Brinkerhoff's Success Case Method: 5 Steps, With Examples

Go from storyboard to an interactive course in minutes

Start Free

Three months after a new-manager coaching programme, the L&D team has a 4.6 out of 5 satisfaction score, a 94% completion rate and no idea whether anyone coaches differently. The executive who funded it asks a simple question: did it work? Averages cannot answer that, because some managers are using what they learned every week and others never tried it once.

Brinkerhoff's Success Case Method is built for that question. I'm an instructional designer at Coursebox, where we make instructional design software for training teams, and I use the method when a programme needs evidence beyond completion data. I'll use that coaching programme, a hypothetical example, to show each of Brinkerhoff's five steps.

What is the Success Case Method?

The Success Case Method investigates where a programme produced useful results and where use stalled. First define the intended impact, then screen participants, often with a short survey, to find contrasting candidate cases. Interview selected high-success and low-success cases, verify the claims and report the conditions behind them. The method explains how results occurred in those cases; it does not establish a causal effect for the whole population.

Brinkerhoff describes it as "a relatively quick and easy method of finding out what is working and what is not, which also provides accurate and trustworthy information that can be used to make timely decisions" (The Success Case Method, 2003). Stories are its main output, with a strict test: "A story that cannot be confirmed is not a success story."

It answers four questions:

The example column refers to a hypothetical coaching programme for new managers. It teaches structured conversations with team members; the evaluation asks whether managers used that approach at work and what results they can substantiate.

Case study: Appraisal meetings for doctors in Denmark

Researchers screened doctors about appraisal meetings, analysing 77 responses from 115 active doctors. They identified eight high-success and seven low-success candidates, then interviewed six and five respectively. Doctors in the successful cases described more useful meetings and longer discussions, around 60 minutes compared with 10. The contrast helped explain the conditions behind perceived value; it did not show that meeting length caused better clinical performance. For your study, use screening to find contrasts, then investigate what actually happened in each case. Read the Danish appraisal study.

Who created it, and what is current

Robert O. Brinkerhoff, a retired Professor of Counseling Psychology at Western Michigan University, set out the method in a 2002 paper and the 2003 book The Success Case Method: Find Out Quickly What's Working and What's Not. He followed it with Telling Training's Story (2006), which, according to a 2007 University of Cape Town study, uses the terms "high success" and "low success".

The method is still in use. The Brinkerhoff Evaluation Institute runs certification in it and has added a lighter sister method, Essential Impact Evaluation, which it describes as "more streamlined and practical, designed specifically for today's time-constrained L&D professionals" (Brinkerhoff Evaluation Institute).

The 5 steps, worked through one programme

Our hypothetical organisation trained 90 new managers to hold structured coaching conversations with their team members. Three months later, completion and satisfaction are high, but the sponsor does not know who has used the approach or what happened afterwards. The study will investigate contrasting experiences before the next funding decision. That uneven use is why this programme fits the Success Case Method: the team needs to understand both helpful conditions and barriers, not just calculate an average.

1. Focus and plan the study

Decide who the study is for, what decision it should inform and which programme or group it covers.

• What you do: agree the evaluation questions with the sponsor and choose a follow-up point when participants have had a realistic opportunity to apply learning. A few months may suit this coaching programme; another task may need a different interval.
• Coaching programme: the sponsor wants to know whether to run the programme for the next 120 new managers, and what would make it work better. The study starts three months after the last session.

2. Create an impact model

The impact model describes what success should look like if the programme worked: what people learn, what they do with it, and what results that produces for the business.

• What you do: map the expected chain from learning to actions, results and business goals. Treat that chain as a hypothesis: interviews and corroborating evidence should test what actually happened, including other influences.
• Coaching programme: managers learn the coaching conversation framework, hold a coaching conversation with each team member every month, see improved performance on agreed goals, and the business keeps good staff and spends less time on performance problems.

3. Screen participants for high-success and low-success cases

Screen a defined participant group using a short survey, existing records or another suitable source. A census can work for a small programme; a larger study may use an appropriate sample. Record who was invited, who responded and how candidates were selected. The screening identifies people to interview. It does not verify their claims or make a respondent percentage representative of everyone who took the training.

• Tried this and had clear and positive results
• Tried this, but had no clear results yet
• Tried this somewhat, but do not expect any results
• Tried this and it did not work
• Have not tried this at all
• Tried this, but not because of the training

The last option raises an attribution question. Ask what changed, when it changed and what role the programme played. Do not classify the result from that answer alone.

• What you do: keep the survey short, a few closed questions, and send it to everyone when the group is small.
• Coaching programme: two closed questions to all 90 managers on how often they have used the framework and what result it produced, plus two short open questions.

4. Interview and document success and non-success cases

Interview a handful of people at each extreme. The first part of the interview screens out anyone whose result does not hold up. Brinkerhoff's interviews collect five "buckets" of information for success cases, as summarised by University of Melbourne evaluators of a Victorian government programme (Bell and McDonald, 2006):

1. What was used that worked? How, with whom, when?
2. What results were achieved? What is the measurable difference, and what evidence is there?
3. What good did it do? Why was it important?
4. What helped? What contributed to the success?
5. Suggestions: what might have increased the success further?

