How to Build a Social Media Strategy From Scratch

Most social media strategies are twenty slides that nobody opens twice. The useful version is five decisions on one page, and the order matters more than the wording.

The order matters because each decision constrains the next. Pick the channel first and you will spend a year making content for people who were never going to buy anything. Pick the audience first and the channel picks itself.

Decision 1: Who, specifically

Not "small businesses". Not "marketers". One person you could describe to a stranger.

The test is whether you can answer three questions about them without hedging. What is the thing that annoys them at work on a Wednesday? What do they already believe about your category? Where do they go when they want to feel better about their job?

If you cannot answer those, your content will be written to a demographic, and content written to a demographic reads like it was written for nobody, because it was.

Narrowing feels like giving up reach. It is not. A post that lands hard with two hundred right people beats one that is politely ignored by ten thousand.

Decision 2: What you want them to do

One action. Not awareness, not engagement, not brand affinity. An action a person takes with their hands.

Reply to this. Sign up. Book a call. Send it to a colleague. Read the thing I linked.

This decision does most of the work later, because it turns every content idea into a question with a real answer: does this move someone towards that action, or does it just fill a slot? A surprising amount of what gets published fails that test and is published anyway, because nobody wrote the action down.

Decision 3: Where those people already are

Now, and only now, pick channels.

Pick one. Two if you have a reason. The instinct is to be everywhere because being everywhere feels like effort and effort feels like progress, but five thin channels lose to one good one every time, and they cost five times as much attention.

Two questions decide it:

Is your person actually there? Not "does the platform have users like them". Do they open it, and are they in a frame of mind where your thing makes sense? Somebody scrolling before bed is not in the same mood as somebody on LinkedIn at eleven on a Tuesday.

Can you make what that channel wants? TikTok wants video. Instagram wants images. LinkedIn wants text and a point of view. Pinterest wants an image that promises something. If you cannot sustainably produce the native format, that channel is not available to you, whatever the audience data says.

Decision 4: A cadence you can hold in a bad week

Everyone sets a cadence based on their best week. Then a launch happens, or a holiday, or a flu, and the whole thing stops. The gap becomes evidence that social media does not work for you.

Set the number you could hit during the worst week of the last quarter. Twice a week that actually happens beats five times a week for a month and then nothing.

You can always add. Nobody has ever been punished by an audience for posting more than they promised.

Decision 5: The one number

Pick the single number that would make you say "yes, this is working" if it moved, and be honest that it is probably not followers.

Followers are the number everyone picks because it is the easiest to see. It is also the one least connected to anything. A thousand followers who never click is a worse outcome than two hundred who reply.

Better candidates, depending on decision two: replies, saves, link clicks, sign ups attributed to the channel, or simply the number of conversations that started with "I saw your post".

Then check it monthly, not daily. Daily numbers are noise, and reacting to noise is how a strategy turns into a series of panics.

Writing it down

One page. Five headings. Something like:

• Audience: founders who run their own social and have no time for it.
• Action: reply to the post or sign up for the free plan.
• Channels: LinkedIn primary, X secondary.
• Cadence: three posts a week on LinkedIn, whatever happens.
• Number: replies per post, checked monthly.

That page is more useful than any deck, because you can hold it in your head while you write, which is the only place a strategy ever actually gets applied.

What comes next

The strategy is the frame. The content is the work. How to build a social media content strategy is the next layer down: pillars, formats and how the week actually gets filled.

FAQs

How long should a social media strategy be?

One page. Anything longer is a document you write once and never consult, which means it has no effect on what gets published.

How many channels should I start with?

One. Two if the second one costs you almost nothing extra, for example a text post that works on both LinkedIn and X.

How often should I revisit it?

Quarterly. Monthly is too often to see anything real; annually is too slow to correct a bad channel choice.

What if I do not know who my audience is yet?

Then that is the strategy for the first quarter: publish for the narrowest guess you have, watch who replies, and rewrite the audience line based on the people who actually turned up.

Should followers ever be the goal?

Only if you have a specific reason to need a large list, for example because you launch to it. Otherwise pick a number closer to the action you named.