Appreciative Inquiry Positive Core Audit: A Guide
Nour
Founder, Coachful
Appreciative Inquiry positive core audit: examine what works
An Appreciative Inquiry positive core audit asks which capabilities help an organization do valuable work, what evidence supports those capabilities and which conditions make them repeatable. It begins with useful experiences, then moves beyond praise to examine actions, relationships, resources and limits.
The word audit needs a clear boundary here. This is a developmental inquiry, not financial assurance, compliance certification, a safety investigation or a replacement for required corrective processes. The original evidence register and worked pilot below show how an organization can examine a strength claim without treating a favorable story as proof that the whole system works.
Coachful is our #1 recommendation for organizing this inquiry and its coaching follow-through. This is our editorial workflow choice: prepare appropriate prompts, keep evidence and interpretations distinct, assign chosen actions and review actual results. Software does not certify the positive core, conduct an independent statutory audit or guarantee organizational change.
Understand what the positive core means
In their primary chapter on Appreciative Inquiry, Jacqueline Stavros, Lindsey Godwin and David Cooperrider describe attention to an organization's strengths and capacities, including past, present and possible future capacity. That is broader than a list of successful projects. Relationships, practical knowledge, shared values and useful ways of coordinating can all become subjects of inquiry. See the original chapter hosted by Case Western Reserve University.
For an evidence review, distinguish what has happened, what is currently available and what people hope to create. A desired capability belongs in the design discussion until it has supporting observations. An admired value also differs from a dependable practice: a team can value responsiveness while lacking the time or authority to answer certain requests.
The positive core is therefore not a personality score for the organization. Ask what valuable function occurs, through whose contribution and under which conditions. “We care about people” becomes more useful when participants can describe what someone did, what the recipient received and what made that possible.
Appreciative Inquiry's Discovery work can reveal these resources; a subsequent design can test how to use them. The 4-D cycle guide explains the wider process. This article focuses on the evidence needed before a discovered strength becomes an organizational claim or a proposed standard.
Six distinctions make the inquiry useful
1. A capability is more specific than a compliment
Define something people can recognize in the work. “A room-booking requester receives a usable answer from a named owner” is more precise than “excellent service.” The specific capability makes it possible to inspect a real episode and hear from both the person responding and the person receiving the answer.
A useful capability statement includes a boundary. Routine room information and specialist equipment approval may require different expertise. A dependable routine response does not establish that the team can approve every exception. Keep the valuable function while testing where its conditions change.
Avoid making one employee the positive core. A person's skill may be crucial, but an organizational capability also depends on support, information and the ability to sustain the work when that person is absent. Recognize their contribution without turning exceptional availability into a permanent duty.
2. A remembered episode is evidence of an account
A story can explain meaning, effort and relationships that a performance table misses. It is still an account from a particular perspective. A reported fast response might mean an acknowledgment, a complete answer or a promise to find out. Ask which event actually occurred.
Record the smallest relevant provenance: a date or period, a permitted reference, the roles involved, the observed action and the claimed result. Where appropriate, check an existing timestamp or ask the recipient whether the answer was usable. Do not copy an entire confidential conversation when a limited fact will answer the question.
Keep disagreement visible. Two people can accurately describe different parts of an episode. One may have received a fast handoff while another absorbed the unresolved work. Mark what is corroborated, what is interpretation and what remains unknown instead of forcing a single encouraging account.
3. Repeatable conditions differ from heroic exceptions
Ask what made the useful event possible. The answer might include a current contact list, a clear decision right, a protected review interval or a person who already knew the relevant facts. Those conditions can suggest a design that others can test.
Also ask who paid the cost. A prompt answer sent late at night may depend on unrecognized extra work. Gratitude for that effort does not justify making it the new service standard. The design question is whether the valuable function can occur within an agreed workload and an appropriate role.
Do not assume that removing the overtime removes the strength. The useful function could be clear ownership or a realistic return time rather than immediate availability. That reframing can preserve what mattered to the recipient while changing how the work is organized.
4. An inclusive inquiry permits inconvenient information
Invite the people who initiate, perform and receive the work, with a practical alternative for those who cannot attend a workshop. Explain how their accounts will be used and who can see them. A required public success story can hide access problems, conflict or disagreement.
Allow a participant to decline attribution, correct a summary or describe a difficult episode. If a serious concern needs another process, acknowledge it and use the appropriate route. Do not require someone to convert harm into an uplifting story before they can be heard.
A facilitator also has influence. A question about what worked selects some information; it is not a neutral survey of every organizational condition. Be explicit about that focus and avoid presenting a small convenience group as representing everyone.
