Measuring the Results of Change

Change is not successful merely because a new system has been installed, a policy has been announced or employees have attended training. Real change is visible in behaviour, performance and results. Measuring those results helps leaders determine whether an initiative is working, where adoption is weakening and what needs to be adjusted.

Effective measurement is not about collecting as many figures as possible. It is about selecting evidence that connects the change initiative to the organisation’s objectives. Whether a business is introducing digital payments, restructuring a department, improving customer service or implementing a new health and safety process, the central question remains the same: what has improved, for whom and as a result of what?

What Does Measuring Change Results Mean?

Measuring the results of change means assessing whether an intended change has been implemented, adopted and converted into useful organisational outcomes. It involves more than checking whether planned activities were completed.

For example, a company may report that all staff completed training on a new customer relationship management system. That is an implementation result. It does not prove that staff are using the system correctly, that customer information is more reliable or that service has improved.

A complete measurement approach usually examines four connected levels:

1. Change activity: What was done, such as training, communication, system installation or process redesign?
2. Adoption: Are the intended users applying the new behaviour, process or technology?
3. Performance: Is the new way of working improving efficiency, quality, service or control?
4. Business or organisational outcome: Is the change contributing to a meaningful strategic result?

These levels are related but not interchangeable. Activity is necessary, yet activity alone is not evidence of value.

Start with the Intended Result

Measurement becomes difficult when the change initiative has a vague purpose. A statement such as “improve the organisation” is too broad to guide useful measurement. Leaders should first describe the problem, the desired future state and the people or processes expected to change.

A practical result statement might be: Within six months, branch staff will use the new digital loan process for routine applications, reducing avoidable paperwork and giving applicants clearer updates.

This statement is stronger because it identifies:

• the people affected: branch staff and applicants;
• the behaviour required: use of the digital process;
• the expected operational benefit: less avoidable paperwork; and
• the service benefit: clearer communication with applicants.

Once the intended result is clear, leaders can ask what evidence would demonstrate progress. They should avoid choosing metrics simply because the information is easy to collect. A convenient measure may have little connection to the change objective.

Establish a Baseline Before the Change

A baseline describes the situation before implementation. Without one, it is difficult to judge whether conditions have improved, stayed the same or deteriorated.

Baselines can include numerical and descriptive information. Depending on the initiative, they might cover processing time, error rates, staff workload, customer complaints, sales conversion, absenteeism, safety incidents, system usage or employee confidence.

For instance, before changing an invoice approval process, an organisation could record:

• the average number of days from invoice submission to payment approval;
• the percentage of invoices returned because of missing information;
• the number of manual hand-offs between departments; and
• staff and supplier perceptions of the process.

The baseline should be practical rather than perfect. If historical data is incomplete, the organisation can document the limitations, collect a short period of consistent observations and use interviews or sample reviews to add context. It is better to have a transparent, reasonable baseline than to delay measurement indefinitely while seeking ideal data.

Distinguish Outputs, Outcomes and Impact

One of the most common measurement errors is treating an output as an outcome. Outputs are the immediate products of an initiative. Outcomes are the changes that occur because people use those outputs. Impact refers to broader and longer-term effects.

Consider a Kenyan wholesaler introducing an inventory management application:

• Output: The application is configured and 25 employees receive training.
• Short-term outcome: Stock records are updated more consistently and staff can locate product information more quickly.
• Operational outcome: Fewer stock-outs and emergency purchases occur.
• Business impact: Customer service and margins improve over time.

The initiative may achieve its output but fail to produce the expected outcome. Perhaps the application is unreliable in some locations, staff do not trust the data or managers continue to approve purchases outside the new process. Measuring each level helps reveal where the chain is breaking.

Use a Balanced Set of Measures

A useful change dashboard normally combines several types of measures. Relying on one metric can create a distorted picture.

Adoption measures

Adoption measures show whether the intended users are taking up the new way of working. Examples include the proportion of cases handled through a new process, active use of a system, completion of required steps and the percentage of teams following a revised procedure.

Adoption should be examined for depth as well as breadth. If 90 per cent of employees have logged into a new platform once, that does not necessarily mean the platform is embedded in daily work. Frequency, consistency and correct use may provide a more reliable picture.

Capability measures

Capability measures assess whether people have the knowledge, skills, access and confidence required to sustain the change. These may include assessment results, observed task performance, requests for support, supervisor evaluations and employee feedback.

Training attendance is an activity measure. Demonstrated competence is a capability measure. The second is usually more useful for understanding readiness.

Process measures

Process measures show whether work is becoming faster, safer, more consistent or less wasteful. Examples include turnaround time, rework, defects, queues, hand-offs, compliance with required steps and first-time-right performance.

People measures

Change affects the experience of employees and other stakeholders. Relevant measures may include confidence, perceived clarity, workload, trust in leadership, willingness to use the new process and the quality of feedback received.

People data should not be dismissed as soft or unimportant. Low confidence or high frustration can be an early warning that adoption will weaken, even when operational results still look acceptable.

Business measures

Business measures connect change to wider organisational performance. Depending on the purpose of the initiative, these might include revenue, cost, retention, service quality, risk exposure, cash flow, productivity or customer satisfaction.

Business measures often move more slowly than adoption measures. Leaders should therefore avoid declaring failure too early or success too soon.

Combine Leading and Lagging Indicators

Leading indicators provide early evidence about whether the change is moving in the right direction. Examples include participation in practice sessions, use of a new workflow, completion of coaching, the number of process questions raised and the percentage of managers conducting follow-up conversations.

