devYAN
Copy to my workspace

Thornbury benefits realization review

Thornbury benefits realization review

A finance and operations team reviewing measured benefits

Benefits review Thornbury Components Corbel Partners

Owner: Claire Dubois. Updated Oct 1, 2026.

Due for measurement

Thornbury order-to-cash program 2 benefits not measured Observed, not attributed

Each benefit claimed at scoping, set beside what has been measured since go-live, including the benefits that were never measured. It is the basis on which the next phase is discussed, and an unmeasured benefit is never counted as realized. Prepared by Claire Dubois, engagement partner at Corbel Partners, with Tom Becker, PMO lead, for Robert Lang, chief operating officer at Thornbury Components and program sponsor. Record as of Thursday, October 1, 2026, with readings dated September 15 to September 30. Every latest value is a reading from the client's own source named in the register. No figure on this page is modeled or forecast, and each target is a goal agreed at scoping, not a prediction.

Invoice disputes a month at plant one41down from 64 at May 31; target under 30
Thornbury claims with a baseline14of 17 on the register
Realized2re-keys and order entry error rate
Not measured2cycle time at plants two and three, manual journals

Claims due for measurement fall into two groups: Baseline set claims need their first reading and Measuring claims their next reading, on the dates in the measurement plan. Claims without a baseline appear in the claims table with what was never measured.

Claims against measurements

These are the Thornbury claims that have a reading or a stated gap. Dates in brackets are the baseline date and the date of the latest reading.

Benefit claimedBaselineTargetLatest readingStatusBenefit owner
Fewer invoice disputes at plant one64 a month (May 31)Under 30 a month41 (Sep 30)MeasuringSunita Rao
Order-to-cash cycle at plant one58 days (May 31)45 days51 days (Sep 30)Partly realizedSunita Rao
Fewer manual order re-keys120 a week (Apr 30)Under 40 a week28 a week (Sep 15)RealizedSunita Rao
Order entry error rate at plants one and two6.4 percent (May 31)Under 2 percent1.8 percent (Sep 25)RealizedRobert Lang
Invoice accuracy at first issue91 percent (May 31)98 percent95.5 percent (Sep 28)Partly realizedSunita Rao
Order desk overtime at plant one95 hours a month (Jun 30)Under 40 a month52 hours (Sep 30)Partly realizedRobert Lang
Receivables over 60 days past due$1,840,000 (May 31)Under $1,200,000$1,530,000 (Sep 30)MeasuringSunita Rao
Month-end close duration9 days (Jul 31)6 working days6 days (Sep 4, one close only)MeasuringSunita Rao
Duplicate customer accounts1,260 (Aug 15)Under 100410 (Sep 28)MeasuringRobert Lang
Invoices issued within two days of shipment, plant one72 percent (May 31)95 percent73 percent (Sep 28)Not realizedSunita Rao
Shorter cycle at plants two and threeNone capturedNo target recordedNoneNot measuredRobert Lang
Manual journal entriesNoneUnder 150 a monthNoneNot measuredSunita Rao

Claims with a baseline set but no reading yet, and one claim with neither:

Benefit claimedStatusBaselineTargetNext reviewBenefit owner
Faster credit hold releaseBaseline set30 hours (Sep 15)Under 8 hoursNov 15Sunita Rao
Billing team overtimeBaseline set210 hours a month (Aug 31)Under 120 a monthDec 31Sunita Rao
Orders waiting over one day at plant twoBaseline set47 a week (Jun 30)Under 10 a weekOct 30Robert Lang
Orders shipped over credit limitBaseline set22 a month (Aug 31)Under 5 a monthNov 2Sunita Rao
Credit notes issued within five daysClaimedNone; the three plants log credit notes in separate spreadsheets90 percentOct 23Sunita Rao

Attribution notes

A result is called attributable to the program only when nothing else changed in the same period. Where something did, Claire Dubois records the change as observed and says what else moved.

Invoice disputes at plant one: observed, not attributed

Disputes fell from 64 to 41 a month since the July go-live. List prices also changed in August, so the fall is not attributed to the program alone. Next review Oct 31.

Order-to-cash cycle at plant one: seven days shorter, target not reached

The cycle is 51 days against a 58 day baseline and a 45 day target. Collections staffing changed in the same period, so the seven days are not credited to the program alone.

Receivables over 60 days past due: two collectors started in August

The value fell from $1,840,000 to $1,530,000. Two additional collectors started in August, so the fall is not credited to the program alone.

Order desk overtime and invoice accuracy at plant one

September order volume was lower than usual, so part of the overtime drop from 95 to 52 hours is not from the program. The invoice accuracy gain from 91 to 95.5 percent came mainly from the plant one price table clean-up in July; the new invoice design is not live, so it has not contributed yet.

Two benefits are not measured and are not counted as realized

Shorter cycle at plants two and three. The cycle time baseline was never captured for plants two and three, so this benefit cannot be measured there. Owner: Claire Dubois, with Robert Lang as benefit owner. Next review Oct 15. Consequence: the cycle time claim cannot appear in a next-phase case for those two plants.Manual journal entries. Journals are not tagged by source in the ledger, so manual entries cannot be counted separately. The review planned for Sep 30 was not held; Corbel is waiting for finance to switch on source tags. Owner: Claire Dubois, with Sunita Rao as benefit owner. Consequence: no claim about journals can be made until the tags exist.

Two more readings are weaker than they look. The month-end close duration rests on the August close alone: the review set for Sep 28 was not held because the finance controller was away, and finance wants the September quarter-end close before treating the target as met. The duplicate customer account count of 410 is only partly updated, because the plant three duplicate list is still under review.

Measurement plan

The cycle time gap closes first. Plants two and three have not yet cut over to the new order-to-cash process, so orders at those plants that are paid between October 1 and October 28 still reflect the old process, and their days from order to cash received can serve as the baseline.

✓Baselines recorded where data exists: 14 of 17 Thornbury claims
✓Benefit owner and evidence source named for every claim
Capture the cycle time baseline for plants two and three, orders paid Oct 1 to Oct 28
Reschedule the month-end close review and the journal review that were not held
Measurement reviews, Oct 15 to Oct 31
Benefits review with Robert Lang, Tuesday Nov 17, before any next phase is discussed
Next reviewBenefitReviewed with
Oct 15Shorter cycle at plants two and threeRobert Lang
Oct 20Duplicate customer accountsRobert Lang
Oct 23Credit notes issued within five daysSunita Rao
Oct 30Receivables over 60 days past due; orders waiting over one day at plant twoSunita Rao, Robert Lang
Oct 31Fewer invoice disputes at plant oneSunita Rao
Nov 6Invoices issued within two days of shipmentSunita Rao

How a result is labelled on this page

Observed. The measure moved after go-live and nothing else is known to have changed. It is still reported as a reading, not as proof.

Attributed. Only used when a comparison or a control shows the program caused the change. No result on this page is labelled attributed yet.

Not measured. No baseline or no data exists. It is listed, owned and dated, and never counted as realized.

Who is on this review

RLRobert LangChief operating officer, Thornbury Components, program sponsor
CDClaire DuboisEngagement partner, Corbel Partners, prepared this review
TBTom BeckerPMO lead, Corbel Partners, keeps the measurement plan

Demo client portal. Corbel Partners, Thornbury Components, the people, fees and every figure are fictional sample data. Built on FuseBase.