Work Understanding - Common misinterpretations
Objective. Give Sales and CSM a quick-reference catalogue of the metric misreads specific to Work Understanding, so a rep can catch the wrong reading in the moment and state the correct one, grounded in the app's own interpretation checks.
Work Understanding infers process shape and time loss from signal, which makes it unusually easy to over-read. This is the catalogue of its common misreads, grouped by theme: inference read as fact, friction read as proven waste, conformance read as a quality grade, waiting confused with handling, and opportunity reported as result. For each: the wrong reading, why it is wrong, and the correct reading. Use it as a pre-demo refresher and an in-call recovery tool.
See it in the demo
Theme 1 - Inference read as fact (discovery, variants, detection)
Misread A - 'This is the process.'
Wrong reading: the discovered workflow is the authoritative map of how the process works.
Why it is wrong: the flow is inferred from captured signal, not a definitive map, and coverage may be partial. The customer still owns what 'correct' means.
Correct reading: an evidence-based draft of how work actually moved, to confirm with the customer. The gap between it and their understanding is itself a finding.
Misread B - 'A detected work item is a fact.'
Wrong reading: a labelled session is certified truth, so skip the confidence.
Why it is wrong: detection is a best estimate with a confidence; everything downstream inherits that confidence.
Correct reading: check confidence and coverage first - high confidence on good coverage can anchor a finding; low confidence means treat it as provisional.
Misread C - 'These variants are errors.'
Wrong reading: every variant off the main path is a deviation to eliminate.
Why it is wrong: variants can be legitimate handling of genuinely different cases.
Correct reading: variants show the real spread of how work is done; investigate which ones add time, not all of them.
Theme 2 - Friction and rework read as proven waste
Misread D - 'You found four frictions of X hours, so that's 4X saved.' (the app's core interpretation check)
Wrong reading: sum the friction list into a single savings number.
Why it is wrong: overlapping findings may touch the same time, and fixing one can absorb another. Each finding is an observed contribution with a confidence, not proven waste, and the headline is opportunity, not realised saving.
Correct reading: de-duplicate, treat each as a candidate, and let savings come only from measured before/after on validated findings.
Misread E - 'Friction equals waste.'
Wrong reading: every detected friction is time that can be removed.
Why it is wrong: the view does not prove the friction is waste or that removing it releases the stated time.
Correct reading: friction is a hypothesis to validate; only an experiment shows whether removing it actually releases time.
Misread F - 'Rework count is a quality score.'
Wrong reading: read a rework or handoff count as a direct measure of how bad the process is.
Why it is wrong: some rework and some handoffs are necessary; the count is bounded by capture and does not establish cause.
Correct reading: rework and handoffs localise where time may be lost - a prompt to investigate, not a verdict.
Theme 3 - Conformance read as an absolute grade
Misread G - 'Is 72% conformance good or bad?'
Wrong reading: treat a conformance percentage as an absolute process-quality score.
Why it is wrong: conformance is measured against a chosen reference path, and the reference choice drives the number - change the reference and the same process scores differently.
Correct reading: state the reference path, agree it with the customer, then investigate the deviations rather than grading the number.
Misread H - 'Low conformance means non-compliance.'
Wrong reading: low conformance proves the process is broken or non-compliant.
Why it is wrong: deviation can be correct handling of genuinely different cases (in RCM, different payers force different routes).
Correct reading: low conformance is a prompt to ask which deviations add days, not a compliance grade on its own.
Theme 4 - Waiting, step time and capture-gap confusions
Misread I - 'The longest active step is the problem.'
Wrong reading: attack the step with the longest active handling time.
Why it is wrong: the elapsed days usually sit in waiting (for information, a payer, an approval), not in active handling. Speeding up a non-constraint step barely moves cycle time.
Correct reading: separate active handling from waiting, and target where the elapsed time actually concentrates - usually between steps.
Misread J - 'This step is fast.'
Wrong reading: a step showing almost no time is efficient.
