A social account can be busy without being useful. You can publish often, collect reactions and still be unable to answer three ordinary questions: Who is this for? What are we helping them do? What would tell us whether it worked? This chapter helps you write those answers down before you make a calendar of posts.
You do not need to be on every platform or know every current feature. By the end, you will have a one-page social media strategy: an audience question, a useful objective, a channel decision, a promise to readers, a realistic way of working and a review plan. The next chapter can turn that strategy into specific content. If you have never managed a social account, you can use the blank sheet near the end without inventing results.
The simple meaning: a decision about whom to help and why
Social media marketing means using social platforms as part of an organisation's communication with people. That may include explaining, listening, answering questions, hosting a community, encouraging a visit to a useful page or supporting an existing relationship. A social media strategy is the set of choices that gives those activities a purpose: whom you want to serve, which channels make sense, what value you will offer, what action is appropriate and how you will learn. It is different from a list of posts or a promise to “go viral”. A current practical guide from Hootsuite describes a strategy as a plan linking social activity to business objectives, audience, platform, content and measures (Newberry and Wood, 2026).
Organic social refers to content and conversations distributed without paying for a particular advertisement placement. It may still take substantial time and production cost, and a platform does not promise that every follower will see a post. Paid social buys a defined promotional opportunity through a platform. The same organisation may use both, but they need separate budgets and clear measurement. An owned destination is a page or service the organisation controls, such as its own website; a social profile lives inside another organisation's platform. Moving someone from a short post to a detailed page can make sense when the full answer will not fit the post.
An academic way to think about social platforms also helps here. Kietzmann et al. (2011) describe several social-media functions, including conversation, sharing, relationships and reputation. Their point is that platforms differ in which functions they emphasise. For a beginner, this means that “be on social media” is too broad to guide a useful decision: decide which kind of interaction your audience actually needs. The article is a conceptual framework, not current evidence that a specific platform will reach a specific audience.
The strategy is not a prediction that a chosen platform will produce customers. It is a reasoned working hypothesis that you can test. Write down what you believe, what evidence supports it, and what remains unknown. This is especially important when a new account has no useful baseline.
Figure 1
Figure 1 in words: In SOSTAC®, first understand the situation, choose an objective and a strategy, then specify tactics, action and control. In RACE, a social interaction may help someone discover you (Reach), take a useful step (Act), buy if appropriate (Convert) or maintain a relationship (Engage). The diagrams are a simplified original application of the frameworks, not a rule that every post must sell something (Smith, no date; Chaffey, no date).
Start with the problem the audience is trying to solve
A target audience is the particular group of people you intend to help or reach. “Everyone on Instagram” is not a usable audience. Neither is a demographic label on its own. A person can match an age band and still have no reason to care about the topic. A better starting point is a question, difficulty, decision or interest that you can describe in everyday language.
Look for evidence you can inspect lawfully: questions people ask in public comments, your own customer-service themes, search queries, consented research, frequently visited help pages, or conversations you have actually had. Record where an observation came from and its date. One loud comment is an idea to investigate, not proof that a whole market wants the same thing. If you have no audience evidence yet, label your audience description hypothesis. You can test it by showing a draft question or explanation to a small number of suitable people rather than claiming an invented persona is research.
Ask what the person knows before they encounter your post. A newcomer may need a definition; someone comparing options may need criteria and limitations; an existing customer may need a short explanation or support route. A customer journey is a way to describe such changes in needs over time. RACE names four broad types of activity: Reach, Act, Convert and Engage. Chaffey explains that Act is a meaningful intermediate interaction, while Engage concerns longer-term customer relationships (Chaffey, no date). Use these as prompts, not labels that must be visible to the reader.
For each audience question, write what a genuinely helpful answer would contain. Perhaps it needs a definition, a short demonstration, a checklist, a comparison, a correction to a common misunderstanding or a route to ask a human. You can decide the format later. If the topic cannot be explained honestly in a short social post, make the post a clear doorway to a longer explanation. Do not remove important conditions just to force the answer into a fashionable format.
Figure 2
Figure 2 in words: Write the person's question, note your evidence and its limits, identify their likely stage, decide what an adequate answer must include, and mark any remaining unknown. The loop starts with the question rather than a platform trend.
