KOGSY Clinical, our work on post-discharge follow-up workflows, is at kogsycare.com

KOGSYThe sprintScope

What a two week measurement involves

The calculator gives you an estimate from figures you supply. This replaces the estimate with a measurement, for your agency specifically.

01 · The two numbers

What gets measured

Two numbers, and only two. One is estimated from a sample and carries an interval. The other is a complete count and carries none. We say which is which everywhere they appear.

  • Where paid office time goes. Established by work sampling: over ten working days, consenting office staff get pings at random moments inside their own declared hours and press one key to say what they are doing. The share of observations in a category estimates the share of paid time in it, which is then converted to hours and cost at your own payroll figures. Nothing is recorded — no audio, no speech recognition, no transcript. Each ping captures a timestamp, a person code and one number. Reported with a confidence interval, never as a point estimate, and if the response rate falls below 70% we report it as indicative and annualise no cost from it at all. The full protocol is published, including how non-response and the observer effect are handled.
  • Where the schedule and the invoice disagree. For every shift in the window, the scheduled record is compared with what was invoiced, and each disagreement is classified: under-billed hours, over-billed hours, shifts recorded twice, shifts never recorded at all. Both records already exist, so this is a census and not a sample: there is no confidence interval because nothing is estimated. The output is a count and a rate with the working shown, so you can check any individual case.

02 · How

How

Ten consecutive working days. The office staff who consent get nine pings a day at random moments inside their own declared hours, and answer with one keypress. Each call lasts about ten seconds and captures three fields: when, who, and which category. Nobody is asked to fill in a research form, because a research form measures the form rather than the job.

The ping schedule is generated once before day one and stored. It is not drawn at runtime, which is what makes the randomisation auditable rather than a claim.

The billing comparison needs nothing from anybody: the schedule and the invoices already exist, so we read them.

At the end we walk you through the readout in a single session, and you get the document and the underlying counts.

One capability note, stated plainly. The schedule-to-invoice comparison is schedule-time only. It compares what was scheduled against what was invoiced.

It does not cross-check electronic visit verification data, and we are not going to imply otherwise. If your question is whether the caregiver was physically present when they said they were, this measurement does not answer it and you should discount it accordingly.

03 · What you keep

What you get, and keep

Three things, all yours:

  • The baseline document. The two numbers for your agency, dated, with the window and the sample stated.
  • The method. Written out to the standard on our method page, so someone else can repeat the measurement and get a comparable figure, or challenge ours.
  • The raw counts. The underlying per-shift records the figures were computed from, so nothing rests on trusting our arithmetic.

These are yours to keep, use, share with your board, or hand to a different vendor entirely. There is no clause that stops you, and no version of this where the numbers stay with us.

04 · The fee

What it costs

The fee is fixed, quoted in the first call, and small enough that the measurement is the decision rather than the purchase. We do not publish it, because a number on a page invites you to compare it against a scope you have not discussed yet.

05 · If it is small

If the numbers come back small

Then they come back small, we will say so, and the honest recommendation is to do nothing.

That is not a rhetorical flourish and there is no counter-offer behind it. A measurement that talks you out of a purchase has done its job. You keep the baseline either way, and the next time somebody pitches you a percentage you will have your own number to hold it against.

06 · What it is not

What this is not

  • Not a pilot of a product. Nothing is being trialled on you, and the sprint does not end with software you now need to keep paying for.
  • Not a free trial. It is paid, deliberately. Free work is worth what it costs and gets the attention it deserves from both sides.
  • Not a proposal. There is no deck at the end and no second meeting where the real pitch happens. The output is the document.
  • Not an audit. We are not checking your compliance, and nothing in the readout goes to anyone but you.

07 · If two weeks is too much

An hour on your numbers

Documentation Cost Diagnostic · $500 · 60 minutes

Two weeks is not always the right size, and it is a lot to agree to before you know whether anything here applies to you. So there is a smaller version.

One hour, paid, on a call. You bring your own figures: roster, rates, what you think a shift note takes, how often the schedule and the invoice disagree. We fill the model together while we talk, and at the end you have it, with every assumption in it labelled and every number traceable to something you told me.

What you get

The filled model, with every assumption in it labelled and every number traceable to something you told me. The assessment on this site already ranks the eight costs for nothing, and it will keep doing that. What the hour adds is the model underneath the ranking, and somebody to push back on your inputs while it is being built.

Then, within two business days, a one-page Measurement Readiness Memo: the assumptions we used, the records you are missing, and a recommendation that is one of three things — do nothing, prepare first, or run the two-week measurement. Here is what one looks like, filled in for an invented agency so you can read the format before you pay for it.

Best for an owner, operations lead, or finance lead at a US non-medical home-care agency with schedule and invoice records.

Bring your roster, rates, and a rough view of schedule-to-invoice differences.

Do not send client or patient information. Not before, not during, not after. There is nowhere in this that needs it.

The fee is $500 and it is published, unlike the sprint fee, because the scope here is one hour and it cannot move. There is nothing to compare it against that you have not already read on this page.

What it is not

It is not a sales call with a price on it. There is no deck and no second meeting where the real pitch happens. It is also not a substitute for the sprint, and I will not pretend otherwise. An hour replaces your guesses with better-informed guesses. The sprint replaces them with counts. If what you need is a number you can defend to a board or a buyer, the hour will not get you there.

Rescheduling and refunds

Reschedule any time up to 24 hours before the call, by replying to your confirmation email or writing to henrik@gokogsy.com. Cancel at any point before the call and you get a full refund, no reason needed and nothing to argue about.

Book and pay $500 →

Booking and payment run on Google Calendar. You are charged when you book, not on the call.

08 · Is this for you

Who this fits, and who it does not

Best fit: US non-medical home care agencies, roughly 20 to 200 caregivers, private pay or a mix.

What it needs from you: access to your scheduled records and your invoices for one two-week window, and an owner or operations lead who can agree the comparison rules before we start. On the office sampling side, the people taking part have to consent and can stop at any point.

Usually not a fit: if what you actually need is electronic visit verification, a compliance audit, a clinical opinion, or software installed next month. This measures two things and hands you the numbers. It is not any of those.

Four ways in, in increasing order of commitment.

Run your own figures first. The calculator will tell you within a minute whether the order of magnitude is worth anyone's attention. It asks for nothing before it shows you the result.

Work out your own figure →

Not sure which cost is yours? Fourteen questions, a ranked list, no score, and again nothing asked before the result.

Find out which to measure first →

Or just ask. Twenty minutes, free, and it ends in one of three things: this is not a fit and here is why, you have the problem but not yet the records or the person to run it, or here is the window, the records, the people and the fixed fee. Any of those three is a good outcome. I would rather tell you not to buy this than sell it to the wrong agency.

Book twenty minutes →

Booking runs on Google Calendar.

Or pay for an hour. The hour described above is the one option here that costs money, and it is the one that ends with an artefact rather than a conversation. It does not replace the sprint and I would say so on the call.

09 · Questions

Questions we get asked

What is a measurement sprint in home care?

Ten working days, producing two numbers. Where paid office time goes, established by work sampling on consenting office staff and reported with a confidence interval, never as a point estimate. And the count of shifts where the schedule record and the invoice disagree, which is a complete count rather than a sample and carries no interval. Nothing is recorded on any call. The agency keeps the baseline document and the method regardless of whether it buys anything afterwards.