Why it is wrong: step time depends on capture coverage - a step no agent sees looks short or absent.
Correct reading: confirm the step is genuinely fast and not simply unobserved before celebrating.
Misread K - 'All waiting is waste.'
Wrong reading: every waiting period is removable delay.
Why it is wrong: some waiting is a payer's or third party's own turnaround and is not controllable.
Correct reading: confirm with the customer which waiting is actually addressable before proposing a change.
Theme 5 - Opportunity reported as result, and overclaimed capture
Misread L - 'Report the opportunity estimate as the quarter's saving.'
Wrong reading: present the modelled opportunity as the realised result.
Why it is wrong: opportunity is a hypothesis; released time is the result. Only a measured before/after on a validated finding, same scope and metric, earns a savings claim.
Correct reading: report measured time released against the customer's own baseline, and keep the opportunity estimate out of the headline.
Misread M - 'Two experiments released 3 hours each, so that's 6.'
Wrong reading: sum experiment results.
Why it is wrong: if two experiments touch the same time, results are not additive.
Correct reading: check overlap and de-duplicate before claiming a combined total; a null result is also a valid, reportable finding.
Misread N - 'The tool sees everything / a manager can drill to one person.'
Wrong reading: overstate capture or access granularity from the demo.
Why it is wrong: capture, aggregation and access are customer configuration and jurisdiction-dependent; overclaiming breaks trust fastest with the exact stakeholders who can block adoption.
Correct reading: state precisely what evidence is shown, at what aggregation, under whose access policy, and escalate the real configuration to product and the customer's security owner.
Interpretation check
The Work Understanding misreads all share one root: reading an inferred, confidence-bounded, reference-dependent hypothesis as a proven fact or a realised result. Discovery and variants are inferred and confirmed with the customer; friction and rework are hypotheses, not proven waste; conformance depends on the chosen reference path; opportunity is never the result - only measured before/after is.
In the field
KnackGlobal RCM: a claim process scoring low conformance (misread G) is explained by payers forcing different routes, and a stacked friction total (misread D) is replaced by one measured before/after on a waiting step - the two corrections that keep the analysis defensible in front of the largest customer.
Practical task
Pick three misreads from this catalogue that you think are most likely in your next real conversation. For each, write the one-line correct reading you would say out loud, and cite the screen you would open to show the honest version.
Self-check
A prospect totals the friction list to a single saving. Which misread is this, and what is the correct reading?
Misread D, the app's core interpretation check. Overlapping findings may touch the same time and each is an observed contribution with a confidence, not proven waste. De-duplicate, treat each as a candidate, and let savings come only from measured before/after on validated findings.
'72% conformance' - what is wrong with asking whether that is good or bad, and what must you always state?
Conformance is measured against a chosen reference path and the reference drives the number, so 72% has no absolute meaning. Always state the reference path, agree it with the customer, then investigate whether the deviations actually add time.
A step shows almost no time. Why is 'that step is efficient' a misread?
Step time depends on capture coverage - a step no agent sees looks short or absent. Confirm the step is genuinely fast and not just unobserved before concluding it is efficient.
What is the single root error behind most Work Understanding misreads?
Reading an inferred, confidence-bounded, reference-dependent hypothesis as a proven fact or a realised result - treating discovery as the definitive map, friction as proven waste, conformance as an absolute grade, or opportunity as the saving.
Why is overclaiming capture (misread N) the most damaging misread to make in a demo?
Capture, aggregation and access are customer configuration and jurisdiction-dependent. Overstating them breaks trust fastest with employees, works councils, security and legal - exactly the stakeholders who can block adoption - and it is unrecoverable once said.
Verified vs demo. All screens referenced are demo-shown on seeded data; finding counts, conformance figures, step durations and confidence values shown are SME-validation-pending. The misreads and correct readings are the app's own interpretation discipline; exact detection method, coverage and privacy configuration for a customer are confirm with product.