Sharp

Capture Discovery

This review is private. Enter the access code to continue.

Incorrect access code

Sharp Capture Discovery
OverviewThe BottleneckWhat It CostsWhat ChangesProving ItQuestions
Section 02

What it costs, on your numbers

Every field below is editable and every result recalculates as you type. We have put our starting guesses in, but they are guesses. The useful version of this page is the one where you have replaced them — including the version where you replace them and the answer comes out too small to bother with.

Your fleet and volume

Nothing here is measured. Replace every number with yours.

Speeds

Device speed comes off the spec sheet. OCR throughput is the one worth measuring.

Architecture

Engines today versus workers you could run when it is busy.

People and money

The two inputs most likely to be argued about in a boardroom.

Effect one · the device comes back
Held per large job today
—
Held per large job after
—
Paper handling — irreducible Released back to the fleet

Across the fleet that is — of device availability returned per year, or about — handed back to every machine, every day — without buying a machine.

Effect two · the backlog clears

The busiest hour, cleared with the engines you have versus workers you could add.

Last document lands, today
—
Last document lands, after
—

—

Effect three · people stop standing there
Staff hours returned per year
—
At the rate above
—

This is the only figure here we would put in front of a finance person, and even then with the caveat that recovered minutes are only worth money if they get spent on something useful. Treat it as the ceiling, not the forecast.

Sanity check

—

Deliberately excluded

What this model does not count

Some of these would make the case stronger and some would make it weaker. They are left out because we cannot size them honestly yet, and a model with an invented number in it is worth less than a model with a hole in it.

  • Re-scans caused by somebody assuming the job failed. Real, and impossible to size without asking.
  • The knock-on cost of a document arriving after the person who needed it went home.
  • Time spent manually renaming, splitting and filing what arrives. Genuinely large, but it belongs to the second conversation.
  • Capital and licence cost of the current capture server, which would partly offset any new spend.
  • The cost of the proposed platform itself. Until the proof of concept runs, any figure would be invented.
  • Staff goodwill, which is not a number but is the reason this conversation started.
The one input to measure first

Share of OCR time that holds the device.

Set it to zero and effects one and three collapse to nothing — the device was never held, and the only problem left is time-to-document. Set it to one hundred and it is the largest number on the page. We have assumed the worst case because that is what was described, but assumption is the wrong word to have at the centre of a business case.

One person, one stopwatch, one big job. It is the cheapest fact on this site to acquire and the most expensive one to be wrong about.

How we would prove it →
Prepared with the Sharp team Discovery phase — nothing built, nothing committed Private · not for distribution