What Manual Processes Cost, Timed
Six routine tasks in a Kenyan SME, timed rather than estimated, with what each costs a year. The surprise is not the total — it is which one is largest, and that it is not the one anybody complains about.
Manual process cost is usually argued in the abstract, which is why the argument never wins. "You are wasting time on admin" is unanswerable and unpersuasive at the same time. What persuades is six specific tasks with a stopwatch against them, because then the conversation moves from whether time is being spent to whether that particular hour is worth what it costs.
The figures below are typical of a forty-staff Kenyan distributor. Yours will differ; the exercise of timing them is the point rather than the numbers.
The six, timed
| Task | Time | Annual cost | Who does it |
|---|---|---|---|
| Matching M-Pesa receipts to invoices | 45 min/day | KES 145,000 | Whoever is most trusted, not cheapest |
| Answering "do we have any?" | ~20 × 4 min/day | KES 170,000 | The storekeeper, interrupted |
| Month-end stock and figures reconciliation | 2 days/month | KES 240,000 | Finance, plus the store |
| Chasing an approval by phone and WhatsApp | ~15 min each, 40/month | KES 96,000 | Whoever wants the thing |
| Finding a document — a delivery note, an invoice | ~12 min each, 30/month | KES 58,000 | Anybody, repeatedly |
| Rebuilding the same report each month | 3 hrs/month | KES 43,000 | Usually the best-paid person in the room |
That is roughly KES 750,000 a year of time. Two things about that number: it is not recoverable in full, and the largest line is the one nobody ever complains about.
The interruption cost is the biggest and the most invisible
Twenty stock enquiries a day at four minutes each is eighty minutes, which sounds survivable. What it actually costs is more than eighty minutes, because each one interrupts something else — a count being done, a delivery being received, a shelf being tidied — and the resumption is never instant.
It also degrades the answers. A storekeeper interrupted for the fourteenth time gives an approximate answer from memory rather than walking to the shelf, and that approximate answer becomes the basis of a customer promise. The cost of the interruption is therefore paid twice: once in the storekeeper's time and once in a commitment made on a number nobody verified.
Nobody complains about answering stock questions, because answering them feels like being helpful. It is the most expensive line on the list.
What is genuinely recoverable, and what is not
This is where honest arithmetic separates from a vendor calculator. Not all of it comes back, and the parts that do come back arrive at different times.
The same six, after software, honestly
Roughly 65% of the total, not 100%, and it does not all arrive at once — payment matching improves within weeks, month-end takes two quarters because it depends on the upstream records being trusted first. Note also that "recovered" means those hours go to other work rather than that anybody is dismissed; if the case depends on headcount reduction, say so explicitly rather than implying a saving that will not appear in the payroll.
The three that do not improve
-
Anything requiring judgement
Deciding whether a slow-moving line should be discounted or returned, whether a supplier's excuse is credible, whether a customer gets credit. These are the tasks worth your best people's time and no system reduces them — it just gives them better inputs.
-
Bank reconciliation
There is no statement feed, no import and no auto-matching here, so this stays exactly as manual as it is today. If it is one of your six timed tasks, take it out of the benefit column — it belongs in the unchanged column, and a case that says so is a case that survives review.
-
Statutory filing
PAYE, NSSF and SHA figures compute, and somebody still files them. eTIMS transmits invoices and POS sales for Kenya and does not transmit credit notes. The preparation gets faster; the filing does not disappear.
And one task that gets slower
Recording a movement. Writing "12 out" in a book takes four seconds; entering a check-out with an item, a location, a quantity and a reason takes forty. That is a real cost, paid by the person at the sharp end, and it is the reason adoption fails when nobody has explained what the extra thirty-six seconds buys. Say it out loud: it buys the twenty enquiries a day that nobody has to interrupt them for.
How to time them properly
Two weeks, five minutes a day
- Time the task itself, not your memory of it. People underestimate routine work by roughly half, consistently, in both directions from what you would expect.
- Count the interruptions with a tally, not a stopwatch. Five marks on a sheet by the storekeeper's till. The count matters more than the duration.
- Use the actual person's cost, not an average. The tasks on this list are done disproportionately by trusted senior staff, which is precisely why they are expensive.
- Note which tasks are done twice because nobody trusts the first version. That duplication is a separate cost and it is usually the fastest thing to remove.
- Write down what you would do with the time. A saving with no destination is not a saving — it is a reallocation nobody made, and by month four it has silently filled with something else.
