Define the audit question and window
A reputation audit identifies gaps between customer promises and observed experiences. An audit should answer a question that management can act on. “How is our reputation?” is too broad to investigate cleanly. “Which delivery and refund failures generated unresolved public complaints during the previous quarter?” gives the audit a population, a time window, and a decision. Write that question before exporting any data.
Choose the event date you will use. A complaint posted this quarter may concern a purchase made last quarter. Keep both dates when available, and avoid moving a case into whichever period makes the results look better. Freeze the initial audit dataset and note subsequent corrections in a revision log.
The FTC Consumer Review Fairness Act guidance [2] addresses restrictions on honest assessments. During an online reputation management audit, keep policy questions separate from service findings so a reviewer can assess the applicable context without treating criticism itself as a process defect.
Inventory the evidence sources
List owned social accounts, public review profiles, support case summaries, search results for relevant brand queries, and customer survey material you are authorized to use. For each source, record the covered period, access method, exclusions, and collection date. A missing channel is an audit limitation, not proof that nothing happened there.
Keep public evidence and internal case information connected through a case identifier rather than copying sensitive records into every worksheet. Someone reviewing the audit needs to understand the finding and inspect the permissible evidence. They do not need unrestricted access to every customer's address or payment history. Vendor listening coverage documentation illustrates why source availability must be checked rather than assumed.
Sample response quality deliberately
Review more than the most visible negative posts. Include routine resolved cases, unresolved cases, high-severity cases, and a small sample of apparently positive interactions. This makes it possible to find problems in the process rather than simply confirm that complaints exist. Describe how the sample was selected and avoid presenting a purposive sample as a statistically representative survey.
Score the response against observable criteria: it addresses the actual issue; it separates known facts from assumptions; it protects personal details; it names a next step; and promised follow-up is documented. Two reviewers should independently score a shared subset. Discuss disagreements and revise ambiguous criteria before reviewing the rest.
Connect the findings to brand promises
Build an audit table with four columns: promise, observed experience, evidence, and corrective owner. A complaint that customer support was “slow” becomes more actionable when the published support commitment and actual case timeline are available. If the promise is vague, the finding may concern misleading expectations rather than a missed formal service level.
Distinguish communication failures from product or service failures. An inaccurate public update may be corrected by the communications team. A refund process that requires repeated contacts needs an operational owner. One case can contain both. Avoid assigning all reputation work to marketing just because the evidence appeared on a social platform.
Worked example: an audit of refund friction
All figures in this example are hypothetical. A subscription company reviews 120 public messages and a permitted summary of 50 support cases. The analyst identifies 18 public refund complaints, but eight are duplicate messages from the same customers. The relevant unit is therefore ten unique public complainants, with message volume retained separately.
Of the ten customer cases, six have verified payment timelines. Four show that an approved refund was processed correctly but the expected timing was not explained. Two show an internal approval delay. The remaining four cases lack enough evidence to determine a cause. The analyst reports those four as unresolved audit questions rather than assigning them to the largest category.
The resulting action plan has two owners: support updates refund timing language, while finance investigates the approval queue. The company schedules a follow-up audit on new cases using the same definitions. That follow-up can assess whether the identified failure recurs, although a smaller sample or changed collection scope would limit the comparison.
McKinsey's customer-feedback guidance [3] emphasizes collecting, analyzing, and acting on feedback. Apply that principle here by separating the observed complaint, verified operational finding, and assigned repair. The audit then produces work that a responsible team can inspect and complete.
Prioritize with severity, recurrence, and fixability
A sensible action list considers potential harm, repeat occurrence, evidence strength, and the feasibility of corrective work. Do not rank issues only by message count. A small set of credible account-security complaints can require specialist attention, while many comments about a temporary stock shortage may call for a straightforward information update.
For each priority, write the proposed action, owner, due date, verification method, and residual uncertainty. “Improve customer trust” is an aspiration. “Remove conflicting refund instructions from three customer touchpoints and verify them against the current policy” is an auditable action. Keep the broader aspiration, but assign work using concrete changes.
Make the audit repeatable and honest
Save the search terms, export filters, exclusion decisions, coding rules, and review rubric. A future analyst should be able to reproduce the reasoning even if the underlying platform changes. Retain source links where permitted, and record when an item is no longer available instead of treating an inaccessible post as never having existed.
The audit cannot establish the views of all customers from a social sample. It also cannot prove that every allegation is true or false. Its value is a defensible account of observed evidence and the decisions it supports. If review practices are part of the audit, consult current FTC guidance and qualified counsel rather than inventing methods to suppress unfavorable but honest customer feedback.
The audit should examine policy and source scope separately. Compare the collection inventory with listening availability documentation [1], then route review-policy questions to an appropriate assessment of FTC consumer review guidance [2]. A coverage gap needs a research repair; a questionable suppression practice needs policy review. Combining them into one reputation score would conceal the different owners and remedies. Record evidence for each finding before prioritizing the audit actions.
Questions teams ask
How often should a reputation audit run?
Choose a cadence based on product change and operational risk. A quarterly audit with targeted checks after significant changes is a practical starting proposal. High-risk events may require immediate specialist review rather than waiting for the calendar.
Can we audit without paid listening software?
Yes, within a clearly documented scope. Owned-channel records, permitted support summaries, and manual public review can identify useful failures. State the source and scale limits rather than claiming complete coverage.
Should the audit produce one reputation grade?
A single grade can hide the cause of a problem. Preserve separate findings for promises, response quality, unresolved cases, and coverage. An executive summary can prioritize them without collapsing the evidence.
References and source scope
- Sprout Social listening coverage documentation
Sprout-specific source objects, history, updates, and limitations; not evidence of Penguin ePulse coverage or universal availability.
Source link checked 2026-10-03 - FTC consumer review guidance
FTC explanation of restrictions on honest assessments; actual policies still require qualified contextual review.
Source link checked 2026-10-03 - McKinsey: Are you really listening to what your customers are saying?
Collecting, analyzing, and acting on customer feedback across a journey; historical operating principles, not current financial benchmarks or a representative social-media survey.
Source link checked 2026-10-03