Model My Context · Resources · Whitepaper

Up to 70% less time per person

A transparent calculation, and why your number differs by processMethod & rationale · EN

The value of time

Time is the one resource you cannot buy more of. Every hour spent retyping data, chasing missing documents and checking things by hand is an hour that does not go to your client, your advice or your growth. Time saving is therefore not a cost-cutting story but a value story: the same people do more of the work that counts, clients are helped faster, and room opens up for judgement and personal contact — exactly what your people are good at. The question is not “how many people can we lose?”, but “how much value is freed up when the friction disappears?”. This document shows how much time a process can realistically save, how we calculate it, and why your own number differs by process.

Where the saving comes from

Every process consists of two kinds of work. There is friction: retyping data, chasing missing documents, checking by hand, waiting on someone else, doing the same thing three times. And there is judgement: assessing, deciding, advising, the personal contact. AI removes the friction. The judgement stays with your people — and that is exactly where you keep control.

That makes the rule simple: the larger the share of friction in a process, the higher the saving. A process that is mostly admin gains a lot. A process that is mostly judgement gains little. There is no universal percentage.

A worked example

Take an illustrative example: handling an incoming application at a service provider. The figures below show the method of calculation; they are not measured client figures. The step in amber is judgement — it stays with the person and does not change.

Step in the processTypeBefore
min
After
min
Retrieve application and retype dataroutine81
Check completeness, request documentsroutine61
Verify data against sourcesroutine102
Assess and decidejudgement77
Draft the rationale and advicemixed72
Record in the fileroutine50.5
Reply to the clientmixed61.5
Update status and adminroutine30.5
Total per application5215.5
Result: from 52 to 15.5 minutes per application — about 70% less time. Note where it comes from: the routine work drops from 32 to 5 minutes, while the judgement (7 minutes of assessing) stays fully intact. The gain is in the friction, not in cutting out human work. How much that is for your process depends on how much friction it holds: an administrative process gains more than one that is mostly judgement. That is why we never calculate “up to 70%” on an example, but on your own process.

That is why the pilot

We don't calculate on an example, but on your own process. In a contained pilot we build a working version in a simulation environment that mirrors your own situation, and we measure the difference:

  1. Together — we pick one concrete process.
  2. You — an intake of about an hour, plus limited additional information.
  3. We — analyse your process and build a working version on a simulation of your own environment.
  4. Together — we measure old versus new on your own situation.
  5. You — receive a personal report with your own saving and the costs, and decide whether to proceed. No obligation.
What you are left with is not a promise but a measurement. A number that is true for your process, your people and your sector — including what the implementation delivers and what it costs.

Start a pilot

Curious what one improved process delivers for you? Start a pilot and see on a real process of your own how much time there is to gain — no obligation, with you in control.
Start a pilot →
Arjan Noordhoek
“A system is not the sum of its parts, but the product of the interactions between them. That is why we don't model the individual steps, but the outcome that matters.”
Arjan Noordhoek — founder & CEO, Model My Context
Download Whitepaper as PDF →

The figures in this document are illustrative and show the method of calculation; they are not measured client results. Your own saving follows from a pilot on your own process.