Setting up proper spend control for one client is easy. You know what it needs. Decide who's allowed to approve what, put a threshold or two in place so the big stuff gets a second look, make sure the person raising a payment isn't also the person releasing it, and keep a record that ties each approval to the payment it authorised. An afternoon's work, and the client is in far better shape than they were that morning.
The trouble starts at client number two. And forty. And two hundred.
Because the obvious way to do this is the manual way, client by client, building each one from nothing. A spreadsheet of who approves what here, an email rule there, a mental note to check the payment run every Friday. That works right up until the point it's your whole book, and then it becomes the thing that quietly eats your team alive. The firms that look at spend control, agree it's a good idea, and never actually roll it out almost always die on this exact hill. Not because the service is hard, but because doing it forty times by hand is.
So the real question was never whether spend control is worth offering. The last post, the work you're already doing for free is a service, settled that. The question is how you deliver it across an entire client base without it collapsing under its own weight. And that turns out to be a solvable problem, as long as you stop treating each client as a blank page.
Here's the insight that changes the economics. Your clients are not as different from one another as they feel when you're sitting inside each one.
A five-person agency and a thirty-person contractor look like completely different animals, but the control they need has the same skeleton. Someone raises a commitment. Someone with the authority checks it. A second person releases the money. A record captures who did what and when. The specifics change, the thresholds move, the names in the boxes are different, but the shape is identical from one client to the next.
Which means the work that doesn't scale isn't the setup. It's the reinvention. If you design the control process once, properly, as a template, then onboarding a new client stops being a build and becomes an adjustment. You're not asking "how should this business control its spend," a question that takes days. You're asking "where does this business differ from the standard," a question that takes an afternoon. That is the whole game. Get the template right and the marginal cost of the next client falls through the floor.
The research on firm onboarding backs this up bluntly: firms bleed hours, often more than a dozen per client, on disorganised, rebuilt-every-time setup, and that time comes straight out of your margin. Standardising the process is the single biggest lever you have. It's true for bookkeeping onboarding, and it's just as true for control.
The instinct, once someone says "template," is to worry you'll be forcing every client into an identical box. You won't, and you shouldn't. The skill is knowing which parts to fix and which to leave loose.
Fix the structure. The stages a payment passes through, the principle that raising and releasing are always separate jobs, the rule that everything above a certain size gets a second approver, the requirement that every approval leaves a trail. These are your firm's standard, they don't bend per client, and they're the thing that makes the service consistent and defensible. If a client ever gets audited, or ever gets hit, this is what protects them and you. If you want the mechanics of that separation, how to set up segregation of duties for accounts payable walks through it.
Leave the details loose. Who specifically approves. At what numbers the thresholds sit. How many layers a given client needs. Whether a particular department head gets a say. This is where you flex to fit the business in front of you, and it's the part that takes an afternoon rather than a fortnight, because you're adjusting a known thing rather than inventing an unknown one.
Get that division right and you've built something rare: a service that's genuinely consistent across your book, and genuinely tailored to each client, at the same time. Those usually pull against each other. A good template is what lets you have both.
You can, in theory, run all of this on spreadsheets and discipline. Plenty of firms start there. But the manual version is exactly the thing that doesn't scale, because every client you add is more approvals to chase, more payment runs to eyeball, more trails to reconstruct by hand when someone asks who signed off on what. The discipline holds for ten clients and starts to fray at thirty.
This is the point where a tool built for the job stops being a nice-to-have. ApprovalMax exists to do precisely this: it puts a real approval workflow around a client's spend, keeps the person raising a payment separate from the person approving it, enforces the thresholds and the delegation of authority you've set, and captures a complete trail automatically, all sitting on top of the QuickBooks Online or Xero the client already uses. Crucially for a firm, you build a workflow structure once and apply it across clients, adjusting the specifics rather than rebuilding the logic. The template lives in the tool, not in your team's heads.
There's also a partner setup designed around firms doing exactly this, with a success manager to help you get the model right, staff access so your whole practice can work in it, a free licence for your own books so you're running on what you recommend, and a referral arrangement once you're set up. The point isn't the feature list. The point is that the hard part of scaling this service, the standardise-once-adjust-per-client engine, is a solved problem, so you don't have to solve it with your team's evenings.
Step back and look at the whole arc this series has traced. What accountants miss when they only look at the books made the case that there's a control gap sitting in nearly every client you have, and you're the only one positioned to see it. The work of closing it is advisory, it's valuable, and you're already most of the way to delivering it. And the one real obstacle, doing it across a whole book without drowning, comes down to a single decision: design the process once and reuse it, rather than starting fresh every time.
That's not a heavy lift. It's an afternoon building a template, a tool to hold it, and a new line on the engagement letter for work you were largely doing anyway. What you get back is a service that's hard to copy, hard to automate away, and stickier than almost anything else you sell, offered to the clients who already trust you, at the moment the rest of the profession is scrambling to figure out what advisory even means for them.
You were already doing the work. This is just the part where it starts paying you for it.