How to Establish a Baseline Before Improving

To establish a baseline before improving, you measure the process as it currently runs, for long enough that normal variation is visible. Without it, any result can be explained away. Why improvements go unproven The measurement usually starts after the change. * No comparable figure from before * Improvement claimed and disputed

How to Set Realistic SLA Targets for Your Team

To set realistic SLA targets, start from what your process currently does rather than from a number someone chose. A target detached from capacity is missed consistently and then ignored. Why arbitrary targets fail A target nobody meets stops functioning as a target. * A figure borrowed from an industry benchmark

Which Metrics to Track for Contract Management

Choosing metrics for contract management means measuring exposure rather than volume. The number of contracts matters far less than what is about to expire and what renews automatically. What goes unmeasured Contract registries usually track existence and nothing else. * No view of what expires in the next quarter * Automatic renewals

Which Metrics to Track for a Vehicle Fleet

Choosing metrics for a vehicle fleet means moving from totals to comparisons. Total fuel spend tells you what the fleet costs; per-vehicle comparison tells you which vehicle is the problem. Why totals hide the answer A fleet total averages away the vehicle that is costing you money. * Fuel spend

Which Metrics to Track for Internal Service

Choosing metrics for internal service matters more than for external support, because internal teams usually have to justify their existence with data nobody collected. Why internal service goes unmeasured It is treated as overhead rather than as a service with a cost and a value. * Volume unknown, so resourcing is

Which Metrics to Track for Recruitment

Choosing metrics for recruitment means measuring the pipeline rather than the outcome. Hires made is the result; where candidates are lost is the thing you can act on. Why hires made is not enough It arrives too late and explains nothing. * Strong candidates lost to delay rather than to a

Which Metrics to Track for Stock Operations

Choosing metrics for stock operations starts with an observation: a quantity on its own prompts no action. It only becomes actionable against a threshold. Why quantity alone is not a metric Knowing you hold forty units answers nothing without knowing what forty means. * Stockouts discovered when an order cannot be

Which Metrics to Track for Project Delivery

Choosing metrics for project delivery means finding figures that warn you before a deadline slips. Completion percentage rarely does, because it is usually an estimate. Why percentage complete fails It is reported by the person least incentivized to report it pessimistically. * Progress estimated rather than derived * Ninety percent complete for

Which Metrics to Track for a Purchasing Process

Choosing metrics for purchasing should include one figure most teams never capture: how much comparing quotes actually saved. Without it, the comparison work looks like bureaucracy. Why purchasing is judged on the wrong things Procurement is usually measured on speed and complaints rather than on value. * Requests counted, savings unmeasured

Which Metrics to Track for a Sales Pipeline

Choosing metrics for a sales pipeline means distinguishing leading indicators from trailing ones. Closed revenue describes the past; stage distribution describes what is coming. What closed revenue does not tell you By the time it moves, the decisions that produced it are months old. * No warning that the pipeline is

Which Metrics to Track for a Support Operation

Choosing metrics for a support operation comes down to which figures change what someone does today. Anything that does not is reporting rather than management. Metrics that rarely change behaviour They are reported often and acted on seldom. * Total tickets handled, with no context * Average resolution time across all types

Which Metrics to Track for an Approval Process

Choosing metrics for an approval process means separating volume from behaviour. Counting requests tells you how busy the process is; timing tells you whether it works. What volume alone hides A count of approvals says nothing about the experience of waiting for one. * Requests approved quickly and slowly, averaged into