TURN OPERATIONAL QUESTIONS INTO WORKING DECISIONS.

Six practical tools for the work behind a better business. Diagnose a delay, agree a handoff, compare approaches and build a plan your team can use.

Choose your tool Download the offline toolkit (ZIP)

Bring a real workflow. Leave with a working document.

Choose the problem you are working on below. Complete the inputs, generate a result, then print it or save an editable copy. Work with your team to check the evidence and agree the next step.

Inputs stay in this page’s memory. They are not submitted to Regal Nexus Group or saved automatically. Reloading clears your work. Save an editable copy before leaving; downloaded files contain your business information, so store and share them deliberately.

Prepared by the Regal Nexus Group editorial team. Toolkit version 1.0 · 21 October 2026.

Find the delay worth investigating

Compare stage timings and backlog to create an evidence-led investigation brief. Use one sequential workflow and the same sample for every stage.

You will create: A ranked delay table, waiting share and next investigation. Read the companion guide.

Define the review

Workflow name

For example, customer onboarding.

Responsible owner

Required.

Review period

For example, the previous four weeks.

Cases completed in this period

Use the same period for all observations.

Sample and measurement method

Record sample size, timestamps used and any missing evidence.

Workflow stages

Use at least two sequential stages. Do not add parallel or overlapping times.

Stage 1

Stage name

Required.

Mean active work per case (hours)

Convert minutes to hours: minutes ÷ 60.

Mean waiting time per case (hours)

Exclude active work and overlapping waits.

Cases waiting now

A snapshot count, measured consistently.

Evidence / suspected cause

Separate what you observed from what you suspect.

Stage 2

Stage name

Required.

Mean active work per case (hours)

Convert minutes to hours: minutes ÷ 60.

Mean waiting time per case (hours)

Exclude active work and overlapping waits.

Cases waiting now

A snapshot count, measured consistently.

Evidence / suspected cause

Separate what you observed from what you suspect.

Agree how work changes hands

Write a shared operating agreement between two teams. Capture what must arrive, who accepts it and what happens when work is incomplete.

You will create: An editable handoff agreement for review by both teams. Read the companion guide.

Choose the simplest workable approach

Compare rules, bounded AI and agent autonomy against the actual task. The screening result explains which assumptions drive the next step.

You will create: An approach recommendation with reasons and evaluation requirements. Read the companion guide.

Describe the task

Task, trigger and intended output

Required.

Decision owner

Required.

Can decisions be explicitly defined?

Required.

Input type

Required.

What freedom is actually needed?

Required.

Evidence and controls

Can acceptable outputs be evaluated?

Required.

Consequence of a wrong action

Required.

Are human approval points defined?

Required.

Are data use and access permissions confirmed?

Required.

Evidence, constraints and unresolved questions

Required.

Build a business case you can explain

Separate released staff time from cash savings. Model total costs, cash ROI, payback and three benefit scenarios in one consistent period.

You will create: A transparent cash business case and sensitivity comparison. Read the companion guide.

The model uses steady cash flows for simple payback. It does not account for the time value of money. See ACCA’s explanation of payback and discounted payback for the distinction.

Scope and period

Proposed workflow

Required.

Business-case owner

Required.

Currency (no conversion)

Required.

Evaluation period (whole months)

Between 1 and 120 months.

Work and time

Cases per month

Required.

Current handling time per case (minutes)

Required.

Expected minutes saved per case

Cannot exceed current handling time.

Share of cases using the improvement (%)

0–100; account for exceptions and adoption.

Fully loaded hourly labor cost

Required.

Released labor value that becomes cash (%)

Use 0 unless avoided overtime, contractor cost or another actual budget reduction is evidenced.

Costs and additional cash effects

Total upfront cost

Include implementation, integration, internal effort, training and transition.

Monthly running cost

Software, model usage, maintenance, monitoring and review.

Other avoided cash cost per month

Exclude labor savings already captured above.

Incremental contribution per month

Use contribution after incremental costs, not gross revenue. Avoid overlap with other benefits.

Have you checked that benefits do not overlap?

Required.

Evidence and assumptions

Record the basis for volume, time, adoption, costs and financial realization.

Compare providers on evidence

Score providers consistently while keeping critical access and ownership requirements separate from weighted preferences.

You will create: A comparison table with scores, critical gaps and a selection record. Read the companion guide.

Buying brief

Project / requirement

Required.

Evaluation owner

Required.

Required outcomes and acceptance criteria

Required.

Provider assessments

Use the same evidence standard for every provider. Critical requirements override scores.

Provider 1

Provider name

Required.

Diagnostic evidence (0–5)

Weight 25%. Relevant investigation method and supporting evidence.

Integration capability (0–5)

Weight 20%. Actual systems, dependencies and recovery.

Testing and controls (0–5)

Weight 20%. Failure tests, monitoring and access controls.

Ownership and handover (0–5)

Weight 15%. Documentation, training and exit arrangements.

Total cost clarity (0–5)

Weight 10%. Full implementation and ongoing cost.

Support and maintenance (0–5)

Weight 10%. Responsibilities, response and change terms.

Required access controls evidenced?

Required.

Required ownership / handover rights confirmed?

Required.

Supporting evidence and unresolved gaps

0 = no evidence; 5 = strong, directly relevant verified evidence. Name the documents or references used.

Provider 2

Provider name

Required.

Diagnostic evidence (0–5)

Weight 25%. Relevant investigation method and supporting evidence.

Integration capability (0–5)

Weight 20%. Actual systems, dependencies and recovery.

Testing and controls (0–5)

Weight 20%. Failure tests, monitoring and access controls.

Ownership and handover (0–5)

Weight 15%. Documentation, training and exit arrangements.

Total cost clarity (0–5)

Weight 10%. Full implementation and ongoing cost.

Support and maintenance (0–5)

Weight 10%. Responsibilities, response and change terms.

Required access controls evidenced?

Required.

Required ownership / handover rights confirmed?

Required.

Supporting evidence and unresolved gaps

0 = no evidence; 5 = strong, directly relevant verified evidence. Name the documents or references used.

Plan an AI pilot with boundaries

Define the operational outcome, data permissions, allowed actions, tests and recovery plan before live implementation.

You will create: A pilot plan, evidence record and review gate. Read the companion guide.

Outcome and accountability

Workflow and pilot boundary

Required.

Pilot owner

Required.

Success metric

Required.

Current baseline and measurement source

Required.

Target and measurement period

Required.

Person who authorizes the pilot

Required.

Data and action boundaries

Permitted data, systems and provider handling

Required.

Allowed reads and writes

List exact systems and actions; identify prohibited actions.

Human approval points

Specify who approves, what they see and what cannot proceed without approval.

Are data permissions confirmed?

Required.

Are approval boundaries confirmed?

Required.

Test and recover

Representative test cases

Include normal work, missing/conflicting data, duplicate triggers, failures and untrusted instructions.

Test outcome

Required.

Observed results and evidence

For an untested plan, state what will be measured. Do not invent results.

Stop conditions and monitoring owner

Required.

Rollback and manual fallback

Who stops the pilot, restores state and reconciles incomplete work?

Are recovery ownership and fallback confirmed?

Required.

Review date

Required.

Proposed next decision and remaining conditions

Required.

Turn the plan into a tested improvement.

Bring the result to an assessment. We can discuss the evidence, the operational constraint and whether a diagnostic or implementation is the right next step.

Request an assessment

Explore all guides · Explore our services