Social media strategy for B2B in 7 steps

A social media strategy for B2B, when you are one to five people, is not a content plan. It is seven decisions, made once, that turn each week's client work into the proof a buyer looks for when they check your name. Here they are in order, with the choice at each step and the mistake that undoes it.

Step 1: Who checks your name, and when?

A B2B buyer rarely finds a supplier by scrolling. They get a name from a colleague, a directory, a quote request or a trade show. Then they check it. That check is the moment your feed does its work, and an empty profile answers in your place.

The decision: write for the person who already has your name and is deciding whether to email you. Not for a stranger an algorithm might show you to. A post for a stranger has to stop a scroll. A post for a verifier has to answer four questions in under a minute: are you active, do you know my industry, is the work real, will you be easy to deal with.

The mistake: building the strategy around reach. Reach counts strangers. Your verifier is one person, already on your profile, reading downward.

Step 2: Where does that person check it?

Pick one primary network: the one your verifier has open during work hours. For services sold to other businesses, such as consulting, bookkeeping, IT support, training or translation, that is LinkedIn, and it rewards specific craft over corporate polish. For B2B work judged by eye, such as a photographer shooting for brands, a designer, or a baker supplying cafés, it is Instagram: the buyer wants to see the work before the words. Facebook holds local trade, checked in a town's groups. X serves a narrow slice: developers, tech, media. Pinterest rarely hosts a B2B check; it hosts sourcing, which is a different strategy.

The other four networks still receive each post, adapted to their format. Done by hand, that adaptation costs 30 minutes per platform. Automated, it is the cheapest part of the week. The strategy decides where the verifier lands, not where the post exists.

The mistake: choosing the network you enjoy. The verifier's habit decides, not yours.

Step 3: What does the check need to find?

Four proofs, in the order a verifier reads them:

1. Active this week. A verifier reads the date of your last post before they read the post. Keep it under 7 days, always.
2. Knows my industry. One post in the last five that uses a word only someone in the buyer's trade would use.
3. The work is real. A specific job with a number, a constraint, a before and after. Anonymized is fine. Vague is not.
4. Easy to deal with. How you explain a choice to a client, how you say no, how you handle a change of scope.

Each proof maps to a post type, and the four rotate. The mistake: producing only proof 3, the portfolio, and none of proof 4. A buyer who already has your name assumes you can do the work. What they are checking is what working with you will be like.

“

A verifier reads the date of your last post before they read the post.

”

Step 4: Which unit of work becomes a post?

Not the deliverable. The decision. Every client job contains three to five decisions you explained to someone: why this option, why not the cheaper one, what the constraint was. Each decision is one post, and it carries the industry vocabulary with it, because you made it inside the client's industry.

This is also the mechanism behind generic B2B content. Give a text generator "IT support for small businesses" and it returns the average of every IT support page on the web: tailored solutions, peace of mind, proactive monitoring. Give it one decision from a real job, "moved a 12-person accountancy off a shared mailbox to per-user accounts because their insurer's policy required it", and the output cannot be mistaken for a competitor's. The fact drags its vocabulary along: policy, per-user, accountancy. You can test this in 30 seconds with any assistant, both inputs side by side. Once the decision is on the page, LinkedIn posts written in that vocabulary are the easy part.

The mistake: writing about your service category instead of the decision. "We help SMEs with cybersecurity" is proof of nothing.

Step 5: Does the post survive being forwarded?

In B2B, the person who reads your post is often not the person who signs. Your post gets screenshotted and dropped into a chat with four words: this one looks good. The screenshot loses your profile header. A post that only makes sense with your bio next to it dies at that moment.

The test: screenshot your last post and read it as someone who has never heard of you. Does it say what you do, for whom, and what the point is? If not, the first line has to carry it. "For a 12-person accountancy this month, we moved…" does it in one line.

The mistake: relying on the profile to carry the context. The profile is never forwarded.

Step 6: What is the cadence floor?

Two posts a week on the primary network, so the last-post date is never older than 7 days. That is the floor, not the target. Below it, proof 1 fails and the other three are never read.

A week where nothing was decided still has material: a question a prospect asked, a quote you declined, a change to how you invoice. Any of those is a post. The floor only holds if the week's posts are produced in one sitting, from what already exists, inside the weekly loop that runs social media at your size, rather than written one at a time once the day's client work is done.

The mistake: publishing a month of posts about your service in one burst, then nothing for five weeks. The gap is the failure, not the volume.

Step 7: Which number tells you it works?

Not reach and not likes. Named inbound: the number of new contacts in a month who mention something you posted. Ask every new contact one question, "did you see anything of ours before writing?", and log the answer. After 8 weeks that log says whether the feed is doing its verification job.

The analytics tab measures strangers. For a B2B check, the verifier is one person, and one person never shows as a spike.

The mistake: judging the strategy on week two. The verifier arrives when the referral happens, and referrals do not follow your posting schedule.

What you have after eight weeks

A primary network with a post every three to four days. Each one built on a decision from a real job, readable without your profile, in your buyers' vocabulary. A verifier arriving from a referral finds all four proofs in under a minute.

The seven decisions take one afternoon. What remains is the production of two posts a week, and that is the only place the strategy can still fail: not in the plan, but in whether those two posts get written by you after the client work, or produced from the client work while it is still fresh.