A posting calendar with a goal written at the top is what most B2B companies mean by a social media strategy. Ask what the posting is supposed to do for the business and the answer is usually visibility, which is a word that cannot be wrong and almost never gets measured. I run a small program of my own, so this is less a lecture from the stands than what I would tell an owner who asked me across a table.
A B2B social media strategy is one sentence that says what social will do for a specific business outcome, plus the evidence you will accept that it is working. Most of what follows, platforms, cadence, content types, is a consequence of that sentence. If a post cannot be traced to it, it is activity, not strategy.
Writing that sentence is harder than it sounds, because it forces three decisions most companies have been avoiding, and then a fourth that almost everyone skips. One thing first, though, because it changes what evidence you should expect. The B2B buyer in 2026 rarely sees a post and fills in a form. They see a video on TikTok, they come across your paid ad on Google, they catch a text post on LinkedIn, and then one day they search your name and land on your homepage. The attribution will not tie back neatly. Social is usually one of several touches that reinforce the same decision, and increasingly it is part of the material an AI assistant may draw on when someone asks it who to call. I would be careful about tying it straight to leads and sales, even though that is the ultimate goal.
What a B2B social media strategy actually is
The sentence is built from three lines, written in this order.
The business outcome is the thing the company needs, in the company’s own words. More qualified pipeline this year. A premium position in a category that is commoditising. Being the name a private equity buyer already knows when the banker’s deck lands. If you cannot name this, stop here, because nothing downstream can be judged.
The marketing objective is what marketing specifically contributes to that outcome. A number of sales-qualified conversations. A measurable change in who recognises the company and for what. This is the line most social plans skip, and it is why social so often floats free of the business.
Social’s job is one sentence that says what social will do to move the marketing objective, and it should be a job no cheaper channel is already doing well. That last clause is there to stop social being handed work that paid search or outbound already does better. Not “build awareness.” Something closer to “put the founder’s thinking about the buyer’s actual problems in front of operations leaders at mid-sized manufacturers, in the places they go to compare vendors.”
Then the fourth line, in two parts. The evidence you will accept, and the result that would justify continuing. For most B2B companies the honest first evidence is not a lead. It is a prospect saying “I saw your video” on a sales call. That proves exposure and recall, not value, so write down the second part too: what the list of mentions has to look like by the review date for you to keep going. Who is on it, how many, and whether those calls went anywhere.
| A founder-led services firm | A specialty manufacturer | |
|---|---|---|
| Business outcome | Win larger retainers from PE-backed companies | Shorten the sales cycle on a technical product |
| Marketing objective | Be known to portfolio operators before the RFP | Engineers arrive at the demo already convinced the spec is real |
| Social’s job | The founder’s judgment on the problems those operators have, where they look | Show the product being tested, in the formats engineers actually watch |
| Evidence we will accept | Operators mentioning the content unprompted on calls | Demo requests that reference a specific test video |
| What justifies continuing | Mentions from operators at target companies, on calls that progress | Demo-to-proposal rate improves on referenced demos |
| Read date | Weekly for mentions, quarterly for pipeline | Monthly against demo notes |
Fill in your own column and much of the rest of this article becomes a set of consequences. Not all of it. The same outcome can support more than one audience, more than one proof type, and very different spending levels, and the sentence does not settle those on its own. What it settles is what social is for, which is the thing most plans never say.
Mine, for the record. The outcome was a proof problem: I had built a business on referrals and was unknown outside my own network, at premium rates. The marketing objective was credibility with owners before I asked anyone for anything. Social’s job became the owner saying what he actually thinks about the viewer’s problems, on camera, in the places buyers look. The evidence I agreed to accept was buyers bringing it up without being asked. And the result that justifies continuing: by the end of the year, if the people mentioning the videos are not the owners and operators I sell to, the sentence was wrong and I rewrite it.