Choose an objective that an account can reasonably help
An objective is the intended result, not the work itself. “Post three times a week” is an activity. “Help suitable newcomers find and use our introductory guide” describes a possible outcome. SMART is a check for whether an objective is specific, measurable, achievable, relevant and time-bound. It does not guarantee that the chosen target is sensible. The number and deadline must come from a real baseline, available resources and a decision about what improvement matters; leave them blank while those facts are unknown. Hootsuite's strategy guidance also uses SMART goals to connect social plans to business outcomes (Newberry and Wood, 2026).
For a beginner, it is often enough to choose one primary purpose for a small initial period. You may want relevant people to discover an explanation, ask better questions, begin a sign-up process, use a support resource or return for further help. The action should fit the person's stage. Someone who has only just encountered a new concept may reasonably save an explainer rather than make a purchase. Do not write a sales objective for a post whose actual job is to answer a support question.
Now decide on an indicator: a number or observation that would give you some evidence about progress. A measure close to the desired action is usually more informative than an easy but distant number. For example, if the purpose is to send people to a guide, qualified visits and what visitors do there are more relevant than total likes alone. A key performance indicator (KPI) is simply a chosen priority measure. It is not a certificate that the measure captures the whole outcome.
Set a baseline before a target if possible. A baseline is a starting observation collected using a stated definition and period. If there is no earlier data, the first review may be to learn the baseline. Record the date, platform, unit and source. “Reach” and “views” are platform-defined measures and may use different counting rules; do not add unlike numbers or turn them into a claim about unique people without checking the definitions. A click is an action at the link, while a website visit can be affected by loading, consent, attribution and analytics settings. Even a later purchase may have been influenced by other channels. Good measurement supports a decision without pretending to prove more than it can.
Figure 3
Figure 3 in words: Publishing is an activity; relevant response or visits are signals; a useful action or customer outcome is the intended result. Choose a period and baseline for each measure before comparing. A large activity number cannot substitute for an outcome.
Take stock before adding another channel
A social audit is a short inventory of what you already have and what it does. It is not a judgment about whether a profile looks fashionable. Write down the current accounts, who owns access, their audience and purpose, what they published recently, visible questions, current links, response arrangements and any credible performance records. Check whether profile descriptions and destination pages still tell the truth. If an account is dormant, note it plainly. An old account may be useful, confusing or costly to maintain; the strategy should decide what happens to it.
Look at the content your intended audience can already find, including questions and answers on other organisations' public pages. This is competitive observation, not an instruction to copy a rival. What do they explain clearly? Which questions appear unanswered? A public feed shows only part of another organisation's performance; it does not reveal their private audience, cost, targeting or conversions. Keep a distinction between what you observed and what you inferred.
You may also use social listening: looking for recurring themes and language in relevant public discussion. A small, dated manual sample can be more useful than an unexplained dashboard. Avoid collecting personal details you do not need, and do not treat an individual comment as permission to quote or identify the person in marketing. Make notes about themes, not a dossier on strangers. If a topic concerns sensitive circumstances, get the right review before publishing advice.
The audit should end with a decision, not an endless spreadsheet. Which existing account has a clear role? Which needs a corrected profile or working link? Which audience question appears strong enough to test? Which assumption is still unverified? There may be an excellent reason to maintain one channel well before opening another. Newberry and Wood (2026) place audience research, a review of existing activity and measurement within social strategy; use that sequence as a prompt rather than a fixed number of required accounts.
Figure 4
Figure 4 in words: Inspect account ownership and access, audience evidence, recent content and questions, destination links, response capacity and measurement. Record evidence and one decision for each area; mark unavailable data as unknown.
Select a platform by fit, then choose a manageable role
There is no permanently “best” platform for all organisations. Platforms change features, audiences and rules. Begin with what your audience is trying to do, where you have evidence they spend time, which format helps the topic, and whether you can sustain both publishing and replies. An interesting platform can still be the wrong choice if you cannot explain your subject well there or maintain the account responsibly.
Think of a channel role as a sentence. One channel might help new people discover a topic. Another could hold a longer demonstration. A community space could support questions among existing users. A professional network could help a specific work audience compare an approach. These are possible roles, not claims that a platform always works a particular way. Test the role against your own audience evidence and the platform's current capabilities.
