Aristo's read Grounded in page evidence

Recoverable licence opportunities are ranked while their evidence remains attached.

Why it matters: The page turns savings potential into a governed action queue.
Recommended action: If a defensible opportunity is counted, assign the highest-value item and track it through closure.

License chargeback

Where spend can be reduced without damaging productivity — each position counted, evidenced, and carrying the risk of acting on it.

Generate report Export

Recoverable, from your own recordscounted

$43,200

$518,400 a year · across 2 counted positions

no prior period held

Source: Microsoft licence records + Copilot usage signals As of: 2026-08-29

Recoverable, on a modelled priceassumption

—

nothing on this page rides on an assumed price

no prior period held

Source: Microsoft licence records at this product's default seat assumption As of: 2026-08-29

To spend deliberatelyestimate

$33,600

a capped pilot for the cohort assessed as ready — never subtracted from, or added to, a saving

↑ 0.2% vs July 2026 — the cohort's own movement

Source: Aristo readiness assessment over counted adoption signals As of: 2026-08-29

$43,200 a month sits in positions read out of your own records — $518,400 a year — across 2 counted positions.

A cost page is where the pressure to name somebody is highest, and the answer is still no. There is no list of who holds a cold seat and no ranking of the least active people. 1 team is withheld from the counted-utilisation column for the same reason.

How the floor holds on a cost page

Every quantity on this page is a count of a group: an estate total, or a team of five people or more. There is no export that produces one either — not because it is switched off, but because this product does not compute it.

There is no grand total on this page, and its absence is the point.

Why adding any two of them cannot be defended

The first figure is money read out of your own records. The second rides on a price this product supplied because you have not set yours. The third is money to spend rather than money to save. Adding any two of them produces a number that reads as precise and cannot be defended line by line, which is exactly the number a finance team will be asked to defend. 1 further position is counted in seats and deliberately carries no dollar at all.

What is worth the most

Recoverable positions ranked by monthly money. A ranking, never a total.

Recoverable positions, ranked by what they are worth a month

current position US$ per month Counted

Put the 1,280 paid seats that are doing no work back to work

From your own records · 1,280 paid seats, held by people

$38,400

Assign or release the 160 Copilot seats that have never been handed to anyone

From your own records · 160 seats, held by nobody

$4,800

Priced from your own recordsPriced on this product’s default seat assumption

Evidence & data — Recoverable positions, ranked by what they are worth a month

Comparison

These bars are a RANKING and never a total: the ones on your own price and the ones on a modelled price are not addable, which is why the two subtotals above are reported apart. 1 further position is counted in seats and carries no dollar, so it has no bar rather than a bar at zero.

Source & method

Microsoft licence records and Copilot usage signals, per position.

Basis: Counted

Accessible data

Recoverable positions, ranked
Position	Per month	Per year	Priced from
Put the 1,280 paid seats that are doing no work back to work	$38,400	$460,800	Your own records
Assign or release the 160 Copilot seats that have never been handed to anyone	$4,800	$57,600	Your own records

The rows are ranked, not summed. Two of these prices come from different places and the page reports their subtotals apart for that reason.

How strongly each is evidenced

Every position raised this run, counted once.

How strongly each position is evidenced

this run positions Counted

Strongly evidenced6 positions (85.7%)Reasonably evidenced1 position (14.3%)

Evidence & data — How strongly each position is evidenced

Comparison

Every position raised this run, counted once. Confidence here is about the EVIDENCE behind a position, not about how likely somebody is to act on it — each row states its own reason beneath it.

Source & method

The counted signals behind each position, assessed at composition.

Basis: Counted

Accessible data

Positions by how strongly they are evidenced
Evidence	Positions	Share of positions
Strongly evidenced	6	85.7%
Reasonably evidenced	1	14.3%

The opportunities, ranked

Each one states what it is, what it is worth, who it touches, what proves it, what could go wrong, who owns it and what to do next.

