Plan and measure a small test
Define the starting point, use the same measurements before and after, write pass rules, and review the result after workers have had time to settle in.
Measured results
Customer results should show how they were measured—not only the best number.
We publish a customer result only when the starting point, measurement rules, sample, problems, implementation effort, and limits are clear. No customer case studies are published yet.
Not sure about a term? Open the plain-English glossaryCurrent status
No customer case studies are published yet.
Until a customer-approved result meets the rules below, the Lab provides the worksheets and calculators needed to run a fair trial. Calculator examples explain the math; they are not customer results or performance promises.
Before a result can be published10 required items
Starting operation and layoutSame measurement rules before and afterMeasurement dates and sample sizeEquipment and computer-system setup effortProblems found during the trialOutside changes that may have affected the resultWarehouses where the result may or may not applyResults when workers handled problems or used the backup methodCustomer review and permissionNamed person responsible for updates and correctionsWhat you can use today
Three ways to build a trustworthy result
Calculators that show the math
Prefilled examples show how a calculation works and how the answer changes. Replace them with measured warehouse numbers.
How evidence is judged
Every future result must separate customer data, Voodoo data, calculated estimates, and opinion, and it must state where the result may not apply.
Measure a real trial
Use task, completion, item, order, and problem records to build a stronger before-and-after result.
With customer permission, system records can show what happened more clearly than a testimonial alone.