Consider the format honestly. Does your idea need sound, moving images, a diagram, readable text, a searchable longer explanation or live discussion? Can it be made accessible? Can you answer comments at an appropriate speed? If you need a detailed response, a profile that directs people to an owned guide and a contact route may be more useful than squeezing the whole explanation into one graphic. W3C guidance recommends meaningful image text alternatives and captions for audiovisual material; plan that work when you choose the format, not after publication (W3C, no date).
Record a reason for not adding a second channel yet: insufficient audience evidence, no owner, no suitable format or no capacity to respond. That is a valid strategic choice. You can reconsider after a review. A platform list becomes a strategy only when each channel has a distinct audience, job and owner.
Figure 5
Figure 5 in words: Check audience evidence, topic and format fit, accessibility, capacity to publish and respond, and the route to a useful next step. Select a manageable primary channel, test it and record the decision before expanding.
Give people a reason to return
A content promise states what someone can expect from you. It might be a commitment to explain a difficult subject clearly, share practical answers, show how something works, or respond to recurring questions. This promise must match what you can actually deliver. A vague promise to provide “amazing content” gives nobody a reason to follow. A narrow promise can be changed as you learn, but consistency helps a reader decide whether the account is for them.
Content pillars are a small set of recurring subject areas that support that promise. They are planning buckets, not hashtags or compulsory categories. For instance, you might choose “explain”, “show how”, “answer questions” and “changes or limitations” as functions. Do not fill these with made-up customer stories. You can use verified information, your own process, a documented question that you have permission to discuss, or a blank worksheet. The next chapter will develop specific formats and a calendar.
Decide how the organisation will sound. Voice is the stable manner of speaking: clear, careful, warm, technical when necessary or direct about uncertainty. Tone can change by situation. A routine explanation and a response to a service problem should not sound identical. Write a few rules that a person could apply: define unfamiliar terms, distinguish fact from opinion, say when information changes, link to the full source and avoid claiming a result you have not checked. A real human reviewing the copy can protect that promise better than a stock phrase pasted into every post.
The offer of value matters more than a quota of promotional messages. A useful post can answer the question on its own and still offer a relevant deeper link. A call to action is an invitation to do something next. It should say what the person will find and fit the stage: “read the steps”, “compare the options”, “ask a question” or “try the blank worksheet” each makes a different promise. Avoid pressing a newcomer toward a purchase before they have the information needed to choose.
Figure 6
Figure 6 in words: The audience's recurring question informs a content promise. A few useful themes support it; a suitable format and next step make each explanation usable. The promise stays stable while individual posts can change.
Plan for replies, errors and uncomfortable questions
Social media is not merely a publishing pipe. People may ask for clarification, raise a problem, challenge a claim or share something private. A strategy needs a response route. Decide who checks the account, how often it can reasonably be checked, who can answer subject questions, and when to move a conversation to a private support channel. Do not ask someone to disclose sensitive account or personal details in a public reply. State realistic service hours if you are unable to monitor continuously.
Set a simple escalation rule: if a question involves a factual error, personal data, safety, a legal or regulated claim, or a distressed customer, pause the standard reply and send it to the right owner. The owner should know what information to preserve and who can correct a published post. It is acceptable to say that you are checking a point; it is not acceptable to fabricate an answer to look fast. Distinguish moderation of abuse from suppression of legitimate criticism.
Plan for access and continuity. Use the platform's current security options, appropriate role-based access and a second responsible owner where the organisation requires one. Keep a record of approved copy, sources and corrections. If a person leaves the team, the account should not depend on their personal password. Tool names and settings change, so check current platform documentation before writing a step-by-step access policy.
If another person is paid, gifted or otherwise incentivised to post on your behalf, Chapter 19's creator-partnership principles apply: the commercial relationship and claims need proper review and clear identification. A brand's own organic post, a creator's sponsored post and a paid platform advertisement are different situations. Do not assume one label or approval process covers all three. The strategy should say who checks a collaboration before it goes live.
