Social Media Content Audit: A Practical Step-by-Step Guide

An audit answers six practical questions. What was published? Why did it exist? Is it still accurate? Do we have the rights and approvals needed to keep using it? Can the intended audience understand and access it? What should happen next? Those questions are simple, but teams often split the evidence across platform dashboards, shared drives, campaign briefs, email threads, and individual memory.

Start by treating every published item as a record. The record should identify the account, platform, publication date, format, campaign, owner, public URL, source media, destination link, approval or rights evidence, accessibility elements, available performance fields, and action decision. A missing field is a finding, not a reason to invent a value.

A useful audit reviews more than performance. A popular post can still contain an expired offer, broken link, outdated disclosure, inaccessible image, unclear media license, or former employee contact. A low-performing post may contain a strong evergreen explanation that deserves a better format. Editorial judgment sits beside metrics because metrics do not answer every risk or usefulness question.

Editorial recommendation: define the smallest scope that can produce decisions. A first audit of the last ninety days on one account is usually more useful than an ambitious all-time review that never reaches an action plan. The correct period depends on publication volume, campaign cycles, staffing, record obligations, and risk.

Write a one-paragraph audit brief before opening a dashboard. Name the accounts, date range, campaigns, formats, purpose, final decision owner, and deadline. Also state what is outside scope. If the audit covers organic posts only, say that paid creative, comments, direct messages, and influencer usage outside the owned account require separate review.

Choose the unit of review. For a simple feed, one row per published post works. A campaign audit may need one parent row for the campaign and child rows for every platform version. A video series may need separate fields for the original file, captions, thumbnail, description, music or audio rights, and every published URL. Do not compress materially different versions into one row just to make the inventory shorter.

Define the action rules before seeing results. Keep means the content remains accurate, authorized, accessible for its purpose, and strategically useful. Refresh means a fix can preserve the core item. Repurpose means the underlying idea or approved asset can become a new format under current review. Retire means the team should stop reusing the item and, when appropriate and permitted, archive, unpin, unlist, correct, or remove it on the platform.

Assumption example: a team may decide that an expired offer automatically moves to Refresh or Retire regardless of engagement. That is a team rule, not a platform standard. Records, contractual, or regulatory obligations may limit deletion. Document who can authorize removal and when specialist review is needed.

Begin with platform exports and account records, then reconcile them with the public profile and your scheduling system. First-party exports can reduce copying mistakes, but they may not contain every field you need. LinkedIn currently lets Page administrators export content analytics for a chosen timeframe as an XLS file. Meta says people with full control can request a Facebook Page download and select information categories, format, media quality, and date range. These features and menus can change, so confirm them before each audit.

Save the raw export separately from the working inventory. Record the platform, account, requested period, request date, completion date, export format, timezone if known, and person who obtained it. Do not edit the raw file. A working copy can normalize dates, split combined fields, add campaign labels, and connect each record to its public URL and source asset.

Next, compare the export with the scheduler calendar and live account. Look for posts that were scheduled but failed, published through the native platform, deleted after publication, duplicated, or moved between campaigns. If you cannot confirm a live URL, mark the status Unknown and assign a verification owner. Do not treat an absent row as proof that a post never existed.

Use stable identifiers when available. A platform post identifier, scheduler record identifier, and source asset identifier are more dependable than a caption opening. Captions can be edited or repeated. The inventory can display a readable title, but the identifiers help the team reconcile versions without guessing.

Every metric needs a name, definition, source, date range, and extraction time. A field labeled engagement can mean different things across platforms, exports, and internal dashboards. Store the original field name when possible. If you calculate a rate, show the formula and inputs rather than presenting the result as a universal benchmark.

Sourced fact: LinkedIn's current Page admin documentation separates super admin, content admin, and analyst roles, and lists analytics export access for all three. That is useful when assigning an audit contributor because someone may be able to export analytics without receiving the broader power to manage administrators or edit the Page. Confirm the current role table for the account before changing access.

Use NIST's least-privilege principle as a security reference: access should be limited to what a person or process needs for assigned work. An auditor who needs a content export does not automatically need publishing credentials, account recovery access, or a shared password. If the platform only offers a broad role, document the limitation, use the narrowest available access, set an end date, and review removal after the audit.

