Planning your Social Media content

Use Social Media Plan to decide who speaks for your business, what they should share over time, and how they should respond. A useful plan gives each voice a clear purpose, real evidence, suitable destinations, and boundaries your team can follow.

A publishing voice, also called a persona, is the speaker: your founder, brand, product, expert, or community. A customer profile describes the audience. This guide uses an illustrative bicycle-repair business to show the difference.

Open the plan and define its purpose

Open Assistant → Strategy → Plans → Social Media Plan. Choose Fill it in myself for an empty plan, or Add details, Edit, or Review suggestions on a section. Select Save after editing each section; you can save work in progress.

Your Brand Guidelines, Customer Profiles, and Marketing Strategy must be ready before this plan can show Ready to use. Use them to keep your social activity aligned with your audience, voice, priorities, and approved offers.

All three plan fields are essential: What social should achieve, Publishing voices & narratives, and Reply rules & coordination.

In Purpose & Audience, use What social should achieve to state the business purpose and audience. For the workshop, that could be helping everyday cyclists understand available services and start suitable booking conversations. Give the plan a useful outcome beyond publishing more posts.

For shared editing and answer states, see Setting up your business strategy in Cedros. Saving this plan does not connect social accounts or start a publishing schedule.

Add a voice and make it your own

In Publishing voices & narratives, choose the speaker you need: Add founder, Add brand, Add product, Add expert / employee, or Add community. Start with the voices you can support consistently.

For example, the workshop owner might share personal lessons and decisions, while the brand account explains services and answers common customer questions. The same event may be relevant to both, but each should have a distinct reason to discuss it.

New cards include suggested wording and recurring stories. Review these against your actual business. If you choose Review suggested starting point for the whole field, review the proposed voices, select Accept reviewed answer, then Save.

For each voice, complete the strategy details:

• Persona name, Persona type, and Primary goal: identify the speaker and their main purpose. Available goal choices depend on the type; choose Custom goal when the standard choices do not fit.
• Strategy and Goal definition: explain what the voice is trying to achieve and what useful progress would look like.
• Audience relationship: describe why this audience should listen to this speaker, such as practical expertise from someone who works with bicycles every day.
• Subject focus and Point of view: choose the subjects and judgments the speaker can support with real experience or evidence.
• Voice and writing style: describe how the speaker sounds while staying consistent with the brand.
• Content shapes: choose useful forms, such as a short explanation, a practical checklist, or a reflection on a real decision.
• CTA behavior: explain how posts invite a next step. A call to action might ask a useful question, point to a guide, or direct someone to the correct service page.
• Cadence and capacity: record a rhythm the team can sustain without inventing updates to fill gaps.
• Engagement rules, Success signals, and Boundaries: explain how the voice responds, what outcomes matter, and what it must not claim or discuss.

Use recommended shape replaces the card's strategy wording with a preset for its current type and goal. It keeps the existing name, narratives, sources, and accounts. Review your custom wording before using it, and check the result before saving.

Do not give a founder or employee invented personal experiences. A brand voice can explain an approved service without pretending that someone personally used it or spoke to a customer.

Build recurring stories with clear limits

Under Editorial narratives, use Add narrative to create a story the voice can develop across several posts. A narrative is a recurring subject with a point to make, rather than a queue of unrelated post ideas.

For each narrative, record:

• Narrative name: a clear label, such as “Understand your repair options.”
• Status: choose Active, Testing, Paused, or Completed. Configured social routines use active or testing narratives when planning content.
• Priority: choose Primary, Supporting, or Experimental to express its place in the voice's strategy.
• Thesis: the central idea you can support, such as “Clear service information helps cyclists ask better questions before booking.”
• Audience promise: what readers should understand or be able to do after following the story.
• Territory: the subjects that belong in the narrative.
• Story arc: how the subject develops over time, such as recognizing a need, understanding options, preparing for a visit, and knowing what happens next.
• Recurring questions: useful questions that can produce fresh, relevant posts.
• Narrative boundaries: claims, details, or topics that should stay out.

Review the narrative's Allowed sources after setting up its evidence sources. Select particular sources to limit it to those sources. Leaving all boxes unchecked allows every source belonging to that voice; it does not mean “use no sources.”

Select Use when no timely source opportunity exists only for a narrative that remains useful without a fresh event. For example, an explanation of the current booking process may be suitable. A claim about a new service or recent customer outcome needs current evidence.

