. Both are just “how we do it here.”

So the first job isn’t to fix anything. It’s to separate the decisions that are still earning their keep from the ones that are only habit.

Get it out of their head while you still can

The most valuable thing you can do in the first few months costs nothing. Sit down with whoever ran the place before you and write down why things are done the way they are.

If you bought the business, the previous owner may be around for a transition period. If you took over from a parent, they’re probably still a phone call away. Either way, that access won’t last forever, and the knowledge that matters most is the kind nobody thinks to mention.

Ask about the decisions, not the procedures. Your staff can show you how to receive a shipment. Only the person who set the business up can tell you why the parts markup is what it is, which customers get special terms and why, and which vendor will take a return when the others won’t.

Write the answers down somewhere other than your own notebook. They’re the start of the documentation the business never had.

What to check first

Some things pay back faster than others. This is the order that tends to matter, because each one affects decisions you’re about to make.

Each of these has its own article if you want to go deeper: month-end, the physical inventory, slow-moving stock, the service department and money earned and never collected.

What to leave alone for now

The temptation in the first year is to change things, especially if you came in with ideas. Some changes can’t wait: anything costing real money every week, anything unsafe, anything that wouldn’t stand up to your accountant. Most of the rest can.

Changing something before you understand why it was done that way has two costs. Sometimes you break something that was working for a reason you couldn’t see. And every unexplained change tells the staff that the way they’ve always done it was wrong, which is hard to recover from in your first few months.

That doesn’t mean waiting a year to do anything. It means changing things in the order you understand them, and being able to say why each time.

Coming from outside, or from inside

People arrive in this seat from two directions, and each has its own blind spot.

Neither starting point gives you a standard to measure against. Coming from outside, you don’t know what normal looks like here. Coming from inside, what you’ve always seen is normal.

Make it a number, not an opinion

The answer to both is the same: replace opinions with numbers everyone can see.

“I think our parts pricing is too low” starts an argument with everyone who set it. “Here’s what this department earns, here’s what it should earn, and here are the items causing the gap” isn’t an argument at all. It’s a conversation about what to do next.

In a business you’ve just taken over, being right isn’t enough. You need to be able to show your work: to the staff, to a lender, and sometimes to the person you took it over from.

It’s also how you bring people with you. Staff who can see the number you’re looking at can help fix it. Staff who are only told something is changing will wait to see whether the change lasts.

A first year in four parts

The order matters more than the speed. One change made in the second month, with the reason written down, will outlast three made in the first week.

The same story from the other chair

Most of what makes a takeover hard is what the last owner never wrote down. If you’re building a business you’ll pass on one day, the work you do in your first year is also the start of making it easier for whoever comes next. The Value of Your Business Is Set Before You Sell It covers that side.

It’s a familiar pattern. The new procedure gets explained, everyone nods, and a month later the work is being done the old way again. It’s tempting to blame the staff. More often, the training was built in a way that couldn’t survive a busy day.

Training that lasts has a few things in common, and none of them are expensive.

Teach the reason before the screen

A lot of system training is screen training: click here, enter this, press that. It’s necessary, but on its own it doesn’t stick, because nobody remembers a sequence of clicks they don’t understand.

Someone who knows why receiving has to match the purchase order will find the right screen. Someone who only learned the screen will skip a step the first time they’re in a hurry, because they don’t know what it protects.

So start with the basics: what the step is for, what goes wrong later when it’s skipped, and who that affects. Then the screens.

Teach each role what it needs

Receiving and the bookkeeper don’t need the same training. Put everyone in one room for the same session and most of it is wasted on most of the people there, while the parts that matter to each of them get lost in the rest.

Train by role. There are two groups, and they need different things.

Good decisions need accurate numbers, and accurate numbers depend on daily habits. Train only one group and the other half of that stays broken.

Use your own business as the example

Generic examples are easy to nod along to and hard to apply. Your own departments, your own items and your own numbers aren’t. When the example is the shelf they restocked this morning, what someone learns on Tuesday gets used on Wednesday.

It also takes the theory out of it. “Slow stock ties up cash” is a principle. “These items on aisle six haven’t sold since last spring” is a job.

Leave something behind they can use on a busy day

The session ends. The busy season doesn’t. What carries the training through is something at the counter or on the wall that shows the routine in a few steps: a one-page job aid.

A good job aid is short enough to read while a customer waits, written around your system and your procedures, and specific about the steps that get skipped. It isn’t a manual. It’s the part of the manual that matters on a Saturday in April.

An example of a receiving exception checklist. A real one follows your own system and procedures.

Job aids also solve the problem of the person who wasn’t there. The counter hire in March wasn’t in the room in January. A one-page guide, and a recording of the session, means they learn the same routine everyone else did.