Separate fact from interpretation in the working file. Fact: the export reports a value for a named field. Calculation: the team divided one documented field by another. Assumption: the selected date range represents the current content program. Editorial opinion: a post should be repurposed because the idea remains useful. This separation makes later review possible even when platform definitions change.

Read each item in context. Ask what audience problem it addressed, what action it requested, and whether those choices still match the current strategy. A launch announcement may have completed its job but still contain an evergreen demonstration. A repeated tip may be accurate but no longer distinct. A high-traffic pinned post may deserve the strictest review because more people rely on it.

Check factual elements one by one: names, dates, prices, offer terms, product availability, locations, links, contact details, statistics, quoted claims, and instructions. Attach the current source for material claims. If a source has changed, do not assume the old statement remains true. Mark the finding and route it to someone with authority to verify it.

Then examine audience value. Does the content answer a real question, demonstrate a process, tell a useful story, or help a person decide? Is the opening understandable without campaign context? Does the media carry essential information that the caption never explains? Is the call to action still possible? These questions are editorial judgments, so write a short reason rather than hiding the decision inside a color score.

Avoid a single composite score unless every component and weight is documented. A total score can make a rights failure look harmless because other fields scored well. Use hard gates for expired offers, unsupported claims, missing permissions, sensitive data, or inaccessible essential information. Use strategic scores only after the hard gates pass.

For every reusable photo, video, illustration, audio track, quote, testimonial, or creator contribution, record who owns it, what permission covers the use, where it may appear, how long the permission lasts, whether editing is allowed, and where the evidence is stored. A file in a shared drive is not proof of permission. The U.S. Copyright Office explains that finding a work online does not resolve its copyright status and that, unless a limitation applies, the user generally needs permission or a license from the copyright owner. Ask qualified counsel about uncertain rights or limitations.

Review disclosures separately. FTC staff guidance says a material connection to a brand should be obvious, the disclosure should be hard to miss and placed with the endorsement, and clear language should be used. For an audit, record whether the original disclosure remains visible in the caption and media, whether later edits preserved it, and whether a reusable asset would need fresh review for a new placement or audience. Other jurisdictions may impose different requirements.

Test every destination link in an ordinary browser session. Record the final destination, status, redirect, offer validity, and whether the destination matches the promise in the post. A working link can still be wrong if it now reaches a generic home page, expired event, different product, or region that excludes the intended audience. Never publish credentials or private tracking data in the audit sheet.

Review accessibility in context. W3C's image decision tree advises a brief text alternative when a simple image contributes meaning, an empty alternative for purely decorative images, and an equivalent presentation elsewhere when a graph or complex image carries information. Social platforms implement accessibility features differently. Confirm the current platform controls, review caption files and burned-in text, and ask an accessibility professional when the audience or consequence requires specialist evaluation.

Lena Park is an illustrative food creator reviewing six months of recipe and partnership posts before renewing sponsor packages. She limits the first pass to published feed posts and short videos on two accounts. Stories, comments, and raw direct messages are outside scope because they require a separate privacy and retention review.

Lena exports the available account information, reconciles it with her calendar, and links each post to the original recipe notes, media folder, sponsor brief, approval, and disclosure evidence. She does not rank sponsors or invent an engagement benchmark. She records the platform's available fields with the extraction date and uses them only to compare items measured under the same documented definition.

One evergreen knife-skills video still answers a common beginner question, but its caption sends viewers to an expired workshop. Lena marks it Refresh, replaces the destination only after reviewing the revised call to action, and confirms that the media permission covers continued use. A gifted cookware post has no easily located approval record, so it moves to Hold for evidence instead of being scheduled again.

Assumption: Lena controls the original recipe text and footage. Limitation: contracts, talent appearances, licensed music, sponsor terms, and local advertising rules may alter what she can reuse. The example shows an operating process, not a conclusion about legal ownership.