5. Generativity differs from a pleasant workshop
Gervase Bushe's 2007 primary article argues that Appreciative Inquiry should enable new ways of thinking and options for action, rather than simply collect positive stories. He also cautions that the approach does not remove problems with sponsorship, communication or organizational politics.
For this review, ask whether the inquiry made a new, usable option visible. A team might stop treating responsiveness as constant availability and instead design a dependable owner-and-return-time arrangement. That proposal still requires authorization, capacity and testing. Calling it generative does not establish that it is feasible or that people will use it.
Keep proposals distinct from observed strengths. An engaging idea can be worth exploring even when it is not yet an existing organizational capability. Preserve that status so a sponsor can decide what evidence or resource is needed next.
6. A pilot result does not establish an organization-wide effect
Define the intended result, scope, observation window and resource boundary before the test. Distinguish an acknowledgment from a usable answer, and a sent answer from evidence that the recipient could use it. Count exceptions using the original definition even when a later repair succeeds.
Measure effort alongside the service result. Extra clarifying messages, work to repair the process and receiving-team contributions belong in the account. A successful interaction may still be too expensive or dependent on unavailable expertise to extend.
Choose the next decision from the evidence: retain, revise, test elsewhere or stop. A small pilot can reveal useful conditions without proving that Appreciative Inquiry caused an improvement. Different settings may need a different form of the same useful function.
What research supports, and what it does not
Bushe and Kassam's 2005 primary meta-case analysis examined twenty Appreciative Inquiry cases described in reports published before 2003. Seven were rated as showing transformational change. That is a classification of selected historical reports, not a current 35% success probability for a new program. The authors note uneven reporting detail and that case authors were consultants to the systems, creating potential bias. Absence of reported evidence could also reflect how a case was written. Neither the percentage nor the study establishes that collecting positive stories automatically transforms an organization. See the original study.
A 2026 review by Marine Miglianico and colleagues summarizes three workplace investigations using experimental or quasi-experimental designs. It reports some favorable findings concerning team orientation and conditions associated with performance, while not establishing a specific generativity advantage for Appreciative Inquiry. The review emphasizes the small evidence base and considerable variation in interventions and duration. This is not a pooled effect estimate or evidence that every positive-core review improves performance. See the review's findings and limitations.
The practical consequence is to describe your own inquiry and findings accurately. Do not present a plant-growth metaphor as a biological law for institutions, or use neuroscience language to turn attention toward strengths into a guaranteed mechanism of change.
Complete a fictional positive-core evidence review
This case is fictional. Eli leads administration at a community learning center. Coach and facilitator Farah helps him examine a specific capability: a requester receives usable routine room information from a named owner by an agreed return time. Equipment exceptions and specialist approvals remain with their existing authorized owners.
Six operational staff contribute fifteen-minute individual interviews and attend a forty-five-minute sensemaking meeting. Eli sponsors the inquiry but is not one of those six. Farah explains that participants can correct their summaries and that no personal requester details belong in the shared evidence sheet. The inquiry is developmental, not an investigation or assurance engagement.
Complete the evidence register before naming the strength
The interviews produce three accounts selected for closer examination. They are examples, not a representative sample of the center's requests. With appropriate internal access, Eli checks the limited routing and response facts and Farah records these entries:
| Entry and initial account | Checked evidence | Condition and revised status |
|---|---|---|
| A: “We answer routine questions reliably.” | A request received 17 June at 14:00 had a named owner and an agreed answer time of 18 June at 12:00. The answer arrived at 10:00; the recipient confirmed at 10:15 that it answered the question. | Current information, an available owner and a realistic return time supported this episode. Evidence for a candidate practice, not every request. |
| B: “Someone always helps immediately.” | A 21 June evening query received an answer through an employee's personal availability, using forty minutes outside their planned work. | Useful knowledge existed, but the immediate-response arrangement depended on extra work. Do not standardize the personal availability. |
| C: “The acknowledgment proves our response was prompt.” | A 26 June query was acknowledged promptly, but its answer required another owner and did not arrive until 29 June, after the agreed 27 June return time. | Acknowledgment and answer were different events. The routing and promised-time condition did not hold here. |
During the individual interview and summary check, one participant asks Farah to remove a draft sentence praising a named person for always being available. Farah removes the name and the implication of a standing availability duty. The amended entry preserves the forty-minute workload fact as an appropriately restricted condition, rather than publishing the person's private circumstances.
