The cost of a business question that takes five days to answer is larger than the time analysts spend preparing the answer. The organisation also carries the cost of delayed decisions, a longer period of exposure to a problem, and opportunities that may expire while teams reconcile the evidence.
That cost should be measured, not guessed. A slow answer is not automatically expensive; it becomes expensive when the delay changes a decision or leaves a material issue unresolved.
What actually happens during a five-day investigation?
Consider a hypothetical finance team asked to explain an unexpected supplier-spend variance. The team exports ERP transactions, requests payment records, looks up contract terms, reconciles duplicate supplier names, checks credit notes and asks procurement to validate exceptions.
- Start Question raised
- Collect exports
- Reconcile identifiers
- Apply business rules
- Validate exceptions
- Outcome Decision made
Illustrative. Elapsed time can include both active work and waiting time — different operational constraints.
The elapsed time can include both active work and waiting time. These are different operational constraints.
| Cost category | What to measure | Why it matters |
|---|---|---|
| Analyst effort | Hours spent collecting, cleaning and reconciling | Indicates recurring manual workload |
| Waiting time | Time waiting for source owners or approvals | Reveals process bottlenecks |
| Decision latency | Time from question to decision-ready evidence | Shows how long the business operates without clarity |
| Exposure window | Time a potentially material issue remains unresolved | Relevant to leakage, risk or service failures |
| Rework | Repeated investigations caused by inconsistent definitions or missing lineage | Indicates whether the process is repeatable |
| Opportunity cost | Decisions delayed or options lost | Requires careful evidence, not a generic multiplier |
How to build a defensible business case
Do not multiply every delayed hour by an arbitrary financial rate. Instead, start with observable facts.
- Select a recurring question with a named business owner.
- Record how often it occurs and the elapsed time to answer it.
- Separate hands-on effort from time waiting for information.
- Identify the decision that follows the answer.
- Determine whether an earlier answer would have changed the decision.
- Estimate financial exposure only where transactions or outcomes support the calculation.
- Re-measure after a process change using the same definition.
A simple baseline table is often enough to start:
| Measure | Baseline to capture |
|---|---|
| Frequency | Number of investigations per month |
| Time to answer | Median and range, not just the best case |
| Active labour | Hours per investigation |
| Rework rate | Share requiring correction or a second pass |
| Decision owner | Person or function responsible for acting |
| Outcome | Action taken and whether the result was verified |
Where cross-system intelligence can help
If most of the delay comes from repeatedly joining the same sources, resolving identities and reconstructing the same definitions, a reusable cross-system intelligence layer may reduce repeated preparation. If the delay comes from unclear ownership or slow approvals, software alone will not fix it.
This distinction is essential. Technology can improve evidence access and repeatability, but it cannot make an organisation act on information it has not assigned anyone to own.
Qwry.AI’s platform is intended to make data from existing sources easier to investigate with traceable, query-backed answers. Whether it improves a particular workflow must be established through a bounded use case and measured against a baseline.
Next step: Choose one slow, recurring business question and measure its current elapsed time, labour, rework and downstream decision. That baseline is the beginning of a credible business case.
FAQ
Frequently asked questions.
Start with labour, rework, decision delay and the duration of unresolved exposure. Monetise only those elements for which the organisation has credible data.
No. The right turnaround depends on the decision's urgency, materiality and risk. A monthly strategic analysis may reasonably take longer than a payment exception that blocks a time-sensitive action.
Choose a recurring investigation with a clear owner, accessible evidence, a measurable baseline and an action that follows from the answer.
AI can assist with data discovery, query generation and evidence synthesis, but source quality, business rules, validation and human judgement remain important.