Harbor & Pine Studio is an illustrative small agency auditing three client calendars after a staffing change. It creates separate inventories for a cafe group, a home-services company, and an arts nonprofit. The agency does not merge client records into one unrestricted sheet because rights, approvals, contacts, and account access belong to different organizations.

The account lead defines a shared field dictionary, while each client names its final decision owner. Analysts receive the narrowest available analytics access. Publishers are not asked to share passwords with auditors. The team records temporary access and schedules an access review after delivery. These are editorial operating recommendations informed by least privilege, not a claim that every platform offers an ideal audit-only role.

The cafe inventory reveals duplicated seasonal posts for locations with different hours. The home-services inventory contains an old phone number in image text. The nonprofit has strong event photos but uncertain reuse evidence for two volunteer portraits. The findings produce three different actions: location-specific Refresh, urgent factual correction, and Hold pending permission evidence.

Harbor & Pine gives each client a decision log containing the affected URL, finding, source, proposed action, approver, due date, and verification result. The audit is complete only when the client accepts the action list and ownership is clear. A colorful presentation without owners would not resolve the operational risk.

Trailcrest Cycles is an illustrative regional bicycle brand reviewing last year's spring commuting campaign before building the new calendar. The campaign included product photos, repair tips, route-safety reminders, retailer links, and a giveaway. The team audits the campaign as a group but keeps one record for every platform version.

The giveaway posts move to Retire from reuse because the dates and terms have ended. A repair checklist remains accurate after review and moves to Repurpose as a captioned short video. A route graphic contains essential information but lacks an equivalent text explanation, so it moves to Refresh with an accessibility owner. A product post links to a discontinued color, so the merchandising owner must approve any replacement destination.

The brand does not claim that last year's engagement predicts this year's result. It uses historical fields as context and checks current inventory, pricing, safety language, retailer relationships, and media rights before a new approval. The prior post is evidence of what happened, not permission to repeat every claim or asset.

Editorial opinion: campaign-level review is most useful when a team expects to reuse a theme. It reveals which components remain valuable while preventing an expired detail from traveling into the new calendar. The campaign brief should record the new audience, objective, facts, and constraints rather than treating the old creative as a complete brief.

Estimate the work from count, sampling rule, and review depth. Here is a transparent example, not an industry benchmark. Assume an account published 180 items in the chosen period. The team performs a two-minute inventory check on all 180 items, a six-minute evidence review on 60 high-priority items, and a fifteen-minute specialist review on 12 flagged items.

Calculation: 180 multiplied by 2 minutes equals 360 minutes. Sixty multiplied by 6 minutes equals another 360 minutes. Twelve multiplied by 15 minutes equals 180 minutes. The review total is 900 minutes, or 15 working hours. Add three hours for setup, exports, reconciliation, and final decisions. The planning estimate becomes 18 hours before meetings or remediation work.

Assumptions: the records are accessible, one person can perform the first two passes, specialist time averages fifteen minutes per flagged item, and remediation is scheduled separately. If exports are incomplete, rights records are scattered, accounts use different definitions, or many posts require live verification, the estimate increases. Track actual time so the next audit can use local evidence instead of a generic guess.

If the inventory is too large, choose a documented sample rather than claiming a full audit. You might review every pinned post, every active offer, every sponsored item, every top landing-page referrer, and a random selection from the remaining period. State what was sampled, how it was selected, and what the sample cannot prove. High-risk categories should not be excluded simply because they performed poorly.

Keep an item when its facts, link, permission, disclosure, accessibility treatment, and strategic purpose remain acceptable for its current use. Keeping does not mean reposting immediately. It means the team can leave the existing item in place and may consider reuse after a new pre-publication review.

Refresh when a bounded correction preserves the core value. Examples include replacing an expired destination, correcting image text, adding an accessible explanation, updating an approved product detail, or revising a caption. Platform editing limits vary, and changing a material element can require a new version, fresh approval, or a new post. Confirm current platform behavior before promising an in-place edit.

Repurpose when the idea remains useful but the format, framing, audience, or channel should change. Return to the source evidence instead of copying the old post blindly. Confirm rights for the new format, update facts and disclosures, adapt accessibility elements, then send the new item through the normal review and scheduling workflow.