Edit an existing narrative as the story develops. Use Paused or Completed when appropriate instead of repeatedly deleting and recreating the same story. Before changing a narrative used by a routine, review the related scheduled work separately.

Choose evidence sources and review expectations

Under Context sources, review the sources already present and use Add source when another source is needed. Newly added source rows start with GitHub settings; change Source type to the kind of evidence you intend to use.

For a simple manual plan, Manual research or evidence can describe material your team has checked. Complete:

• Source label: a recognizable name for the evidence, such as “Approved service information.”
• Filters and editorial guidance: what is relevant, what should be excluded, and what the writer must verify.
• Freshness window (days): how recent the evidence should be for this source. The control accepts 1–365 days.
• Review policy: choose the review expectation that fits the work. Owner review before external scheduling records that the owner should review before content is scheduled externally.

For the workshop, the guidance might identify the approved service descriptions and require a check whenever prices or availability change. Keep private customer details out of public examples unless you have appropriate permission to use them.

A source row describes where evidence should come from. It does not prove that the evidence has been retrieved, is accurate, or is suitable for publication. Check the actual material and the capabilities of any workflow using it.

The other policy, May schedule under approved strategy, is a broader planning choice. It does not by itself authorize or enable an external publishing action; a routine still needs the relevant permissions, connection, and approval settings. Choose it only when that is how you intend the work to operate.

If a source no longer belongs, use Remove source and review Allowed sources on affected narratives. Removing the last specifically selected source can leave a narrative with no boxes selected, which means all remaining sources for that voice are allowed.

Record the intended account destinations

Under Account destinations, select Add account and fill in Account label, Platform, and Handle or page. Use a real destination your team intends to use, and keep each voice associated with the correct account.

The platform field accepts the plan's supported names: X, Instagram, LinkedIn, Facebook, TikTok, YouTube, Threads, Pinterest, Email, or Custom. These names describe destinations in the plan; they are not a promise that every destination supports automatic publishing.

For example, a brand card might point to the workshop's actual business page, while an owner card points to the owner's own account. Verify the handle or page rather than relying on the label alone.

Saving an account here does not sign in, connect it, or grant publishing access. Configure and check those separately in the workflow you use. Likewise, a cadence written on the card does not activate recurring work. See Creating a recurring routine if you want to set up a supported routine after the plan is ready.

Coordinate replies and handoffs

In Replies & Coordination, use Reply rules & coordination to describe how the voices work together:

• Decide which voice handles each type of conversation and how to avoid repeating the same post or answer.
• Define what can be answered from approved information and what needs someone to check first.
• Name the route for support, pricing exceptions, complaints, or sensitive requests.
• Explain when to stop a public exchange and ask the appropriate person to help.
• Keep any customer offer consistent with its approved Marketing Strategy terms.

For the workshop, the brand account might explain the standard assessment process and direct a booking question to the service page. A complaint about a specific repair should reach the responsible person without sharing the customer's details publicly. The owner should not promise a refund or turnaround time beyond their authority.

Align these shared rules with each card's Engagement rules and Boundaries. When the plan is used by a configured workflow, the rules guide its work; they do not replace the workflow's actual permissions or approval requirements.

Review readiness before using the plan

A card's Routine readiness message highlights missing configuration, such as an account destination or a complete active or testing narrative. A configured narrative needs a name, thesis, and story arc. The voice also needs a source and supported destination details.

For the publishing-voices field to count as answered, at least one voice must have the required strategy details and complete configuration. Review every voice you intend to use, even if another complete voice already makes the field count as ready.

Before commissioning posts or configuring a routine:

1. Confirm the speaker, audience, goal, point of view, and voice.
2. Replace unsuitable preset wording and check the active or testing narratives, evidence, and fallback choices.
3. Verify sources, review policies, account destinations, and the actual connections needed for the intended workflow.
4. Check the cadence, responsible person, success signals, reply rules, and handoff routes.
5. Save each edited section and check readiness. Ready to use means the essential decisions and prerequisite plans are addressed; it does not verify account access, publish a post, or confirm that a routine is running.

Use Not known yet for unresolved decisions, or Not applicable with a reason where appropriate. If someone else has changed the shared plan, keep a copy of your unsaved wording before choosing Reload latest plan, then compare the latest decisions before editing again.

Optional assistant drafting and routine execution can use your configured AI service; see Understanding AI usage and costs. Review proposed content against the current evidence, offer terms, and destination before approving it.