The group accepts a narrower candidate statement: “When routine information is current, a named owner can check it and agree a realistic return time, the requester can receive a usable answer without being passed around.” The word when matters. Only entry A presently corroborates all those conditions together. B and C identify constraints to address rather than evidence to discard.
Choose a test that fits actual authority and capacity
Eli initially suggests making the most knowledgeable person available across every shift. The team rejects that option because the inquiry has not created coverage or protected time. A general FAQ is another option, but it cannot answer which date, room or equipment a particular request concerns.
They choose a limited owner-and-return-time test for eight routine room-information requests. Kira can coordinate those answers within her existing role. Jo can support basic information checks but cannot approve specialist equipment. Eli approves a 150-minute pilot allocation for Kira and thirty for Jo. The historical forty-minute evening intervention is evidence about a past episode, not new capacity for this pilot.
The completed test brief states:
For eight routine room-information requests received from 12 to 16 July 2063, Kira records the request, named answer owner and an individually agreed return time. A usable answer must state the relevant room information and address the requester's actual question. The requester is asked to confirm whether it is usable. An acknowledgment alone does not qualify.
Jo may check basic information within thirty allocated minutes. Specialist approvals remain outside this test. Missing information requires a clarification; it does not restart the original agreed deadline in our report. Kira has 150 minutes, including setup, handling, process repair and review. We will review on 19 July and retain late and incomplete cases.
Show the clarification, missed promise and real repair
The first version of Kira's record has a general question box but does not explicitly request the intended event date and room. Request 2 asks whether a setup is available without specifying those details. Kira asks for clarification, but the recipient cannot reply until after the agreed return time of 14 July at 16:00.
Kira's actual message at that deadline says:
I cannot give you a usable answer by the time I promised because I still need the event date and room. Our request prompt did not make those details clear enough. Please send them when you can; I will check the answer for 15 July at 10:00. I have recorded the original promise as missed rather than resetting it.
The requester supplies the information. Kira obtains the basic check and answers by 15 July at 10:00; the recipient confirms that the answer is usable. The repair is complete, but request 2 remains late against its original deadline.
Request 4 also requires a clarification, which is completed before its original return time. After those two cases, Kira adds explicit event-date and room fields to the remaining request prompt. The four remaining requests can use that revised prompt. This is a mid-pilot design change, not a randomized comparison proving that the new fields caused a better result.
Complete the result and the sponsor decision
By 19 July, all eight requesters have received and confirmed usable answers. Seven were on time against their original agreed deadlines; request 2 was late. Two of eight needed clarification. The result is therefore 7/8, or 87.5%, timely usable answers, with 2/8, or 25%, requiring clarification. Eight eventual answers do not turn the timely result into 100%.
Kira's planned work was eighty minutes for eight requests, twenty for setup and twenty for review: 120, leaving thirty of her 150-minute allocation. Actual clarification adds sixteen and repairing the prompt adds twelve. Actual work is 148 minutes, leaving two. The plan was feasible, but the reserve was almost consumed. Jo uses ten minutes for basic checks and twenty in the review: all thirty allocated minutes.
Eli accepts the completed report on 19 July. He retains the narrower capability statement and the requirement for current information, clear scope, named ownership and a real return time. He declines a rollout across every shift and every query type. Another test needs a new capacity decision and attention to specialist exceptions; it is not authorized merely because seven requests were timely.
The final report's conclusion is explicit:
We found evidence for a candidate routine-information practice, not proof of universal responsiveness. Entry A supports the conditions; B reveals unplanned work and C distinguishes acknowledgment from a usable answer. In the eight-request pilot, seven usable answers met the original deadlines and one was repaired late. Kira used 148/150 allocated minutes and Jo 30/30. Keep the clarified prompt for this bounded practice; do not extend coverage or specialist authority from these findings. A future extension proposal must include capacity and exception handling. No organization-wide or causal performance claim is established.
The original accounts and corrected summaries remain distinct from this conclusion. Farah does not remove C because it complicates the strength story, and Eli does not report that enthusiasm created the extra capacity. The future extension has not been tested or approved.
