Identify operational friction points

How can you distinguish proven inefficiencies from impressions to prioritise improvements to a procurement process?

1960 € per session, excl. VAT

In short

A seven-hour day, excluding breaks, to map a purchasing process, locate waiting times, repeated data entry and rework, and then document inefficiencies. Generative AI helps structure observations and formulate hypotheses; participants verify the evidence, recalculate the indicators and produce a process map that operational managers can use.

Why this training?

A purchase request can pass through several departments without anyone seeing all the queues. Confusing processing time with total lead time leads to speeding up a task when the bottleneck actually lies in an approval or an incomplete file. Generative AI makes it easier to synthesise interviews and categorise anomalies, but it can invent steps or attribute a cause without evidence. The approach therefore links each finding to a source and separates symptoms, presumed causes and mandatory constraints. Running case study: The fictitious organisation « Ternovia Outillage » identifies delays between purchase requests and the issuing of purchase orders. The learning pack includes simulated interviews, approval rules and a fictitious log of thirty requests, with timestamps, returns for correction and changes of owner. Transfer to the workplace: Within an authorised scope, reuse the observation template, have the map reviewed by the stakeholders concerned and establish a baseline measurement. Then examine a priority action with the process owner, without announcing any gains before they have been observed.

Objectives of this training

  • Define a process by its trigger, output and stakeholders
  • Build a current-state map with decisions, handovers and loops
  • Quantify waiting times and rework using verifiable records
  • Characterise inefficiencies by distinguishing facts from causal hypotheses
  • Prioritise improvement options using evidence and indicators

Who is it for and what are the prerequisites?

Roles and tasks concerned

  • Purchasing and supply managers responsible for streamlining internal requests
  • Operations and continuous improvement managers leading cross-functional diagnostics
  • Business analysts documenting approval workflows

Prior skills

  • Know an operational or purchasing workflow and its main approval rules.
  • Be able to formulate a structured prompt for generative AI and work with durations in a spreadsheet.

What you will work on

  1. Set the boundaries of the workflow under review

    Objective: Define an observable scope. Trigger, output, beneficiary, exceptions and mandatory rules.

  2. Structure the available records

    Objective: Establish a traceable factual basis. Interviews, timestamps, contradictions and missing data.

  3. Represent how the process actually works

    Objective: Make handovers and rework visible. Map organised by stakeholder, with decisions, queues and correction loops.

  4. Measure friction points

    Objective: Quantify without confusing waiting time with activity. End-to-end lead time, active time, frequency of rework; limitations of the available timestamps.

  5. Characterise and rank

    Objective: Select inefficiencies for further investigation. Symptom, hypothetical cause, necessary check; impact, effort and level of evidence.

  6. Deliver a verifiable diagnosis

    Objective: Prepare for the discussion with the relevant business manager. Consistency between map, evidence and measurements; indicators and arrangements for validation in practice.

What you take away

  • Scope sheet and source register
  • Current-state map annotated with friction points
  • Table of inefficiencies, evidence and verified measurements
  • Prioritisation matrix and indicator tracking sheet

Method and assessment

  • Short instructional inputs followed by immediate practice on the running case study.
  • Individual work followed by work in pairs, with peer review of outputs.
  • Facilitated group review: what works, what needs correcting and what remains a human decision.
  • Preparation: Identify a short process, its boundaries and an observed difficulty, without sharing confidential documents. Arrange access to the tools authorised by the organisation; the exercises use only fictitious or anonymised data.

Observable criteria: The map shows a trigger, an output, the owners and every documented loop. ; Each inefficiency refers to an identifiable source or is labelled « hypothesis ». ; Durations are recalculated without AI, with explicit units and rules. ; Priorities distinguish between impact, effort and strength of evidence. ; Each selected improvement option includes a proposed owner and a defined indicator..

Frequently asked questions

Do you need to be proficient in a process notation?

No. A simple representation organised by stakeholder is sufficient, provided decisions and loops remain explicit.

Does AI determine the causes?

It proposes hypotheses; records and validation in practice allow them to be confirmed or ruled out.

What should you do if times are missing?

Document the gap and plan data collection; do not turn an estimate into a measurement.

Can you remove an approval step considered slow?

Not without checking its regulatory, contractual or control function with the manager concerned.

Last updated: