In most small companies the social media strategy looks like one of two things. Either there isn’t one and posts go out “when there’s something to say”, or it’s a 40-slide deck from an agency that was opened once, at the pitch. Both end the same way: two months in, nobody can explain why yesterday’s post went out or how to tell whether social is working.
Why one page
A working strategy fits on a single page. Not because it’s fashionable, but because a single page actually gets re-read: before planning the week, during the “let’s start a TikTok” argument, when hiring a social media manager. Below are seven blocks, each a question you answer in one or two lines, and at the end a filled-in example for a small coffee roaster.
The strategy answers “why and for whom”, not “what to post on Thursday”. The content plan sits on top of it; there’s a separate breakdown with a template for that.
The seven blocks
Block 1. Business goal and social goal
“Increase awareness” isn’t a strategy, because you can’t tell whether it worked. Start from a business number and work backwards to what social can actually control.
A worked example for a language school (illustrative numbers, plug in your own):
| Step | Figure | Source |
|---|---|---|
| New students needed per month | 30 | Sales plan |
| Lead-to-payment conversion | 30% | Your CRM, last 3 months |
| Leads needed | 100 | 30 / 0.3 |
| Profile-visit-to-lead conversion | 5% | Last quarter’s stats |
| Profile visits needed | 2,000 | 100 / 0.05 |
The social goal here isn’t “30 students” but “2,000 profile visits and 100 leads a month”. Content drives the first; the second also depends on your profile and reply speed. If you don’t know your conversions yet, write a hypothesis and mark it as one: replace it with a fact in a month.
For companies that don’t sell directly through social, phrase the goal as retention: “70% of customers follow the channel and see announcements”, “support requests in DMs are closed within an hour”.
Block 2. Audience: one or two segments
Not a persona like “Anna, 32, loves travel”, but the job the person is trying to get done. Four lines per segment:
- Who: parents of 10–14-year-olds in a mid-sized US city.
- Job to be done: get the kid ready for an English certificate in two years without a private tutor.
- Where they are: school parent group chats, Instagram.
- What stops them buying: fear the kid quits after a month; can’t see how this school differs from the one next door.
The last line is the most useful: it’s where the content pillars that remove objections come from. Add a second segment only if it buys a different product or for a different reason. How to see which objections your competitors answer is covered in the competitor analysis article.
Block 3. Channels: one primary, one supporting
Five platforms “just in case” means five accounts run badly. Criteria for the primary channel, in order:
- Where your audience already discusses your topic, not where it merely “exists”.
- Which format you can produce consistently: if nobody on the team shoots video, Reels and Shorts won’t be your base.
- Where the sale happens: for many local businesses that’s DMs or WhatsApp, for B2B it’s a LinkedIn message or an email.
Typical pairings:
| Business | Primary | Supporting |
|---|---|---|
| Local services, food, retail | Facebook Page or a Telegram channel for regulars | |
| Expert, consultant, education | YouTube or a Telegram channel | Threads or LinkedIn |
| B2B, SaaS | X | |
| Brand with a visual product | Threads for the conversational side |
The supporting channel gets adaptations of the primary content, not its own program. Everything else goes in a line called “Not running and why”, so the argument doesn’t come back every quarter.
Block 4. Positioning and tone
One sentence by formula: “We are [what] for [whom] who [job]. Unlike [alternative], we [difference].” For the roaster: “We roast coffee for people who brew at home and are tired of burnt supermarket arabica. Unlike chain cafés, we show the roast date and origin of every lot.”
Underneath: three words that describe the tone and three that are banned. For example “neighbourly, specific, dry humour” and “no flattery, no ‘unique’, no stacked exclamation marks”. This block goes straight into your AI assistant’s knowledge base and a freelancer’s brief; details in the tone of voice breakdown.
Block 5. Content pillars and cadence
Three or four pillars with percentages, each answering an objection from Block 2 or feeding the goal from Block 1:
- Useful 40%: how to brew, how to store, how to pick a grind.
- Product 30%: new lot, roast date, price, how to order.
- Behind the scenes 20%: the roaster, the farm, a rejected batch.
- Customers 10%: reviews, photos of home coffee corners.
Set the cadence from hours, not wishes. Three hours a week on social is roughly three feed posts and five Stories, not “daily”. Write the number you can hold in the worst week of the month: a broken rhythm hurts more with the algorithm than a slow one.
Block 6. Metrics: one primary, three supporting
The primary metric follows directly from Block 1: profile visits, leads, channel subscriptions. Supporting metrics explain why the primary one moves:
| Goal | Primary metric | Supporting |
|---|---|---|
| Leads | Leads from social per month | Profile visits, link clicks, share of non-follower reach |
| Customer retention | Share of customers in the channel | Post views to subscribers, reactions, unsubscribes |
| Expert brand | Inbound consultation requests | Saves, shares, mentions |
Likes and follower count stay out of the table: they also grow from giveaways that bring no leads. Review monthly, 30 minutes, in the same spreadsheet. What to count and how to present it to a boss is in the social media metrics article.
Block 7. Resources and rules
Four lines: hours per week, monthly budget for promotion and tools, who writes and who approves, and after how many days a post counts as overdue. The last one kills endless approval loops: if no edits arrive in two days, the post goes out.
Filled-in example: a small coffee roaster
- Goal. 60 orders a month via social. DM-to-order conversion 40% → 150 conversations.
- Audience. Home brewers, 25–45, within delivery range. Job: coffee better than the supermarket’s, delivered next day. Blocker: don’t understand lot differences, afraid of overpaying.
- Channels. Primary Instagram (brewing Reels, carousels on lots). Supporting Telegram channel: roast announcements, one-tap ordering. Not running TikTok: nobody for daily video.
- Positioning. See Block 4. Tone: neighbourly, specific, dry humour.
- Pillars. Useful 40 / product 30 / behind the scenes 20 / customers 10. Cadence: 3 posts and 5 Stories a week, roast announcement every Tuesday.
- Metrics. Primary: DMs labelled “from social”. Supporting: profile visits, saves on lot carousels, share of non-follower reach.
- Resources. 4 owner hours a week, $150 a month on boosted posts, approval no longer than two days.
That’s the whole strategy, explainable to a new hire in ten minutes.
When one page isn’t enough
- Several products with different audiences. One page per product line, not one shared page.
- A team of five or more with split roles. You’ll need a separate process document: who, when, in which tool.
- A launch with an ad budget of several thousand dollars a month. The media plan lives separately; the strategy only sets its goal.
- The strategy hasn’t been revisited in six months. If the numbers in Blocks 1 and 6 are stale, it’s an archive, not a strategy.
Blocks 5 and 6 need a tool where pillars, calendar and numbers sit side by side. In Repost21 the calendar suggests best slots for your cadence, account analytics for Instagram, YouTube and Threads show the supporting metrics per post, and Block 4 drops straight into the brand knowledge base the Luna assistant drafts from in your tone. You can test the loop on your own strategy for free.