Value model

Set four numbers for your own team and read the time and dollars they get back. Every default is labeled CITED or ASSUMPTION so you can tell a public figure from a placeholder.
What this page is A value estimate for one team, built from numbers you set. It shares only the figures you enter here. No controlled file, supplier name, or program detail is used. Candidate workspace, mock data only.
Labor-value model illustrative · sample assumptions, adjust to your role · no customer figure · candidate pilot scenario only
Stop hunting for the right version

A reviewer stops losing time to finding the current version and lining up sources by hand, and gets that time back for the work that needs their judgment.

At the public role salary that is worth about $10,800 per person every year. The dials below size it for your own role and headcount. No setup and no new tool to learn.

Value in week one

6 hours/week

per person, reclaimed in week one, no training period.

Today
  • Searching across folders and systems for the current version.
  • Fixing the same fact in five places and missing the sixth.
  • Two people working from two versions, each sure theirs is current.
  • Reading two documents side by side by hand to find what changed.
  • A new team member stuck, unsure what the next step is.
With Halcyon
  • One search finds the current, approved version. The hunt is over.
  • Update a fact once and every place that uses it shows the new value.
  • One agreed version per item, so "which one is current?" stops being a question.
  • The difference is shown for you: what changed, what conflicts, what is missing.
  • A plain-language next step tells the person what to do.
Why this works for you
  • Low effort to start. No software to install and no data to move. It works from the status and summary of what you already have.
  • Pricing comes later. Run a four to six week pilot, measure the hours back, then agree on annual pricing.
  • Value in week one. The first search finds the current version. No ramp and no training period.
  • Stays in your systems. It reads the status and summary and replaces nothing. Your source systems stay where they are.
What gives the time back
  • Stop hunting. Ask once and get the current version across folders and systems. ASSUMPTION · folds into the blended hours-back input · basis: knowledge workers spend about 1.8 hours a day searching for information (McKinsey, "The social economy", 2012)
  • Change once, update everywhere. Update a fact in one place and every place that uses it shows the new value. ASSUMPTION · same blended input · counted once (McKinsey 2012)
  • One agreed version. "Which version is current?" stops costing rework and second-guessing. ASSUMPTION · rework avoided, counted inside the blended hours figure (no separate invented number)
  • See what changed. Open two documents or sources and the difference is shown for you. ASSUMPTION · the first catch on day one, modeled in the slider math (hours times salary divided by 2080)
  • Guides a new team member. A plain-language next step, with no tooling or AI knowledge needed. ASSUMPTION · no standalone hours claim, carried by the single blended input
Size it for your role

Set the numbers. Read one result. Plain math.

The hourly rate is the salary divided by 2080 full-time hours. Every result is that rate times the hours you set. No estimate is hidden in the math.

Sample only
What this pilot is worth
Hours back / person / yr

276 hr

hours back each week times working weeks

Value / person / yr

$10,881

hours times (salary divided by 2080 hours)

Company value / yr

$544,038

value per person times people in this role

Share of paid year reclaimed

13.3%

time share, independent of salary

Full-time roles' worth added

6.6 FTE

capacity, with no new hires

Cross-check

Fully-loaded view (times 1.35): $734,452 per year company.

the conservative base leads, the loaded view backs it up

Payback check

Value back against what you pay. Set your pilot cost.

Pilot cost for year one Not set

Drag to your own pilot price to see the value-back ratio. At $0 the ratio is not shown yet.

ASSUMPTION · your own pilot price · target 10 to 1 or better
Value back for every dollar paid

Set a pilot cost

For every dollar on the pilot, this shows the value back in year one. 5 to 1 or better is the usual mark to go ahead.

Sample assumptions only. Adjust to your org. Salary midpoints are CITED to public Glassdoor and BLS analyst-tier bands. Hours back, working weeks, the loaded multiplier, and headcount are labeled ASSUMPTION and should be set with your real org data. No customer figure is shown. The hourly rate (salary divided by 2080) and every result are plain math.

What we will not claim, so you can trust what we do
What the pilot proves
  • Hours reclaimed per person per week, measured.
  • The first search finds the current version.
  • It works from the status and summary and stays inside your systems.
What it does not
  • No customer figure is implied by these sample numbers.
  • No claim that controlled-data readiness has been assessed.
  • No claim that the assistant closes, approves, or prevents anything on its own.
  • No replacement for any source system.
What to measure next
  • Hours back across more role types.
  • Time to find the current version per source.
  • Stop metric: if hours back stays flat, stop before annual pricing.
Pilot ask

Run one paid proof of value. Measure it fast.

Measure firsthours reclaimed per person per week, value per year, full-time roles' worth added
Scopeone role type, one set of source systems, one manager-safe result packet
Duration4 to 6 weeks
Stop metricif hours back cannot show real movement, stop before annual pricing

Result surface

This value model only frames the pilot. Your deliverable stays the reviewed readiness packet: proof the system works, with a person approving before anything is shared.

Open readiness packet