Schedule it around the season, not the calendar

An independent store can’t send eight people to a classroom for two days, especially in spring. Training that ignores the season either doesn’t happen or happens to a room full of people watching the door.

Mix the formats. Short sessions on site for the people who need to hear the same thing at the same time. One-to-one time for the person who carries a responsibility alone. A recording for anyone who joins later. Save anything that needs real time for the slow weeks.

Check whether it stuck

Training isn’t finished when the session ends. It’s finished when the work is being done the new way without anyone watching.

Pick one thing to check for each routine, and look at it a few weeks later. Are receiving exceptions being logged? Are counts matching? Is the slow stock moving? If the answer is no, look at the training, the job aid and the routine itself before you look at the person.

Changing your DMS, ERP or point of sale is one of the biggest projects an independent business takes on. It touches every department, it costs real money, and it happens on top of a full workload, not instead of one.

It’s natural to expect the new system to fix what was wrong with the old one. Sometimes it does. But a lot of what goes wrong with a system was never the system. It was how it was set up, how the data was kept, and how people were trained to use it. Bring all of that across unchanged and you’ll have the same problems on a newer screen.

First: is it the system, or the setup?

Before committing to a conversion, be sure the system is really the problem. Reports nobody trusts, inventory that never matches, margins nobody can explain: these are as often setup and habit problems as software problems.

If the system you have, properly set up, could do the job, a conversion is an expensive way to find that out. Getting More From the System You Have covers how to tell the difference, and Start With the Decision, Not the Report covers the question to ask first.

If it really is time to change, the rest of this is about getting the business side right.

What the vendor does, and what they don’t

A software vendor knows their product. They’ll install it, configure it, convert your data and train your staff on the screens. What they aren’t responsible for is your business: your policies, your decisions, your staffing, your sign-offs, and how one department depends on another.

Owners often assume the implementation includes everything in the second column. It usually doesn’t, and that gap is where conversions get into trouble.

Make the decisions before they become configuration

Every setting in a new system is a business decision in disguise. How parts are priced. How special orders are handled. Which account a deposit goes to. Who can approve a return, and up to what amount. How departments are split for reporting.

If nobody on the business side makes those decisions, someone else will: a consultant working from a template, or a staff member answering a question on a call without knowing it affects month-end. It works until it doesn’t, and by then nobody remembers why it was set up that way.

Write each decision down: what was decided, why, and who decided it. A year from now, when someone asks why the system does what it does, you’ll be glad you did.

Clean the data before you move it

A data conversion copies what’s there. Duplicate customers, part numbers superseded years ago, on-hand quantities nobody has counted, vendors you stopped buying from: they all come across, and now they’re in the new system too.

Cleanup is slow, unglamorous work, and it’s tempting to leave it until after go-live. Don’t. After go-live there’s never time, and every report built on bad data inherits it.

Two of those have their own articles: getting a physical inventory you can trust, and customer deposits, which are easy to carry across wrong.

Test the business, not the buttons

Vendor testing checks that the software works. Business testing checks that your business works in the software. They aren’t the same thing.

Take real scenarios from a normal week and run them from start to finish. A special order with a deposit that gets picked up a month later. A warranty repair with parts. A return on a sale from last season. A commercial account on terms. A cycle count. Follow each one all the way through to month-end, and check that the numbers land where they should.

Anything that fails in testing is cheap to fix. The same problem found on the first busy Saturday after go-live isn’t.

Agree what ready means before the day

Go-live dates have a way of becoming fixed long before anyone knows whether the business is ready. Decide in advance what has to be true to go live: which tests have passed, who’s been trained, and what happens if something goes wrong on the day. Then stick to it.

Postponing a conversion is a legitimate outcome. Going live on a date nobody believed in is how a project that was mostly fine becomes the one everyone remembers.

Don’t skip the part after go-live

The weeks after go-live are when habits form. If people find workarounds for things that don’t work yet, the workarounds become how the business runs. If reports aren’t checked against reality, nobody trusts them, and the old spreadsheets come back.

Stabilization gets skipped more often than any other phase, because the project feels finished. It isn’t finished until the workarounds are gone, the reports have been checked, and everyone is using the system the way it was set up to be used.

The shape of a conversion

Each phase has decisions, owners, and something that has to be true before the project moves on. The vendor runs the software through all six. Somebody has to run the business side.

A well-run hardware store or home center already has most of the habits that matter: counting things and believing the count, controlling shrink, buying to demand instead of a rep’s enthusiasm, and knowing which parts of the store pay the bills. All of that carries over.

What doesn’t carry over is the idea that the new departments will behave like the old ones. An equipment line brings four departments with it, and each runs on different rules from the rest of the store.