Who should be the face
When I ask a B2B company who will be the face of their LinkedIn presence, the most common answer is some version of “anyone is available.” The CEO, the CFO, the head of sales, whoever. That is the wrong answer, and not because any of those people is wrong. The face is a consequence of the sentence, and “anyone” means the sentence has not been written.
There are three honest options, and the buyer picks, not the agency. The founder, when the outcome depends on the market knowing how this specific person thinks, which is most professional services and anything else where the client is buying judgment. A subject expert who is not the founder, when the buyer’s confidence comes from depth the founder no longer personally carries. Or the work itself, when the proof is the product. A specialty materials manufacturer once stopped me mid-pitch to ask whether the videos had to be a person sitting down and talking to a camera, or whether they could be, say, a tensile strength test on one of their products. He was right and I was pitching the format I happen to like. For a company like his, the face is the test.
If the sentence lands on the founder, which it does more often than founders would like, a word on how to actually do it, because “be authentic” is not a method. I identify as shy and introverted and I have been in sales or leadership for about two decades, so I know what the camera does. It removes the listener, not the ability. You can talk fluently for an hour on a sales call and freeze in front of a lens, and the difference is that nobody is there. So I do not script anything. I use a one-page card: a hook memorised word for word, three cue lines that are things I can picture rather than topics, and a last line written out but not memorised, because the ending is where everyone hedges. Before I roll, I name one real person I am talking to. The pauses and false starts come out in the edit, and nobody ever sees them.
Keep in mind, though, that I am an unusual case and I know it. I self-produce because this is my trade and I am curious about the tools. Most owners should not run it this way. The version that works for most B2B companies is an owner who gives an hour on camera every week or two, plus a review pass on what goes out under their name, and a team, in-house or not, that does everything else. The hour and the judgment are the parts that cannot be delegated.
Which platforms, and the test for choosing them
Pick platforms by format strength and by where your buyer’s questions get typed, not by demographics. “Our audience is on LinkedIn” is true for every B2B company, which is exactly why LinkedIn is also where all of your competitors are.
What most B2B platform advice misses is that TikTok and YouTube behave like search engines that happen to show video. A company with almost no following can be found there by the exact question a buyer typed, where LinkedIn reach is mostly a function of the feed and who already follows you. I have written separately about how people actually find businesses now, and the short version is that discovery has moved to wherever a question gets answered, and short video is one of those places.
An earlier version of this very article told B2B consulting firms to skip TikTok and Instagram, and said TikTok required daily posting. My own numbers made me take that back. In the first week of my program, six videos posted once a day, TikTok delivered roughly twice the impressions LinkedIn did and about two and a half times the views Instagram did. And when I checked my account in mid-September, before the daily program began, about nine in ten of the views my TikTok videos had ever received had come from search, almost none from the For You feed. Nobody was scrolling into me. They were looking something up, and I happened to have said the words.
Two cautions on those numbers, because they are distribution numbers. They say how many people a platform showed the video to, and nothing about who those people were. The prospects who mention my videos on calls are the only evidence I have that the right people are among them, and which platform they saw me on is a question I should have been asking from the start. The log now has a column for it.
One finding went the other way. On LinkedIn, a text post out-reached the video version of the same idea by about two and a half to one, once. I kept LinkedIn on video, and I want to be clear about why, because it looks like a strategy protected from evidence. Video is inside my sentence. The proof an owner needs from me is watching me think about their problem, and text does not carry that. One pair of posts is not grounds to change the sentence. If the gap holds across three batches, the sentence gets rewritten, not defended. That is the test. When a platform’s data argues with your strategy, the sentence decides, and the sentence is allowed to lose.
So for most B2B companies the order is LinkedIn because the buyers are there, YouTube because its videos surface in ordinary Google searches in a way LinkedIn posts rarely do, and TikTok or Instagram when the format suits the buyer and someone can keep the cadence.
