5.1 · Purchasing & logistics KPIs

Level 5 · Steering & Specializations

5.1Purchasing & logistics KPIs

Goal: Build an OTIF, service rate, turnover, IRA, savings and transport cost/revenue dashboard with a precise definition and owner.
Estimated time: 12 min · self-paced reading · If you only have 5 min: read “The essentials in 30s” + the diagram.

The essentials in 30 seconds

  • Few but maintained: 5 to 10 indicators reviewed every month beat pages nobody reads.
  • In antagonistic pairs: service with inventory, savings with OTIF, cost with lead time — otherwise perverse effects guaranteed.
  • Defined and owned: each KPI has a locked definition, an owner and triggers decisions.

Mini-case thread — Nadia, buyer at an agrifood SME in Algiers

Nadia handles purchasing at an agrifood SME in Algiers importing packaging and ingredients. Caught between stockouts on one item and overstock on another, she is steering blind. This section shows how she picks five indicators in pairs, locks the OTIF definition with her suppliers and sets up a monthly one-page ritual.

This same case returns in the lesson, diagram, script and 48h challenge — and continues in 5.2, 5.3, 5.4 and 5.5.

1. The indicator pack that matters

Service/lead times: OTD (on time) and above all OTIF (on time AND in full) — both ways: your suppliers' OTIF (reviews 2.5) and yours towards your customers; service rate (demand served without stockout, companion to turnover 3.4).

Inventory: turnover/coverage, IRA (accuracy 3.5), value and share of slow movers. Purchasing: savings (cautious methodology), PR-to-PO processing lead time, share under contract, supplier concentration. Logistics: transport cost/revenue or per unit, picking cost/order, picking error rate, transport claims.

Golden rule: 5 to 10 indicators reviewed every month beat 40 ignored pages — each with an owner and decisions.

2. Opposing pairs and locked definitions

KPIs come in opposing pairs: service AND inventory, transport cost AND lead time, savings AND OTIF/quality (TCO). Managing one indicator in isolation guarantees perverse effects — the graveyard of local optimisation.

Each indicator is defined precisely once and for all: OTIF = on time against confirmed or requested date? tolerance? complete at line or order level? Without a frozen definition, every meeting drifts into scope debates and trends lie.

The dashboard fits on one page: month value, 3-12 month trend, target, variance comment — a decision tool, not a decorative report.

3. Tooling and honest savings

Your P2P and inventory data (if 1.5 is under control) are first exploited in a spreadsheet; Excel and Power BI (site guides) turn raw data into visuals.

Savings is the most manipulable KPI: honest baseline (price actually paid before, or average of bids for new items), actual volumes purchased afterwards, TCO scope (MOQ, lead times, quality deducted), no double counting and no perpetual savings.

A 30-minute monthly ritual: every variance gets an explanation and an action; an indicator that never triggers anything leaves the board.

Common misinterpretation: "more KPIs = better steering". No. Without a locked definition and without an antagonistic pair, every figure is debatable and none drives action. And "savings" with no stated method is not steering, it is communication.
KPIs in balanced pairs Few indicators, but maintained and opposing Service + InventoryOTIF / Service rate+PurchasingSavings ↔ Supplier OTIF+LogisticsTransport cost/revenue ↔ Lead time+FoundationIRA accuracy One page: value, trend, target, comment — one owner per KPI.
Few indicators, but maintained and opposing
Text version of the diagram (SEO / screen readers / translation): 1) Nadia picks 5 indicators in pairs: service with inventory, savings with supplier OTIF, transport cost with lead time. 2) She freezes each definition in writing, assigns an owner and builds a monthly one-page sheet with trend and target. 3) Each month, 30 minutes: every gap gets an explanation and an action. Lesson: few indicators, but maintained and antagonistic, drive decisions.

🔎 Concrete example — The 5-indicator starter pack

An SME starts with: customer service rate, overall coverage in weeks, OTIF of the 5 key suppliers, transport cost/revenue, IRA. Tracked over 12 months with targets.

Reading: the service/inventory duo prevents fake improvement (improving one by sacrificing the other).

Lesson: defining OTIF precisely and measuring error avoids sterile debates and triggers improvement plans.

Section vocabulary

OTIF
On Time In Full: % of deliveries on time AND complete; measured both ways (your suppliers to you, you to your customers) — definition to freeze precisely.

Simple example: Nadia measures OTIF for her two packaging suppliers at confirmed date, agreed tolerance, complete line.

What it is not: counting "roughly on time" with no written rule, differently each month.

Reflex: does my OTIF definition hold in one sentence, with no possible debate?

Service rate
Share of demand served without stockout; ALWAYS read with inventory turnover/coverage — one without the other lies.

Simple example: Nadia reads the customer service rate with coverage in weeks: serving well without inflating inventory.

What it is not: reporting good service achieved by overstocking everything, without looking at coverage.