One line, four departments

Each of the four needs its own setup, its own measures and someone responsible for it. Treating them as one “equipment department” is how problems stay out of sight.

The money is still made on the buy

Retailers already know this, and it doesn’t change. What changes is what a good buy looks like when the catalog runs to thousands of part numbers you’ve never carried.

Parts end up on the shelf in a few familiar ways. A customer asked for one once. A rep recommended a stocking package. It came on an opening order nobody questioned. None of those tells you how the part sells, how many to keep, or what happens to it in the off-season. Three years later it’s obsolete, but the money was spent the day it was ordered.

Before the first season, settle how parts will be stocked: which numbers stay on the shelf, how many of each, who decides, and how often the list gets reviewed. Then decide what happens to special orders that never get picked up.

Need, Want, and Vendor Pressure covers how to buy with a reason, and Freeing Cash Trapped in Dead Stock covers what to do when the shelf is already full of the wrong things.

Parts don’t belong on the store’s markup

One markup rule across a hardware floor is workable. Across a parts catalog it isn’t, because parts sell at very different speeds and get very different price scrutiny. The filter a customer buys every spring gets compared with the shop down the road. The part they need to get a mower running this afternoon doesn’t.

Buying decides what you own. Pricing decides what you earn on it. Both need rules written for parts, not borrowed from the rest of the store.

The department you didn’t ask for

Almost nobody adds equipment because they wanted a service department. It came with the line. The manufacturer required it, so a technician got hired, some floor got cleared, and attention went back to selling machines.

Nothing about running retail prepares you to price labor, cost it or track it. So the labor rate ends up being whatever the shop across town charges, the real cost of an hour never gets worked out, and service gets written off as something the store carries.

Measured properly, it’s often the opposite. Staffed to the work, stocked so jobs aren’t waiting on parts, and measured on billed hours against hours paid, a service department can be the most profitable part of the business. It earns a higher margin than the machines, and it pulls parts across the counter with it.

Turning Work Orders Into a Production System goes further into running service like a production line.

Money that expires

Warranty work is paid by the manufacturer, not the customer, and only if the claim is filed correctly and on time. Co-op funds and program rebates work the same way: real money, owed to you, with rules and deadlines attached.

When nobody owns this, claims get filed when someone has a minute, rejected claims don’t get resubmitted, and co-op money expires unclaimed. None of it shows up as a loss, because it was never recorded as income in the first place.

Give it an owner and a routine from the start. The Loss You Watch Is Rarely the Largest One covers why this kind of money goes missing and how to find it.

Don’t let it disappear into one number

Blended into a store-wide margin, an equipment department can underperform for years without anything looking wrong. A strong hardware season covers a weak parts margin. A busy spring hides a service department that loses money on every job.

Set the system up so each of the four can be measured on its own from the first day. Rebuilding the account structure after two seasons of mixed-up history is far more work than getting it right before the first sale.

Questions to settle before the first season

If you’re already carrying equipment and a season didn’t go to plan, the same questions apply. The difference is that you’ll be answering them with real numbers, which makes it easier to see what needs fixing first.

Shrink is the gap between what the system says you have and what’s actually on the shelf. It is a result, not a cause. Treating every shortage as theft sends people looking in the wrong place, and the receiving errors, unposted parts and unrecorded transfers behind the rest of the gap keep happening.

Theft deserves attention. It’s just one of several reasons the record and the shelf disagree, and the only one that starts as a security problem.

What shrink actually measures

If the system says ten and the shelf has seven, three are missing. Multiply by cost and that is the shrink value. The arithmetic is simple. What it doesn’t tell you is why.

That matters because the fix depends entirely on the cause. A receiving problem needs a receiving fix. A parts-posting problem needs a service fix. Neither one is solved by a camera.

Three kinds of shrink

Classifying a shortage before acting on it is what keeps the response proportionate. A store that treats every variance as unknown spends its energy on suspicion. A store that sorts them first finds out how much of the gap has an ordinary cause, and ordinary causes can be fixed with process.

Where it hides in an equipment and hardware store

Some causes are specific to the way independent dealers and hardware stores work. These are the ones worth checking first.

Each of these can produce a shortage that looks exactly like theft on a count sheet. None of them involves anyone taking anything.

Overages are a warning too

A count that comes in higher than the record is not a gain. It means the record is wrong in the other direction, and it distorts reordering, valuation and purchasing just as much as a shortage does.

Overages come from the same kinds of transaction problems: a receipt posted twice, a sale that never posted, parts consumed on a job but never taken out of stock. Stock in the wrong location shows up the same way — short in one bin, over in another.

A quantity below zero is the clearest sign of all. Nothing can be on a shelf in negative numbers, so a negative on-hand is always a record problem, often a receipt that was missed or posted after the sale. Correcting the number without finding the cause just means it comes back.