7 positions

Recoverable high confidence counted

Put the 1,280 paid seats that are doing no work back to work

$38,400 per month recoverable · $460,800 a year

These seats are assigned to somebody and show no qualifying Copilot activity this period. The money is already committed and the licence is already paid for; what is missing is the work.

New Owner: Unassigned Open chargeback History & status →

Impact, risk & evidence

Impacted: 1,280 paid seats, held by people. Aristo reports groups, never a person: there is no per-seat, per-person view of this anywhere in the product.
Risk: A quiet seat is not a wasted seat until somebody checks. Reclaiming first hits the people who were three weeks from adopting; Aristo's ranked move is activation, which turns committed spend into work and costs nothing extra to try.
Evidence: The License Report — every seat, counted

Source: Microsoft licence records joined to Copilot usage signals As of: 2026-08-29

Recoverable high confidence counted

Assign or release the 160 Copilot seats that have never been handed to anyone

$4,800 per month recoverable · $57,600 a year

These seats were bought and never assigned to a person, so nobody loses anything either way. They are the cleanest position on this page: assign them to the people waiting, or release them at renewal.

Not tracked as an object See the product position →

Impact, risk & evidence

Impacted: 160 seats, held by nobody. Aristo reports groups, never a person: there is no per-seat, per-person view of this anywhere in the product.
Risk: Releasing a seat at renewal is not reversible at the same price. If a hiring plan needs them in the next quarter, assigning is the cheaper of the two right answers.
Evidence: Product utilisation — bought against assigned

Source: Microsoft licence records (Organization.Read.All) As of: 2026-08-29

Recoverable high confidence counted

Data & Analytics uses 46% of the seats it holds

$3,900 per month recoverable · $46,800 a year

130 of Data & Analytics's 240 counted seats show no activity this period, against the 60% utilisation floor this product's Traffic Lights use by default. This is where the estate-wide cold-seat position above actually sits.

New Owner: Unassigned Open chargeback History & status →

Impact, risk & evidence

Impacted: 130 seats in this team. Aristo reports groups, never a person: there is no per-seat, per-person view of this anywhere in the product.
Risk: A team below the median is not necessarily a team wasting money — a team that adopted last month looks exactly like a team that never will. The counted movement between two runs is what tells them apart, and it is on the Analytics variance table.
Evidence: Analytics, scoped to Data & Analytics

Not added to any total. This is part of the cold-seat position above, cut by team — not money on top of it. Adding the team rows to the estate row would count the same seat twice.

Source: Microsoft licence records joined to your directory and to Copilot usage signals As of: 2026-08-29

Recoverable high confidence counted

Supply Chain uses 57% of the seats it holds

$2,700 per month recoverable · $32,400 a year

90 of Supply Chain's 210 counted seats show no activity this period, against the 60% utilisation floor this product's Traffic Lights use by default. This is where the estate-wide cold-seat position above actually sits.

New Owner: Unassigned Open chargeback History & status →

Impact, risk & evidence

Impacted: 90 seats in this team. Aristo reports groups, never a person: there is no per-seat, per-person view of this anywhere in the product.
Risk: A team below the median is not necessarily a team wasting money — a team that adopted last month looks exactly like a team that never will. The counted movement between two runs is what tells them apart, and it is on the Analytics variance table.
Evidence: Analytics, scoped to Supply Chain

Not added to any total. This is part of the cold-seat position above, cut by team — not money on top of it. Adding the team rows to the estate row would count the same seat twice.

Source: Microsoft licence records joined to your directory and to Copilot usage signals As of: 2026-08-29

Recoverable high confidence counted

Manufacturing uses 60% of the seats it holds

$2,550 per month recoverable · $30,600 a year

85 of Manufacturing's 210 counted seats show no activity this period, against the 60% utilisation floor this product's Traffic Lights use by default. This is where the estate-wide cold-seat position above actually sits.

