Social media marketing plan: the one-page version

If you are looking for a social media marketing plan, what you need is a marketing plan with social media on it. Social is one channel. A plan built around a single channel can tell you what to post, but it cannot tell you whether posting is the right use of the money, and it cannot tell you when to stop.

The plan itself fits on one page. Nine boxes, each one a short answer somebody in the business has to agree to.

1. The one outcome.
2. The buyer, and how they decide.
3. The position, in one sentence.
4. The proof that carries it.
5. Three channels, and no more.
6. What you stop.
7. The budget.
8. The monthly number.
9. The review date.

That is all of it. Writing the page takes an afternoon if the arguments are settled and three months if they are not, which is why most firms produce a forty-slide deck instead. The deck is easier than the argument. What follows is each box, with a bad answer beside the good one.

One outcome. Not three. The test is whether your finance director would recognise it as a business result rather than a marketing result.

A good answer names a quantity, a type of customer and a date. Twelve new clients in twelve months, from mid-sized contractors in the Eastern Province. Thirty qualified enquiries a quarter from firms with more than a hundred staff. Something you could be publicly wrong about.

A bad answer is any of these: increase brand awareness, grow our following, become the leading name in the region. Or three outcomes listed together because nobody wanted to lose theirs. Three outcomes is no outcome. When the budget tightens in month seven, a plan with three of them gives you no way to choose, so you cut whatever is hardest to defend, which is usually the thing that was working slowly.

Write the number down even if you are not confident in it. A number you revise in month four is worth more than an adjective you never test.

Two lines. Who buys, and what happens between the moment they notice a problem and the moment they sign.

A good answer is specific enough to pass the naming test in how companies actually buy, and it says who else is in the room. In most Gulf firms that means the owner who signs, the department head who lives with the result, and a junior person who does the research and sends the first message. It also records what nearly stopped the last three deals you won. Your sales team knows all of this. It has never been written down.

A bad answer is a market description: SMEs in the GCC, growing businesses, decision makers aged thirty to fifty. Those describe a population, not a buyer. The other bad answer is an invented persona with a stock photograph, a first name and a list of hobbies, which feels like research and contains none.

The position, in one sentence

What you do, who for, and why you rather than the other two names on the list. Twenty words is plenty.

A good answer would damage you if a competitor copied it, because it commits you to something. It usually contains a narrowing: a sector, a size of company, a type of problem, a method you genuinely use.

A bad answer could be signed by every firm in your category. Quality, innovation, customer focus, tailored solutions, results driven. Test it: put a competitor's name on your sentence and see whether anything looks wrong. If nothing does, the box is empty.

The proof that carries it

A claim with no evidence under it is a preference. List what you can actually show: clients who agreed to be named, a result with a date and a source beside it, research nobody else in your category has published, a client who will take a reference call.

A bad answer is a wall of logos with no story attached, and figures with no date, no period and no note saying who measured them. Research is the most under-used item here, because it is the one form of proof a buyer can read before contacting you. Ours sits on the white papers page, in public, because evidence a buyer cannot read before contacting you cannot do its job. That applies to proof. A tool or a template can sit behind an email address without the same cost, because nobody uses one of those to decide whether to trust you.

Three. The limit is the point of the box.

A good answer names three and gives a reason for each that refers to the buyer rather than to the channel. Search, because people type the problem before they type a company name. LinkedIn through named people, because the researcher checks who works there. Email to a list of companies you chose, because the universe is small enough to name.

A bad answer is a grid of eight platforms with a posting frequency beside each, built on the assumption that presence is free. It is not. Every channel costs attention, and attention is the resource that runs out first. How few, and which, is a decision the strategy behind this page makes: what a digital strategy should include sets out the order.

The other bad answer is a channel chosen because a competitor is on it. That tells you what they decided, not what their buyers did. Social takes more room in most plans than it earns, which is the argument in social media is not marketing.

The box nobody fills in, and the one that turns the page into a plan. A plan that only adds is a budget request.

A plan that only adds is a budget request.

A good answer names activity you are running now, gives a date it ends, and says who tells the person doing it. The monthly newsletter nobody opens. The second Instagram account. The blog, if the blog is not working.

We did this to ourselves. Around four hundred blog articles on this site produced thirty five clicks from people who were not already searching for our name, over seventeen months. Stopping was worth more than anything we could have added, because it returned the time to work with a buyer at the end of it.

A bad answer is an empty box, or the phrase we will review this later. Later has no date and no owner, so it never arrives.

The budget

A figure, and what it buys. A good answer starts from the value of one client rather than from a percentage of revenue copied out of an article. Take the margin on one client, decide what share of it you would pay to win one, then multiply by the number of clients in box one. That produces a budget for the year, not a price per lead. The price per lead is a different calculation from the same figures, set out in what marketing should cost.

A bad answer is whatever is left after everything else has been funded. That is how most marketing budgets are set, and why they are cut first.

The monthly number

One number, on the front of the report, owned by a named person. A good answer is a number the business already cares about: qualified enquiries that became clients, and what one of them cost. Everything else is diagnostic, useful when something breaks and out of place at the top of a page.

A bad answer is a dashboard. Twelve metrics reported every month is not measurement, it is cover. When everything is reported, nothing is accountable.

A date in the diary, a person, and three questions: what did the number do, what do we stop, what do we keep. Monthly suits a small firm. Quarterly suits the position and the channels. A good answer is a specific date with a name against it. A bad answer is the word quarterly with nothing after it, which means it happens in the week somebody remembers.

The discipline is holding the meeting in the month the news is bad, because that is the month it quietly moves. A review that only happens after a good quarter is a celebration. Two rules keep it honest. Book the next date at the end of the current meeting. And open with the number rather than with what was produced. If it cannot be held in half an hour, one of the eight boxes above is still unsettled.

That is the plan. Nine boxes, no appendix. Once the nine answers are agreed the execution is mostly obvious, which is the uncomfortable part: most of the complexity sold in this industry lives in the space where these answers should be. We keep the nine boxes on a single sheet in English and Arabic, and you can have it for an email address, though nothing on it is missing from the article above.