Retire when an item should no longer be reused. The live action may be archive, unpin, unlist, correction, removal, or simply a record that blocks future reuse. Deletion is not automatically correct. Contracts, investigations, records obligations, public accountability, or platform limits may require a different approach. Record who authorized the action and verify the final state.

Product observation: the PostTempo application is built around draft, scheduled, publishing, published, failed, and approval-related states. Its calendar can surface posts that need a fix, approval, or retry. That visible state model informs this guide's recommendation to reconcile the inventory with both the calendar and the live account instead of treating every scheduled record as successfully published.

Product observation: PostTempo's approval interface records pending and decided requests and supports Approved, Rejected, and Changes Requested outcomes. This supports the operational idea that a remediation action should have a clear decision, not an informal comment. The audit still needs the organization to decide which changes require legal, compliance, rights, client, or executive approval.

Product observation: implemented media preflight and platform publishing code checks media and surfaces validation, authorization, processing, retry, and failure conditions. Those experiences inform the recommendation to connect source assets, platform versions, and final status. They do not establish that every historical post was validated under today's rules or that a platform will accept a future upload.

The PostTempo Editorial Team builds and tests social scheduling, media validation, approval, calendar, queue, retry, analytics, and publishing-status workflows. That relevant product experience helps us explain where content records become unreliable and where ownership matters. It does not provide legal credentials, platform authority, accessibility certification, or a proprietary performance dataset.

Days one through three: write the scope, field dictionary, action definitions, access plan, risk escalation rules, and delivery format. Confirm who owns the final decisions. Create a secure location for raw exports and a separate working inventory. Do not request broad account access until you know which data each person needs.

Days four through ten: request exports, collect scheduler records, list source asset folders, and reconcile the account list. Normalize identifiers and dates while preserving raw evidence. Test a small set of records from beginning to end. If the fields do not support a decision, fix the inventory design before expanding it.

Days eleven through twenty: perform the inventory and evidence passes. Prioritize pinned posts, active offers, sponsored or licensed content, regulated topics, high-traffic destinations, reusable campaign assets, and accounts with staffing changes. Record Unknown when evidence is missing. Route high-consequence questions to qualified owners rather than improvising an answer.

Days twenty-one through twenty-six: assign Keep, Refresh, Repurpose, Retire, or Hold pending evidence. Consolidate duplicate findings into projects without losing the affected URLs. Estimate remediation separately, name owners and approvers, and put realistic work into the calendar. Do not let a future due date turn an unsafe item into an approved item.

Days twenty-seven through thirty: obtain decisions, verify urgent corrections, remove temporary access, store the decision log, and write a short summary of scope, method, limitations, findings, actions, and unresolved risks. Schedule the next review based on content volume, campaign cycles, staff changes, and risk rather than claiming that one universal quarterly cadence fits every organization.

Methodology: we inventoried PostTempo's blog registry, SEO pages, comparison routes, canonical redirects, blog hub, sitemap inputs, and existing article intents. We selected an informational content-audit intent because it does not replace the content calendar guide, pre-publication checklist, crisis response plan, approval workflow page, scheduler comparisons, or platform timing pages.

We inspected implemented scheduling, approval, media validation, calendar, analytics, queue, retry, and publishing-status behavior to identify first-hand product observations. We researched current first-party sources from LinkedIn and Meta for exports and access, W3C for image text alternatives, the FTC for social endorsement disclosures, the U.S. Copyright Office for permission, and NIST for least privilege.

Sourced facts are linked near the relevant statements and listed below. Calculations show their inputs and assumptions. Product observations describe repository behavior available to the Editorial Team. Recommendations such as the four action states, hard evidence gates, and thirty-day sequence are editorial opinions that teams should adapt.

The named creator, agency, and brand scenarios are fictional composites. They are not customer stories or performance evidence. Platform documentation, interfaces, permissions, export fields, metrics, edit options, and policy enforcement are volatile. Confirm current details with the platform. Consult a qualified local professional for copyright, advertising, accessibility, privacy, record retention, contracts, employment, regulated claims, and other high-consequence decisions.