Reflex: what is the second figure that stops this one from lying?

Savings
Purchasing savings achieved, measured with an explicit methodology (comparison baseline, scope, validation) — otherwise a communication figure, not management.

Simple example: Nadia computes the gap on volumes actually purchased, prior price paid as reference, validated with the accountant.

What it is not: valuing a discount on imaginary volumes never ordered.

Reflex: are my reference, scope and validator written down?

Logistics cost / revenue
Transport (and picking) costs over revenue or per unit: the ratio that makes logistics costs comparable over time.

Simple example: Nadia tracks transport cost over revenue each month to spot drift.

What it is not: comparing raw transport bills month to month with no link to activity.

Reflex: does my ratio let me compare this month with the same month last year?

Dashboard
Monthly one-pager: 5-10 indicators with value, trend, target, variance comment — each indicator has an owner and triggers decisions.

Simple example: Nadia keeps one page: value, 12-month trend, target, comment — one name per line.

What it is not: a 40-page binder nobody opens before the meeting.

Reflex: does each line have an owner and a possible decision?

In practice — Build your purchasing & supply chain dashboard

  1. Choose 5 to 8 indicators in OPPOSING pairs: service rate + coverage; supplier OTIF + savings; transport cost/revenue + lead time; IRA.
  2. Define each one in writing (exact formula, source, tolerance, scope) — the definition will no longer change.
  3. Build the monthly one-page sheet (spreadsheet: value, 12-month trend, target, comment) and assign one owner per indicator.
  4. Set up the ritual: 30 min per month — every variance gets an explanation and an action; a useless KPI goes out.
You steer with one page of KPIs in balanced pairs, precisely defined and reviewed monthly — figures that trigger decisions.

Ready-to-say script — set up the monthly ritual (1 min)

"Each month, 30 minutes, one page: value, trend, target, gap. Everyone presents their indicator: what moved, why, and the action decided. A figure with no action goes out next month. We start with supplier OTIF and coverage."

Say in the monthly meeting, then note owner and action for each gap.

Copy-paste template — one-page KPI sheet

INDICATOR: _______________
DEFINITION (formula + source + tolerance): _______________
OWNER: _______________
TARGET: _______________
MONTH VALUE: _______________ / 12-MONTH TREND: _______________
GAP AND CAUSE: _______________
DECIDED ACTION (who + when): _______________

Small business / tight budget

Simple spreadsheet: 5 lines, 4 columns (value, trend, target, comment). Nadia updates it herself in 30 minutes each month.

Algeria SME

OTIF tracked at confirmed date with agreed tolerance, useful when import and transport lead times vary: compare at constant rule.

Agrifood import

Add coverage in weeks by sensitive family (stockout = production stop) next to the service rate.

Key suppliers

OTIF of the 5 major suppliers displayed and shared in review: factual basis for the progress plan.

✓ Best practices

  • Freeze the OTIF definition with a quantified example: avoid scope renegotiation every month.
  • Show the 12-month trend more than the isolated value: direction matters more than the point.
  • Have savings validated by finance: credibility earned durably.

✗ Common pitfalls

  • Multiplying KPIs: 40 indicators kill attention — 5 maintained ones beat 30 ignored ones.
  • Fanciful savings baseline (inflated quote): it rewards inflating initial bids.
  • OTIF measured with no clear tolerance: everyone computes it their own way and comparisons lie.

Key takeaways

  • Few but MAINTAINED indicators: 5-10 monthly in trend, with owner and decisions.
  • KPIs in OPPOSING pairs: isolation guarantees perverse effects.
  • OTIF both ways: suppliers towards you and you towards your customers.
  • Define precisely once and for all; tool with Excel/Power BI.

48h challenge — apply tomorrow

Write your OTIF definition in one sentence (reference date, tolerance, complete at which level) and pick your 5 indicators in pairs on the model above. Show the sentence to a colleague: do they understand with no question?

Section exit checklist
  • I can cite 3 antagonistic pairs without hesitating.
  • My OTIF definition holds in one sentence.
  • My 5-indicator page is ready for next month.
Self-rating 1-5: do my KPIs come in balanced pairs?Self-rating 1-5: does each KPI have definition, owner and action?

Frequently asked questions

Which five KPIs if I start from scratch?

Service rate (or customer OTIF), overall coverage (weeks), key-supplier OTIF, transport cost/revenue (or per unit), IRA. They cover service, money and data accuracy — simply calculable — then add savings with a clean methodology and others as you mature.

How to measure savings without cheating?

Honest baseline (price paid before or average of bids), actual volumes afterwards × (old−new), TCO scope (MOQ/lead times/quality deducted), no double counting and no perpetual savings, finance validation. Savings on imaginary volumes will never materialise.

Check your understanding

Why must KPIs come in opposing pairs?

Additional resources (optional, to go further)

Level 5 examFresh, harder questions to test what you've learned