Check the count before you blame anyone

Before a shortage becomes an accusation or a write-off, it should go through the same short sequence every time.

Counts go wrong too — a case counted as one, the overflow location missed, a look-alike part counted in the wrong bin. A variance that turns out to be a counting error should be corrected and noted, not investigated as a loss.

Patterns tell you more than single shortages

One shortage on one item is often noise. The same item coming up short count after count is a signal. So is shortage that clusters in one place.

A number you can measure is a number you can reduce

The basic defenses aren’t complicated: accurate receiving, so stock enters the system correctly; secured storage and controlled access for high-value items; and regular counting, so the record is checked against the shelf more than once a year.

Once shrink is measured and sorted by cause, it stops being an unexplained loss at year-end and becomes a short list of fixable problems. The theft that remains is then a smaller, clearer number, and much easier to act on.

Where to start

Internal controls are the routines that make honest mistakes visible and dishonest ones difficult. They are not a judgment about anyone. They are a judgment about the structure: whether one person, on a bad day or a bad year, could move money and hide it without anybody finding out.

Most months, a control program catches small errors. Its real purpose is the rare, large event it makes impossible — the diversion that runs for years, or the single loss big enough to threaten the business. That’s why weak controls are dangerous even when nothing has gone wrong. They measure exposure to something that hasn’t happened yet.

Trust and controls are not substitutes

Trust answers the question “do we believe this person?” Controls answer a different one: “is the structure sound?” A business needs both, and having plenty of the first does nothing for the second.

The losses that hurt small businesses most are usually committed by a trusted person, precisely because nobody was checking. Long tenure is not the risk. A long-tenured employee working alone, with no second set of eyes, is.

What a control program actually does

A complete control program rests on five activities. Each covers a gap the others leave open.

They work as a set. Separating duties means wrongdoing would take two people working together. Authorization keeps transactions legitimate. Reconciliation catches what slips through. Safeguarding protects the physical assets, and access enforces all of it inside the systems where the work actually happens.

No one person should own a transaction end to end

Every transaction has four parts: approving it, recording it, holding the money or goods, and checking the result afterward. When one person does all four, an error and a theft look exactly the same, and neither will be found.

The most dangerous combination is the person who can both approve a transaction and record it. That concentration is what allows a loss to be committed and hidden by the same hands. Breaking it is the single most important control in the program.

Separating duties when the office is two people

The honest constraint of an independent business is headcount. A small office can’t separate every duty, and pretending otherwise produces a policy nobody follows. The goal is not perfection. It is breaking the most dangerous concentrations first, starting with anyone who can both move money and record it, and using the owner as a checkpoint where staff can’t be split.

None of these adds much work. Each one puts a second person at the point where money could leave without anyone noticing. Owner review isn’t as strong as a fully separated department, but it covers the riskiest points, and in a two-person office it’s the control that is actually available.

Money leaving the business gets the strongest control

Of everything the back office does, sending money out carries the most risk. That’s where the strongest controls belong.

The written list matters more than it looks. When approval authority only exists in people’s heads, it bends under pressure — a vendor calling for payment, a busy week, an owner on the road. Written down, it’s the same answer every time.

Logins are controls too

System access should mirror that approval list. People can do what their role requires and nothing more. Access gets reviewed on a schedule, tightened when someone changes roles, and removed the day someone leaves.

Every action in the system should trace back to a named person and a source document. That trail is what lets a problem be found, and knowing it exists is itself a deterrent. Shared logins and old accounts for people who no longer work there break the trail completely. They’re among the most common gaps, and among the easiest to fix.

A control that gets skipped is not a control

Controls decay when nobody checks them. The reconciliation that used to happen monthly slides to quarterly. The second signature becomes a formality. The approval list stops matching who actually approves things.

Keeping the program real doesn’t take an audit. It takes a periodic look at whether the controls are actually being followed: reconciliations done, approvals genuine, access current. A control that exists on paper but gets bypassed every time things get busy provides no protection at all.

What a review tends to find

Consider a dealer where the owner believed the controls were good. A review found the bookkeeper entered the bills, wrote the checks and reconciled the bank. Nobody had reviewed system access in years, and two former employees could still log in. Cash was locked up every night but never counted against the records.

Nothing had gone missing. But one trusted person could have diverted money for years without detection, and nobody would have known until it was too late to recover.

The fix was small: the owner began releasing payments and opening the bank statement first, the old logins were removed, and cash counts were added on a schedule. The controls cost very little. The exposure they closed was large.

Where to start

Start with the conversation, not the paperwork. The message is not “we suspect you.” It is “we are protecting the business, and protecting you if anything ever goes wrong.”