Non-success interviews focus on the barriers and on what would have helped.

• What you do: ask for evidence of each result, such as who noticed, what records changed or what feedback came back.
• Coaching programme: six success cases, where a coached team member's performance improved and their manager can point to the records, and six non-successes, most of whom never found time for the monthly conversations.

Case study: Student wellbeing support in Victoria

An evaluation of Victoria's student wellbeing initiative combined programme-wide evidence with detailed cases. The initiative funded 256 full-time-equivalent positions across about 450 schools. The case work involved five schools and 36 respondents, helping evaluators understand how schools used the support and where implementation was blocked. The practical value was an explanation of how the programme operated, rather than a claim that the cases proved improved student outcomes across the state. See the Victorian evaluation paper.

5. Communicate the findings

Report the screening results separately from the verified interview cases. State the invitation and response counts, how you chose interviewees, which claims were corroborated and which barriers appeared in the low-success cases. Recommend changes that follow from that evidence.

• What you do: tell the verified stories, report survey categories with their response denominator and sampling limits, and recommend changes to the training and work environment.
• Coaching programme: the report shows the sponsor which managers got results, that most non-successes lacked time and their own manager's backing, and that the next cohort needs a monthly slot protected in their calendars.

How to estimate the value of the training

The survey gives you reported-use categories within its respondent group. Interviews provide verified case evidence about results and the conditions behind them. Neither a selected success story nor an unrepresentative survey establishes the value of the whole programme. If you estimate unrealised value, show the assumptions, the population they apply to and a range of plausible outcomes.

Brinkerhoff gives sample reports with percentages of successful use and scenarios for increasing that use. They illustrate how a report can connect findings to a decision; they are not benchmark success rates or a formula for extrapolating your best interview case. Use your own defensible denominators, verify the value claims and show any projection as an estimate.

The evaluators note that "a traditional summative evaluation would consider this disappointing. However, SCM would deem this highly successful" (Medina et al., 2015). Brinkerhoff is cautious about money figures: "we usually avoid making value and cost-return claims except where such claims are clearly and uncontroversially defensible". When you need a full return calculation, training ROI covers the methods.

Building the follow-up in Coursebox

In a targeted support review between May and October 2026, I found about 13 threads from about 12 customers asking how to add a survey or find responses. Coursebox has a native Survey block and reporting tools. A Success Case follow-up uses those responses to find people to interview about actual use, rather than treating a course rating as evidence of impact.

For this coaching programme, I wanted to invite all 90 managers and collect answers in one place. On 7 October 2026 I built a follow-up in Coursebox, the course builder we make: a short module with a welcome page explaining that answers go to the training team, screening questions, a request for a specific coaching example, another for anyone who had not used it, and a closing page about interviews. Coursebox is our product, so weigh my read with that in mind.

The first outline matched my brief: a welcome page on purpose and privacy, an evidence-of-use section and a closing page, at about 19 minutes. When the builder turned survey answers into a graded quiz, I asked for one ungraded written-response page instead. The final outline had three pages and took about 14 minutes, with all four questions in one assignment. That records narrative responses without a pass mark. For structured survey responses, the preferred next step is to generate the course and add the native Survey block in the page editor; it is added after generation rather than requested as an outline block.

After generating the course, open the intended page in the editor, add a Survey block, enter the screening questions and preview it as a learner. Use the responses to select interviews, then verify claims outside the survey. Coursebox explains which blocks are added after generation in its course-creation guide.

Success Case Method vs the Kirkpatrick model

Kirkpatrick organises evaluation around reaction, learning, behaviour and results. The Success Case Method investigates contrasting cases of use to explain what helped and what blocked results. They can work together: use your evaluation measures to find candidate cases, then interview and corroborate them. Neither framework, on its own, turns a positive result into proof that training caused it.

For programme-level decisions, the CIPP model looks at context, inputs and process as well as outcomes.

Limitations of the Success Case Method

• Biased by design. Brinkerhoff's own answer to "Isn't the SC Method biased because it looks only at a few cases?" is "Yes. It is biased." It does not try to judge a whole programme from a few cases.
• Self-reports overstate success. In one field trial, "the team interviewed three participants who had been identified as "success cases" through the survey, the interviews showed that only one of them qualified as such", and respondents tended "to confuse actions for outcomes" (BetterEvaluation).
• Non-successes are hard to find. A 2024 study of a global faculty development programme named this as a major limitation (BMC Medical Education, 2024), and the field trial above found no non-success cases at all.
• Ceiling effects. In a US hospital's quality improvement training, 51 of 142 respondents gave themselves the top score, so the evaluators narrowed who they interviewed (BMC Medical Education, 2017).

Use the cases to make the next programme decision

Define the result you need to investigate, find contrasting candidates and check their stories against records or other witnesses. Keep the survey counts separate from the cases you verify. Then turn the findings into a specific decision: change a tool, remove a barrier, strengthen manager support or revise the training. The report is most useful when it explains what you will change and which evidence led you there.