Figure 7
Figure 7 in words: Monitor and acknowledge; answer a routine question with a checked source; route private or specialist issues to the appropriate owner; correct an error openly; record a recurring question for future content. The diagram is an operational decision aid, not a service-level promise.
Make measurement possible without pretending to know everything
Your review plan needs two kinds of evidence. Platform evidence can show how content was distributed and how people interacted within the platform, using its own definitions. Destination evidence can show whether visitors reached an owned page and completed a useful task, if analytics and consent arrangements permit it. These numbers rarely line up perfectly. Different platforms, browsers, privacy choices and reporting windows can produce gaps.
For links to your own website, campaign parameters can help identify where visits came from. Google's Analytics guidance defines utm_source, utm_medium and utm_campaign and recommends using them consistently. Parameter values are case-sensitive (Google, no date). You do not need to put an example URL into your first strategy; just record who will create and check the link, and whether the destination works on a phone. Analytics setup and interpretation belong with an owner who knows the site's configuration. Do not put personal information into campaign parameters.
Keep a small measurement ladder. At the bottom, was the planned material published correctly and accessibly? Next, did relevant people encounter or respond to it? Then, did they take the intended next step? Finally, what useful outcome, if any, can you observe? Each level answers a different question. “This post got many views” is not the same as “this post caused sales.” If you cannot observe the later step, mark it unknown and decide whether better measurement would be worth the effort.
Review quality as well as quantity. Did a question reveal a missing definition? Did people click a link but leave because the landing page was unclear? Did support replies take more time than the team can sustain? Did a post prompt people to misunderstand a condition? These observations may justify changing the explanation or destination without changing the channel. A review is useful when it ends in a decision: continue, revise, pause or test something specific.
Figure 8
Figure 8 in words: Check publication and accessibility, distribution, meaningful response, the next action and any observed outcome. Add definitions, dates and limitations. A lower rung cannot automatically prove the next one.
Write the one-page strategy
Here is a blank decision sheet. Complete it with your own evidence, not imagined followers or results. The purpose is to make a small plan that another person could understand and review. If a field is unknown, say what evidence you would need rather than filling it with an attractive guess.
Audience question. What does the intended person need to understand, decide or do? Where did that question come from? Is it observed or a hypothesis?
Purpose. What useful change should social help create? What should the person be able to do after encountering the explanation?
Initial objective. Write one outcome, an indicator and a review date. Add a target only when you have a defensible baseline and resources.
Primary channel and role. Which platform will you test first, for whom, and why does the format fit? What will you leave out for now?
Content promise. What will someone consistently get from you? What are the few themes that support it? What must you verify before you post?
Useful next step. Can a reader get the complete answer in the post, or do they need a linked guide or a support route? Who checks the destination?
Capacity and responsibility. Who makes the material, reviews facts and accessibility, publishes, monitors and responds? What can that team sustain?
Review. Which measures and qualitative signals will you inspect, with what definitions and dates? What finding would make you continue, revise or stop?
Figure 9
Figure 9 in words: The blank sheet has fields for audience evidence, purpose, objective and baseline, primary channel, promise and themes, next step, owners and response route, measures and review decision. Empty spaces are intentional; a reader should fill them with real information or “not verified”.
A good first draft might be short. Its strength is that it connects its parts. If the audience question is about understanding a technical term, the promise should offer a clear explanation, the format must allow enough detail, the next step should be a deeper guide if needed, and the measure should tell you whether the guide was reached or the question was answered. If any one piece is disconnected, revise the plan before increasing the number of posts.
What to do when the first attempt is quiet
Quiet results do not immediately prove that the platform, topic or organisation is wrong. First check whether the account had a real audience to reach and whether the material was actually visible. Then inspect whether the first line explained the value, whether the format fit the question, whether a link worked and whether the next step was worth taking. Compare periods only when their length and definitions match. If the audience evidence was weak, return to the question rather than changing a headline at random.
If people respond but ask the same basic question, the explanation may assume knowledge they do not have. Define the term, show the missing step, or link to a longer answer. If posts are saved or shared but the intended page is not visited, perhaps they already answer the question; perhaps the invitation is unclear; perhaps the platform prevents easy linking. Check before declaring a failure. If a visit occurs but the destination does not help, revise the page rather than producing more traffic.