Lagging indicators show results after the change has had time to influence performance. Examples include reduced complaints, improved delivery reliability, lower error rates or increased retention.

Both are needed. A manager who looks only at lagging indicators may discover problems after considerable time has been lost. A manager who looks only at leading indicators may mistake activity for achievement.

For example, if a service centre introduces a new customer-handling procedure, early measures could track whether staff are using the required diagnostic questions. Later measures could assess repeat contacts, resolution time and customer feedback. The early measures help leaders intervene while the behaviour is still being formed; the later measures test whether the behaviour is producing value.

Measure Quality, Not Just Quantity

Numbers can encourage the wrong behaviour when they are poorly designed. If a team is judged only on the number of cases closed, employees may rush work and create more errors or repeat contacts.

Good measurement considers quality, trade-offs and unintended consequences. A call centre might track resolution time alongside customer satisfaction and repeat calls. A procurement team might track savings alongside supplier quality and delivery reliability. A school or training provider might track enrolment alongside completion and demonstrated learning.

Leaders should also define what a measure means. “System usage” could mean logging in, entering complete records, following the approved workflow or using advanced functions. Precise definitions reduce arguments and make comparisons more reliable.

Gather Quantitative and Qualitative Evidence

Quantitative data helps identify patterns, scale and direction. It can show that processing time has fallen or that use of a new platform differs across branches. However, it may not explain why the result occurred.

Qualitative evidence provides that explanation. Useful methods include short interviews, focus groups, observation, open-ended survey questions, help-desk records and reviews of actual work. These methods can reveal barriers such as poor connectivity, unclear authority, duplicate data entry, inadequate equipment or concerns about how performance is being monitored.

A practical approach is to investigate unusual results. If one Nairobi branch adopts a new process more successfully than other branches, leaders should not simply copy its figures. They should examine what differs: manager behaviour, staffing, customer mix, local support, training quality or access to resources. The answer may identify a practice that can be adapted elsewhere.

Consider Attribution Carefully

Organisational results are influenced by many factors. A rise in sales may reflect a change initiative, seasonal demand, a competitor’s difficulties, pricing changes or wider economic conditions. A fall in complaints may result from improved service, fewer customers or a change in how complaints are recorded.

Leaders should therefore avoid claiming that every positive movement was caused by the change. Instead, they can use several approaches to strengthen judgement:

• compare results with the baseline;
• compare teams or locations progressing at different speeds, where appropriate;
• look for a logical connection between adoption and performance;
• check whether the timing of the result fits the implementation period;
• ask staff and customers what changed in their experience; and
• consider alternative explanations before making a strong claim.

Perfect attribution is rarely possible in complex organisations. The goal is a credible evidence-based judgement, not false precision.

Build a Change Measurement Plan

A simple measurement plan can be organised in a table or shared document. For each intended result, record:

1. Objective: What problem or opportunity is the change addressing?
2. Measure: What will be observed or counted?
3. Definition: Exactly how will the measure be calculated?
4. Baseline: What was the position before implementation?
5. Target or expected direction: What improvement is sought, and by when?
6. Data source: Where will the evidence come from?
7. Owner: Who is responsible for collecting and reviewing it?
8. Review frequency: How often will leaders examine the result?
9. Response: What action will be taken if progress is weak or unintended effects appear?

Targets should be challenging but credible. A target without an owner, time frame or agreed response is unlikely to influence behaviour.

Review Results and Adjust the Change

Measurement only creates value when it leads to decisions. Change leaders should establish review points before implementation begins. A weekly review may suit early adoption issues, while monthly or quarterly reviews may be more suitable for business outcomes.

During each review, ask:

• What is improving?
• Where is adoption uneven?
• What evidence contradicts our assumptions?
• Which barriers can the project team remove?
• What should be stopped, simplified or redesigned?
• Are we measuring an activity instead of a meaningful result?

Measurement should support learning rather than create a culture of blame. If staff fear that every problem will be used against them, they may hide errors and provide unreliable feedback. Clear governance is still necessary, but the purpose of early measurement should be to improve the change while improvement is possible.

Applying This in Practice

Imagine a medium-sized organisation replacing paper-based leave requests with a digital approval workflow. The project team could begin by recording the current approval time, common errors, the number of incomplete requests and staff perceptions of the process.

During implementation, it could track whether employees can submit requests, whether managers approve them through the correct workflow and what support issues arise. After adoption begins, it could examine approval time, incomplete submissions, payroll corrections and employee satisfaction.

If approval time improves but payroll corrections increase, the result is mixed. The organisation should investigate whether the new workflow is missing information, whether managers need clearer guidance or whether payroll integration is unreliable. The appropriate response may be process redesign rather than more general communication.

This example illustrates an important principle: measurement is part of change management, not an administrative task added at the end. It helps leaders test assumptions, protect the intended benefits and make informed adjustments.

Key Takeaways

• Measure the full chain from change activity to adoption, performance and organisational outcome.
• Establish a practical baseline before implementation so progress can be judged fairly.
• Use a balanced set of adoption, capability, process, people and business measures.
• Combine leading indicators with lagging indicators to identify problems early and assess lasting results.
• Pair numerical data with qualitative evidence to understand why results are occurring.
• Test alternative explanations before attributing business results entirely to a change initiative.
• Review evidence regularly and use it to adjust the change, not merely to report on it.