Skip to content
Findings and Outcomes

What our reviews turn up.

Every number here comes from records the schools already held: sign-in logs, purchase inventories, usage exports, and a privacy review of every vendor they pay.

520connected apps per school, on average, that can reach student and staff accounts
61%of connected apps showed no recorded use
9 in 10connected apps we checked weren't in the school's own records
6%of vendors reviewed rated high risk
Under 2%of paid tools had both a confirmed cost and measured use at review
Outcomes

What changed after the review.

Choose an area to focus on, or see everything.

Security

Access narrowed at no instructional cost.

61%

of connected apps had no recorded use. Each one still held standing access to a school account, and revoking them touches no tool in use.

98%

of app registrations were trusted by default in one review. The fix is an approval pathway, so the list doesn't rebuild.

12

browser extensions in staff use turned up in one review, none in any inventory, and all 12 sending data to an AI backend.

123

accounts were found syncing full mailboxes, calendars, and contacts to devices, the broadest access in that review.

Spending

Renewals decided with the evidence in hand.

$27,471

a year in duplicate tools identified in one review, on confirmed or published pricing alone. The full identified figure was $69,141.

0% to 80%

of one district's paid spend checked against real use, after a single usage export.

$20,475

a year in net cuts a district decided on after its review.

91%

of seats were in use on a contract that looked idle. It stayed. Our numbers protect the tools that work, too.

Compliance

Agreements and policy caught up with what was already in use.

427

accounts had authorized an AI chatbot the school never bought, with no data agreement or use policy. The fix starts with an AI acceptable-use policy.

12+

vendors in one district ran generative or adaptive AI on student accounts without confirmed policy coverage.

5

tools in one review needed a data agreement or consent fix before their next use.

6%

of vendors reviewed rated high risk, and in every review at least one high-risk vendor was also a heavily used or core system.

Renewals

The whole portfolio in front of the board at once.

90.7%

of kept spend moved onto one renewal date in one district, covering 22 of its 31 kept platforms.

$0

for a vendor to move its end date onto the common renewal date, pending the district's approval.

2

renewal dates where vendor records disagreed with the district's own inventory, found during alignment.

Days

for schools to sign the narrow letter that lets us request contract details and date changes, and nothing else.

The finding underneath it all

Cost and use never met.

No school could show its board, from records in hand, whether its largest software contracts were earning their cost. At review, under 2% of paid tools had both a confirmed cost and measured use.

That says nothing about waste. It means renewals were being decided without the evidence, and both missing records already existed: usage in vendor consoles and sign-in exports, cost in purchase orders and invoices.

How the work runs

01Reconcile. Match the school's purchase or curriculum inventory against its sign-in records or usage analytics.
02Score. Rate every paid vendor's privacy and security with the PURPOSE framework, anchored to FERPA, COPPA, and federal AI guidance.
03Authorize. A narrow letter that lets us request contract details and date changes, and nothing else.
04Align. Contact vendors to move every renewal onto one common date, logging every request and reply.
What held in every review

Five patterns, whatever the size or type of school.

01Schools can't see their own portfolio. Inventories were incomplete, split across files, or missing costs, renewal dates, or data-agreement status.
02The largest spend is the least visible. The biggest paid lines often sign in outside the systems schools can export from. In one review, 66% of paid spend sat there.
03Vendor records and school records disagree. Renewal dates and student counts didn't match between what vendors held and what schools held.
04A narrow authorization gets signed. Schools signed a letter that lets us request information and date changes, and nothing else, within days of receiving it.
05Honest savings are smaller than headline savings. Every report leads with its conservative figure. That's what makes the numbers defensible to a board.
What we recommend any district do first

Six steps. None needs new software.

01Revoke apps with no recorded use, and set an approval pathway so the list doesn't rebuild.
02Fill in the cost column for every tracked tool, starting with the 20 largest.
03Fill in the data-agreement column, starting with high-risk vendors.
04Adopt an AI acceptable-use policy, including a rule that no student personal data goes into AI prompts.
05Request usage exports for the largest tools that sign in outside Google, before their next renewal.
06Move renewals onto one date, so the whole portfolio is reviewed in one budget cycle.
What this data can't see

Figures are drawn from our 2025-26 engagements, as of October 2026. Sign-in records show that access was granted, not how often anything is used. Usage was measured only where a platform supplied it. Savings carry their own confidence level (decided, estimated, or identified) and are never added together. Schools, staff, and vendors aren't named.

This is the work behind our Reconciliation Report and Third-Party Risk Assessment.

Want this view of your district? We build it from records you already hold, and we do the analysis.

Start the Conversation →