Knowledge

How to measure outsourcing quality: 12 KPIs that reveal more than SLA alone

Timeliness is only the beginning. Measure quality, security, cost and the supplier's ability to improve the process.

How to measure outsourcing quality: 12 KPIs that reveal more than SLA alone

The contract is signed, transition is complete and the first work has been delivered. The next question matters more than a headline SLA result: is outsourcing actually improving the business? A single timeliness percentage cannot answer it. A useful scorecard combines speed, quality, security, cost and user experience.

Current UK Cabinet Office guidance defines a KPI as a factor or measure used to assess a supplier's performance throughout the contract life cycle. ISO 37500 likewise emphasises continuous monitoring and management of the outsourcing relationship to sustain value for both parties. Measurement must therefore lead to decisions, not merely a monthly report.

Separate KPIs from the SLA

An SLA states an agreed service level, such as processing 98% of items within one business day. A KPI shows how the process performs and whether it is moving in the right direction. Not every KPI needs to be contractual. Some should warn both parties early enough to improve the process before an SLA is missed.

Combine lagging indicators such as timeliness and defects with leading indicators such as incomplete inputs or overdue corrective actions. The first group explains what happened. The second signals what may happen next.

12 useful outsourcing KPIs

1. On-time completion

Measure the proportion of cases completed within the agreed time, split by priority. One overall average can hide delays in time-critical correspondence.

2. Cycle time

Measure from receipt of a valid request to final completion. For mail operations this may cover collection, registration, scanning, routing or dispatch. Report the median and a high percentile because the mean hides the slowest cases.

3. First-time-right rate

The share completed without correction, reprint, misrouting or repeat dispatch. This separates fast activity from correct delivery.

4. Exception rate

The proportion that cannot follow the standard route. Classify causes: source data, client decision, equipment, carrier, transport or execution error.

5. Record and evidence completeness

The share with a complete status, timestamp, owner and required proof. Validate the result with samples rather than relying solely on a system declaration.

6. Incident response and resolution time

Measure acknowledgement, start of work and permanent resolution separately. Closing a ticket is not a resolution if the cause remains.

7. Repeat-problem rate

The proportion of incidents returning for the same root cause. A high result suggests symptoms are being cleared without effective corrective action.

8. Security and compliance

Track unauthorised access, wrong recipients, retention failures, missing audit trails and unapproved channels. Zero reported events is not proof of security, so include access reviews and samples.

9. Availability and continuity

Review agreed-function availability, recovery test results, achieved RTO and RPO, and minimum throughput. Link ongoing oversight to the business continuity plan.

10. Unit cost and total cost

A unit cost is meaningful only when the unit is stable. Add exceptions, rework, materials, transport and retained internal work. See the guide to outsourcing pricing models.

11. Throughput and capacity use

Measure correctly completed units and spare peak capacity. Maximum utilisation is not always desirable because it can signal queues and no resilience.

12. Process-user satisfaction

A short recurring survey of people who submit work and use the output reveals unclear reports, difficult communication or poor escalation context. Ask about a specific step, not general happiness.

Give every KPI a definition card

Record the purpose, formula, data source, owner, frequency, scope, exclusions, baseline, target and warning threshold. Define when the clock starts and when it may pause. Without these rules, both parties can calculate different results correctly.

State the numerator and denominator. Replace “98% timeliness” with “cases completed on time / all valid cases accepted during the month”, then define cancelled, client-blocked and reopened cases.

Keep the main scorecard small

Six to ten measures are usually enough for the primary review. Supporting metrics can remain in the operational analysis. Too many KPIs dilute ownership and create a report in which everything is visible but nothing triggers a decision.

Use a clear rating such as good, approaching target, requires improvement and inadequate. Current Cabinet Office guidance uses a similar scale. A colour must never replace the number, trend and explanation.

Create a service-management rhythm

  • Daily: exceptions, urgent deadlines, incidents and backlog.
  • Weekly: volume, throughput, first-time-right quality and open actions.
  • Monthly: full KPI scorecard, trends, root causes, cost, forecast and decisions.
  • Quarterly: business outcomes, risk, scope changes, automation and improvement roadmap.

Every deviation needs an owner, due date, root cause and a way to verify effectiveness. If the same red KPI is discussed at three meetings without a decision, the weakness is governance, not reporting.

Five common mistakes

  • Measuring speed only. The supplier may move faster at the expense of quality.
  • A target without a baseline. The threshold is arbitrary and detached from the real process.
  • An average without distribution. Serious delays disappear inside a good aggregate.
  • Penalising what the supplier cannot control. Late or defective client data is treated as supplier failure.
  • Never changing the measures. The scorecard remains fixed after volume, technology or business goals change.

Start within 30 days

  1. Select three to five business outcomes outsourcing should improve.
  2. Collect a baseline from the internal process or pilot.
  3. Agree six to ten primary KPIs and definition cards.
  4. Assign data sources, owners and warning thresholds.
  5. Run one trial report and reproduce every result.
  6. Set review dates and a corrective-action log.
  7. After one quarter, remove metrics that never drive a decision.

If transition has not started, use our 30-day outsourcing preparation plan. To design measurable oversight for correspondence, print, inserting or bulk mailing, talk to BackOffice Outsourcing. We start with the business outcome and data source, then choose the measures.

Sources

Frequently asked questions

How many KPIs should an outsourcing arrangement have?

Six to ten measures are usually enough for the primary review. Supporting data can remain operational, but every primary KPI should support a decision.

How is a KPI different from an SLA?

An SLA states an agreed service commitment. A KPI measures process performance. Some KPIs sit inside the SLA, while others provide early warning and improvement insight.

How often should performance be reviewed?

Review urgent exceptions daily, operating results weekly, the full scorecard monthly, and objectives, risks and scope quarterly.

Can KPIs change during the contract?

Yes, through controlled change. Measures should evolve with volume, technology and objectives while preserving enough continuity for trend comparison.

Talk to us

Talk to us
← Back to all articles

Let’s talk about your business correspondence

Verify our company

Check BackOffice Outsourcing’s current credentials on the official services.

Quality, security and business continuity

We work in line with recognised ISO standards that structure our processes and information-security practices.

9001:2015Quality management
22301:2020Business continuity
27001:2023Information security
27002:2023Information security controls