New Owner: Unassigned Open chargeback History & status →

Impact, risk & evidence

Impacted: 85 seats in this team. Aristo reports groups, never a person: there is no per-seat, per-person view of this anywhere in the product.
Risk: A team below the median is not necessarily a team wasting money — a team that adopted last month looks exactly like a team that never will. The counted movement between two runs is what tells them apart, and it is on the Analytics variance table.
Evidence: Analytics, scoped to Manufacturing

Not added to any total. This is part of the cold-seat position above, cut by team — not money on top of it. Adding the team rows to the estate row would count the same seat twice.

Source: Microsoft licence records joined to your directory and to Copilot usage signals As of: 2026-08-29

Recoverable high confidence counted in seats

472 idle seats on 5 other paid products

— counted in seats, deliberately not dollarised

Counted from the same licence read as everything else: seats bought on non-Copilot products and never assigned. Aristo holds no price for these products, so the headroom is real and its value is not knowable here.

Not tracked as an object See the product inventory →

Impact, risk & evidence

Impacted: 472 idle seats, unpriced. Aristo reports groups, never a person: there is no per-seat, per-person view of this anywhere in the product.
Risk: None from looking. The risk is the opposite one: a product nobody prices is a product nobody reviews, and these seats renew on their own.
Evidence: Every paid product, in counted seats

Not added to any total. No price backs these seats, so they add nothing to either subtotal. They are counted here so the position is visible rather than absent.

Source: Microsoft licence records — the full paid-SKU inventory (subscribedSkus) As of: 2026-08-29

Deliberate spend medium confidence estimate

Fund the 1,120 people who are ready to consume, with a cap

$33,600 per month, to spend deliberately · $403,200 a year

A capped $33,600 a month would fund the cohort assessed as ready for metered consumption, and only them. The alternative is not "no spend" — it is unbudgeted spend spread across everyone, which is what a cap exists to prevent.

Not tracked as an object Set the monthly cap →

Impact, risk & evidence

Impacted: 1,120 people assessed as ready. Aristo reports groups, never a person: there is no per-seat, per-person view of this anywhere in the product.
Risk: Funding without a cap is the risk this row exists to close. Funding a cohort that has not been assessed is the other one: the assessment is what makes this a pilot rather than a spend.
Evidence: Budget planning — set the cap

Not added to any total. This is money to spend deliberately, not money to recover. It is never added to any saving on this page, in either subtotal.

Source: Aristo readiness assessment over counted adoption signals As of: 2026-08-29

How every figure on this page was arrived at, and what this page cannot see

Every position on this page is a figure the Overview already states. This page ranks and explains them; it re-derives nothing. A leader who reads one number on the summary and a different number one click deeper has stopped trusting the product.

Where a position already exists as an insight object, this page shows THAT object's status, owner and action and links to its history. It creates none: a second place that minted objects would put two documents with two histories behind one sentence.

The two subtotals are never added. One is dollars read out of your own Microsoft records; the other rides on a price this product supplied because you have not set yours. A single "total addressable" across the two would be a number a finance team could not defend line by line, which is the only kind of total worth printing.
The department rows are a BREAKDOWN of the estate's cold-seat position, not additional money. They are excluded from every subtotal for that reason, and each one says so on its face — adding them would count the same seat twice and roughly double the figure.
Duplicate licences, orphaned accounts and redundant products are not detected. Aristo counts seats per product and activity per estate; it does not hold the account-lifecycle signals a duplicate or an orphan is defined by, and inferring one from a quiet seat would name a person's account as waste on an org-wide screen.
Unused agents are not costed. The agent register counts what has been built and what is being reused, and an agent carries no seat and no meter of its own — so there is a real count and there is deliberately no dollar attached to it.
An over-provisioned department is reported against the 60% utilisation floor this product's Traffic Lights already use — a number a leader can move per policy — and never against an external benchmark. Cross-organisation spend benchmarks are not something this product evidences, and a ratio against a made-up peer set is the most quotable wrong number a cost page can produce.