Our take
Time six tasks for two weeks and use the real person's cost, and you will find the largest line is the one nobody complains about — interruptions for stock enquiries. Expect to recover about two-thirds rather than all of it, take bank reconciliation and statutory filing out of the benefit column entirely, and be explicit that recording a movement gets slower before anything gets faster. That last sentence is what makes the case honest and the rollout survivable.
Anything above that you need, we can build for you
Everything listed above as something we can add describes what ships in the standard product today — it is a starting point, not a limit on what AWRA OpsHub can do for your organisation. Kenya's eTIMS integration and its maintained payroll engine are both in the product because clients needed them and commissioned them; neither appeared by itself, and the same door is open for whatever you just read about. One qualification so this is worth what it claims: a small number of things on this blog we deliberately leave to a specialist rather than build — a statutory ledger we will not sign our name to, a rule that would decide a tax question for you, a clinical or member-funds record that belongs in a regulated system — and where that is true the post says so in those words. Everything else is a scope, a timeline and a price.
The operational work, which is what most commissions actually are
An extra approval stage in a chain that does not match the standard one, a custom field set on employees or assets that only your sector needs, an expiry that has to block an order rather than send an email, a report your board asks for in a shape nothing produces, or a scanner or weighbridge feeding the goods-in door. These are the commissions we are asked for most often and the smallest ones we quote — and unlike a revenue-authority pipeline, none of them waits on a regulator.
The module-shaped additions, which are the ones readers ask for most often
A price list with real discount authority, a customer-facing quotation that expires, a bill of materials or recipe costing, a staff advance that is issued, acquitted and chased, a member or unit ledger, a matching rule that holds a payment. Each of these is a build rather than a setting, and each has been quoted before — a bigger piece of work than a custom field, with a written spec and a date instead of a roadmap slide.
The report, document or pack nothing currently produces
The board pack in the shape your board actually asks for, a donor or funder layout, an invoice or receipt template carrying what your regulator or your customer expects, a dataset the report builder cannot reach yet. Usually the fastest thing on this list to deliver, because the data is already in the system.
Systems, rails and hardware you already run
The accounting package, CRM, online store, core banking or custom database you intend to keep — connected through our API so a fact is entered once and appears everywhere it is needed. Plus the physical edge: a scanner, a scale, a weighbridge or a till peripheral feeding the door it belongs to.
How it works: you describe the requirement, we return a written scope, timeline and cost, and once agreed it is built into your environment and maintained as part of the product. Nothing here waits on a regulator or a published specification, which is why operational builds are the ones we quote fastest. Tell us the requirement that would otherwise rule us out — that is a better first conversation than a demo.
Tell us what your operation needsTime six tasks before you decide anything
We would rather look at your six timed figures than show you a calculator. Two of them will not improve at all, and knowing which two before you buy is worth more than the demo.
See plans & pricingFrequently asked questions
What do manual processes cost a Kenyan SME?
Around KES 750,000 a year of staff time for a forty-staff distributor across six routine tasks: matching mobile-money receipts, answering stock enquiries, month-end reconciliation, chasing approvals, finding documents and rebuilding the same report. Roughly two-thirds of that is recoverable, not all of it, and the improvements arrive at different times.
Which manual task is most expensive?
Answering "do we have any?" — about twenty interruptions a day at four minutes each. It is the largest line and the one nobody complains about, because answering feels like being helpful. It also costs more than the raw minutes, since each interruption breaks a count or a receipt, and an interrupted storekeeper starts answering from memory rather than from the shelf.
Which tasks do not get faster?
Anything requiring judgement, which is the work worth your best people's time anyway. Bank reconciliation, because there is no statement feed, import or auto-matching — if it is on your timed list, put it in the unchanged column. And statutory filing: PAYE, NSSF and SHA figures compute and somebody still files them, and eTIMS does not transmit credit notes.
Does anything get slower?
Yes. Recording a movement takes about forty seconds instead of four, and that cost is paid by the person at the sharp end. It is the main reason adoption fails when nobody explains the trade — so explain it: those thirty-six extra seconds are what remove the twenty daily interruptions somebody else was making to ask what is in stock.
How should we time these tasks?
Two weeks, five minutes a day. Time the task rather than your memory of it, because people misjudge routine work by roughly half. Tally interruptions rather than timing them. Use the actual person's cost, since these tasks fall disproportionately on trusted senior staff. And write down what you would do with the recovered time — a saving with no destination quietly fills with something else by month four.