If the team cannot keep up with replies, reduce the scope or improve the support route before making more publishing promises. If a claim is wrong, correct it, review its source and update the approval process. The point of a strategy is to make those decisions with evidence, not to keep an unrealistic schedule alive. Smith's SOSTAC® framework explicitly places metrics and feedback in Control and returns the learning to the next situation analysis (Smith, no date).
Figure 10
Figure 10 in words: Low visibility calls for a distribution and audience check; repeated confusion calls for a clearer explanation; visits without progress call for a destination check; overloaded replies call for a capacity decision. Record the observation, the first check and the next test.
Common misunderstandings
“We must be everywhere.” Every channel adds publishing, monitoring and learning work. Begin where audience evidence, topic, format and capacity meet. Expand when a new channel has a distinct purpose and owner.
“Followers are the goal.” Follower counts may describe part of an account's audience, but they do not tell you whether the right people understood, acted or benefited. Select measures for the stated purpose.
“A strategy is a calendar.” A calendar names dates and pieces of content. The strategy explains why those pieces should exist, for whom, on which channel and with what review. Chapter 21 will turn this into a content plan.
“Organic is free.” You may not buy ad placement, but research, creation, captions, checking, replies and measurement use time and resources. Include them in the decision.
“If a post goes viral, the plan worked.” A large distribution number may involve people outside the intended audience and does not establish a useful outcome. It may create more questions than the team can answer. Review the actual goal and evidence.
“The platform dashboard tells the whole story.” It shows only what the platform measures under its definitions. Use relevant destination evidence and qualitative learning where appropriate. Avoid claiming a causal result from one dashboard number.
Check your understanding
Take an actual social account or a blank plan. Can you write one sentence each for the audience question, channel role, content promise and useful next step? Then ask whether the proposed measure would tell you anything about that next step. If it only counts how much you publish, choose a better indicator or mark the measurement gap. A clear “we do not know yet” is a useful finding.
Watch, if a visual explanation helps
Start with a step-by-step plan:How To Create a Social Media Strategy Plan (FREE Template), HubSpot Marketing, 2024. Its public description lists goals, audience, competitors, platforms, content and measurement. Use it to see a planning sequence, then fill our blank sheet with your own evidence. HubSpot is a commercial software provider and the video's platform details should be checked against current practice. Exact runtime and captions were not independently verified.
Go deeper into the decisions:Ahead of the Game Podcast Episode 16: Practical Social Media Strategy, Digital Marketing Institute, 2021. The public description identifies sections on listening, content, platform choice and whether organic work can stand without ads. It is a longer professional discussion with a consultant, not a peer-reviewed lecture; some platform examples may have aged. Exact runtime and captions were not independently verified. Both videos are optional, and the written strategy sheet contains the complete exercise.
Kietzmann, J.H., Hermkens, K., McCarthy, I.P. and Silvestre, B.S. (2011) 'Social media? Get serious! Understanding the functional building blocks of social media', Business Horizons, 54(3), pp. 241–251. Available at: https://doi.org/10.1016/j.bushor.2011.01.005 (Accessed: 15 November 2026).
Smith, P.R. (no date) SOSTAC® 1-Day Masterclass – AI enabled. Available at: https://prsmith.org/sostac/ (Accessed: 15 November 2026).
World Wide Web Consortium (W3C) (no date) Writing for Web Accessibility – Tips for Getting Started. Available at: https://www.w3.org/WAI/tips/writing/ (Accessed: 15 November 2026).
Further watching
Digital Marketing Institute (2021) Ahead of the Game Podcast Episode 16: Practical Social Media Strategy. 15 April. Available at: https://www.youtube.com/watch?v=JwDSSxwLG6Y (Accessed: 15 November 2026).
HubSpot Marketing (2024) How To Create a Social Media Strategy Plan (FREE Template). 29 January. Available at: https://www.youtube.com/watch?v=VS0Sao1oHlw (Accessed: 15 November 2026).
A beginner-friendly route through 26 short lessons on customers, offers, channels, content and measurement. Start anywhere, with no account or paid tools.
Learn the difference between features and benefits, write a promise customers understand, and check which kinds of online value you can honestly deliver.