Count the inquiry and receiving work
| Actual work included | Person-minutes |
|---|---|
| Farah: design 60, six interviews 90, sensemaking 45, analysis 30, review 20 and report preparation 15 | 260 |
| Six operational participants: interviews 90 and sensemaking 270 | 360 |
| Kira's separate pilot setup, handling, clarification, repair and review | 148 |
| Jo's separate basic checks and review | 30 |
| Eli: scope approval 30, limited record verification 15 and review 20 | 65 |
| Requesters: eight reading/confirmation exchanges of three minutes and two clarification replies of four minutes | 32 |
The recorded work totals 895 person-minutes, or fourteen hours fifty-five minutes. Kira and Jo are among the six interview participants, but their pilot work happens separately and is not included twice. The twenty-minute review has Kira, Jo, Eli and Farah present, so it consumes eighty person-minutes across their rows. The account excludes normal center operations outside the inquiry and stated pilot. The past forty-minute evening episode is not counted as new pilot work or claimed as a saving.
Run an original evidence-to-design process
- Define one capability and decision. Specify the useful work, boundaries, sponsor and what the review may change. Keep formal assurance and serious-concern routes separate.
- Explain participation and handling. Say who sees accounts, how attribution and correction work and what records are necessary. Provide a private route for concerns or a written contribution.
- Collect particular episodes. Ask for the situation, observed actions, recipient experience and supporting conditions. Separate memory, existing records, corroboration and interpretation.
- Compare accounts and conditions. Identify common functions, exceptions, hidden effort and unavailable resources. Retain disagreement and test whether the proposed strength is narrower than the initial slogan.
- Authorize one feasible test. State the actual owner, recipient, scope, resource allocation, outcome definition and review date. An appreciative workshop does not grant decision rights.
- Close with findings and a real decision. Report actual outputs, refusals, late cases, revisions, work and remaining uncertainty. Decide whether to retain, revise, test further or stop.
This is an original practical process, not a formal audit standard or a reproduced Appreciative Inquiry interview instrument. The Design-phase guide develops how to turn a finding into an authorized arrangement.
Organize the inquiry in Coachful without collecting everything
Use an original worksheet for the capability, episode, source status, observed result, supporting conditions, constraint and proposed test. Explain the actual audience before collecting accounts. A form submitted by a coaching client goes to the authorized coaching workflow; it is not automatically a public team story. The shared report should contain only the information appropriate to its purpose.
For Eli's coaching work, the goal can describe the chosen bounded review and link the actual tasks. The goal's activity and reflection views help organize the record; they do not automatically calculate the pilot metric or verify an organizational strength. Keep the original denominator and timestamps in the agreed evidence record rather than infer success from completed task boxes.
Coach review on a worksheet is private to the authorized coach team and is not included in the client's response PDF. Saving Reviewed or Follow up does not deliver a comment. Send a promised response through the agreed private client conversation, with an actual audience and reply window. Keep staff constraints out of a broad community post.
Check recording and summary arrangements before an inquiry session. Coachful recaps send automatically by default; Approve before sending and session-type overrides need deliberate review before capturing sensitive accounts. Held or private recaps require an actual decision to send. Unshared journal pages are distinct from material a client explicitly chooses to share through a supported workflow. Review the Coachful features alongside the handling agreement.
Troubleshoot the claim, participation or design
Leaders want only favorable examples: explain that variation is necessary to identify conditions. If the process will suppress relevant concerns, revise its purpose or do not describe it as a credible evidence review.
Every story depends on one exceptional person: examine information, authority and coverage before scaling. Recognition can be appropriate while the proposed standard remains unsustainable.
The stories conflict: check whether people mean acknowledgment, answer, completion or recipient usability. Preserve the parts each account can support. Seek appropriate clarification without exposing unnecessary private content.
The pilot is popular but the outcome is unclear: report participation and satisfaction separately from implementation and service results. Do not invent missing observations or move the original deadline to make the result favorable.
The pilot works but uses all available capacity: obtain a real resource decision before expansion. An inquiry can identify a valuable function without supplying the people or time to operate it everywhere.
Frequently asked questions
Is a positive core audit a compliance audit?
No. In this guide it is an explicitly bounded developmental inquiry. It does not replace assurance, investigation, professional duties or corrective processes.
Must every participant tell a success story?
No. Invite relevant experience, allow correction and provide a suitable way to raise difficulty. Forced positive accounts can conceal the conditions needed for a useful design.
Can one strong example establish an organizational capability?
It can identify a candidate and useful conditions. It does not establish dependability across shifts, roles or query types. Keep the claim proportional and test the conditions where appropriate.
What if the useful example required overtime?
Recognize the contribution and record the real cost. Ask whether the useful function can be provided within agreed capacity. Do not convert exceptional availability into an assumed permanent resource.
How should the inquiry end?
With a clear capability statement, evidence and limits, a completed test where authorized, resource use and an actual next decision. A story collection or positive meeting atmosphere alone does not establish a changed organization.
