Strategy
Part of A social scheduling strategy that begins with a communication decision, not a calendar
A ninety-day social scheduling plan as a calendar of deliverables and gates
A social media scheduling ninety day plan for England as a calendar: what to deliver in each period, which gate can stop it and what the decision pack holds.
A ninety-day plan for social media scheduling works as a calendar of deliverables, each period closed by a gate that can stop the work. It covers one England audience task and the strategy behind it. The phases show dependencies; they do not promise that procurement, legal review or platform approval will finish in ninety days.
What to take away
- Each period ends with a named deliverable and, in four cases, a gate, so a stop is a planned event and not a surprise.
- Pause the ninety-day clock whenever a critical gate lacks supporting evidence.
- Name the shortlisted tools, such as Buffer, Hootsuite, Later, Sprout Social or SocialPilot, and the platforms they publish to.
- Independent decisions on privacy, security, accessibility and claims cannot be replaced by a weighted score.
- Compare the new route with the baseline using identical definitions, periods, timezones and exclusions.
- Rehearse export, credential return, deletion and manual fallback before expanding the approach.
The calendar at a glance
Six periods carry the plan, and four of them end in a gate. When a gate lacks evidence, pause the clock rather than skip it. A skipped gate turns every later test result into a guess about a route nobody has checked.
| Days | Deliverable | Gate that can stop it |
|---|---|---|
| 1 to 15 | Owner, audience, geography rule, platforms, workflow states and the manual or no-post baseline | Gate 1: stop if the audience, purpose, owner or current workflow cannot be evidenced |
| 16 to 30 | One-timezone baseline log and an inventory of account roles | Gate 2: continue only if the baseline allows a like-for-like test and account ownership is clear |
| 31 to 45 | Release specification for a synthetic item, with claim sources, rights records and accessibility tests | Gate 3: hold when any critical outcome is fail or unknown |
| 46 to 60 | Controlled acceptance tests in non-production | Gate 4: reject the route if rollback, evidence capture or administrator recovery fails |
| 61 to 75 | Comparison against the baseline | None named; unknowns carry forward |
| 76 to 90 | Decision pack and exit rehearsal | The decision itself: continue, revise or retire |
Days 1 to 15: write down the job
Name the accountable owner, the intended audience, the England geography rule, the platforms in scope and the recent communication event. For platforms, think of Facebook, Instagram, LinkedIn, X or TikTok. Record all of it in the planning template, then list the tools that could carry the work.
Map the brief, sources, rights, drafting, review, approval, timezone, release, monitoring, correction and incident states. In a shortlisted tool this becomes the content calendar, the publishing queue, the posting cadence and the approval workflow: who drafts, who approves, when a post leaves the queue and who can pull it back. Record who can move each state.
Define the job workflow
- Name accountable owner
- Define intended audience
- Set England geography rule
- Classify content elements
- Write manual baseline
- Map approval states
Write the manual or no-post baseline beside the shortlist. The ICO's guidance on planning direct marketing says to think about what direct marketing activity you want to carry out before you collect the information. It supports purpose and channel analysis, but it does not clear a campaign.
Days 16 to 30: observe the current route
Observe a defined set of eligible items through the current route on each platform in scope. Facebook, Instagram, LinkedIn, X and TikTok handle scheduling, timezones and failure states differently. Record timestamps in one timezone, approval changes, platform confirmation, failures, corrections and staff minutes.
Do not insert example results or compare the log with an industry benchmark. The baseline is your own route, observed on your own items. A baseline that borrows a benchmark cannot show whether a new tool changed anything.
Baseline observation checklist
- Record timestamps in one timezone
- Log approval changes
- Note platform confirmations
- Track failures and corrections
- Record staff minutes
- Inventory account roles and recovery
Inventory account roles and recovery in the same fortnight. If someone with authorised access is publishing damaging content, the NCSC says you must be able to quickly revoke their access, most likely remotely. Remove unnecessary access through the approved process, and do not disrupt production to satisfy the study.
Days 31 to 45: build the release specification
Complete the plan for a synthetic item. Attach claim sources, rights records and the exact help page for each channel in scope, such as the Meta Business Help Centre for Facebook and Instagram, the LinkedIn Help Center, the X Help Center or TikTok Support.
Check each platform's publishing rules for image and video ratios, caption length and link handling. Define accessibility tests for text alternatives, captions, contrast, reading order, keyboard use and errors. W3C's WCAG 2 material is a technical reference, not proof of the final social output.
Obtain independent decisions for privacy under the UK GDPR and PECR, where the ICO is the supervisory authority. Add advertising and consumer claims under the Advertising Standards Authority's CAP Code and the Competition and Markets Authority's consumer law, plus security, accessibility, intellectual property, employment, contract and finance. No weighted score can compensate for failure in one of these areas.
Days 46 to 60: run controlled acceptance work
Use non-production access and synthetic content where platform rules allow it, such as a trial or sandbox workspace connected only to test accounts. Test role changes, exact formats, release time and timezone, late edits, rejection, alerts, correction, export and recovery. Keep supplier claims separate from observed test records.
Revoke one test role in the tool and show that access disappears, including in the connected platform's own account settings. Simulate an unavailable approver and prove that the queue stops. Do not connect customer data merely to make the scenario realistic.
Days 61 to 75: compare with the baseline
Use the same eligible-item definition, period unit, timezone, inclusions and exclusions as the baseline. Compare staff effort, release evidence, exceptions and full economic cost. Treat reach, response and revenue as separate outcomes needing their own design, because scheduling timestamps do not establish causality.
Ask the service owner whether the operational difference matters. Ask specialist reviewers whether any control changed. Preserve disagreement and unknowns in the pack rather than averaging them away, so the final decision can see where the evidence is thin.
Days 76 to 90: decide and exit
The decision pack holds the task, audience rule, baseline, test configuration, source versions, gate outcomes, incidents, costs and limitations. Continue, revise or retire are the three outcomes. If continuing, set an owner and a review trigger for every dependency.
Rehearse the exit in the tool you are testing. Export the content calendar and post history, disconnect each connected account, revoke the tool's access from the platform side, return or delete credentials under the contract and delete the workspace. Add manual fallback and supplier failure to the same rehearsal.
Retire the approach if the existing route meets the need, the audience task has ended, critical evidence cannot be retained or the business case depends on assumed engagement or sales.
Before you act
- Name the accountable owner and audience rule.
- Write the manual or no-post baseline.
- Record timestamps in one timezone throughout.
- Attach claim sources and rights records.
- Revoke one test role and confirm access disappears.
- Rehearse export and credential return before expansion.
Common questions
What should happen when a critical gate lacks evidence?
Pause the clock. Gate 1 stops the plan if audience, purpose, owner or workflow cannot be evidenced. Gate 3 holds when any critical outcome is fail or unknown. Gate 4 rejects the route if rollback, evidence capture or administrator recovery fails. A skipped gate leaves later results unreliable.
How should the new route be compared with the baseline?
Use the same eligible-item definition, period unit, timezone, inclusions and exclusions. Compare staff effort, release evidence, exceptions and full economic cost across the shortlisted tools. Treat reach, response and revenue as separate outcomes needing their own design, because scheduling timestamps do not establish causality.
What belongs in the decision pack at the end of the ninety days?
The task, audience rule, baseline, test configuration, source versions, gate outcomes, incidents, costs and limitations. Then choose continue, revise or retire. If continuing, set an owner and review trigger for every dependency. Retire the approach if the existing route meets the need or critical evidence cannot be retained.