What to post, by what the buyer is doing
Every post has a job at a stage of the buyer’s journey, and the stages for B2B are fewer than the frameworks suggest.
| What the buyer is doing | The post’s job | Shaped as the question they would type |
|---|---|---|
| Learning they have a problem | Name it before they can | “Why isn’t our content bringing in leads anymore?” |
| Comparing approaches | Hand them the lens that favours good work | “Why does one company quote $5K for a website and another $50K?” |
| Deciding | Remove the last objection | “What happens in the first ninety days if we hire you?” |
| Staying | Make the customer look smart | “What should we tell our board about this?” |
Three weeks of my own data is not enough to hand anyone rules, so take what follows as the things I would test first, with the numbers that made me want to test them.
Lead with a number when you have an honest one. Across five videos in one week, the one that opened on a dollar figure did roughly three times the reach of the next-best video on every platform, and it drew the only saves of the batch. It also had a different topic and a different hook, so I cannot isolate the number as the cause. It is the first thing I would try to repeat.
Watch saves and comments before likes. People bookmark a reference and rarely bookmark an opinion, and on LinkedIn the comments are the distribution, so a post that invites a real argument travels further than one that invites agreement.
Concrete held better than opinion. On the videos that showed something specific, about half of viewers stayed past the first three seconds; on the ones that offered a view about AI, about a third did. Six videos, so a lean, not a law.
Talk about the viewer’s problem in the first person, not about your company. These sound like the same thing and they are not. My first-person accounts of things I had actually done for an owner reached roughly thirty times what my generic tips did on LinkedIn over the past year. My videos about my own company, about clients I let go and projects that went sideways, did worst of everything. The difference is whose problem the story is about.
A topic is not a title. There is an agency owner out there with exactly my buyer who posts every day and gets double-digit views, because his titles are subject headings like “weekly marketing meetings to track metrics.” The same material titled as the owner’s question is worth testing. And on the platforms that transcribe audio, the spoken words are part of what gets indexed, so I say the buyer’s question out loud early in the video rather than leaving it to an ending that might get cut, and I check the transcript afterwards to make sure it got said.
How to measure it when attribution is broken
Start with the honest fact about B2B. In a long sales cycle, the evidence that social is working shows up in sales conversations before it shows up in any dashboard.
Since mid-September I have had a lot of prospects mention on sales calls that they had watched my videos. Most of them never liked, commented, or interacted with anything. They watched, and then they said so when we talked. My follower counts are small and my engagement numbers are small, and the videos are reaching the right people anyway, because each one is written to a very specific person with a very specific problem. That is exposure and recall, which is the first rung, and it is why the first thing worth logging is a mention, not a click. Keep a list of who said it, which video, which platform, and whether the call went anywhere. That list is the evidence for the sentence long before a CRM can be.
Then the platform numbers, in the order they actually tell you something. Watch-through first, because it is the closest thing a platform gives you to whether the idea held. Saves and shares next, because they are a decision. Engagement as a rate against impressions rather than followers, since impressions are the people who were actually shown it. Comment text, because that is where you learn what the buyer calls the problem. Views last, because views are the weakest thing a platform collects and the first thing it shows you. None of these tells you whether the viewer was a buyer, which is why the mention list comes first.
Read on a cadence, then stop reading. I look at 24 hours, 7 days and 30 days; TikTok freezes after about two days and LinkedIn keeps accruing to day seven. And I do not call anything a pattern until it has shown up in three separate batches of videos, because two is a coincidence with a story attached.
One warning nobody gives owners. A flat number can be the platform, not the post. I had a YouTube Short with the best retention on my channel, meaning the people who saw it watched almost all of it, and it got ten views, because YouTube never put it in the feed. That does not prove it would have done well with exposure. It does mean the content was never tested, and that is a different conclusion from “it failed.” Check where the traffic came from before you judge what you made.
Then the follower chart, which is the one thing everyone looks at and the one I would read last. About one in five of my LinkedIn followers has a buyer-side title. The rest are largely other marketers and freelancers, which is normal for anyone who posts about marketing. And my follower count has risen by roughly the same amount every month since March, in months I posted once and in the month I posted every weekday. A number that moves without you is weak evidence of anything, and the follower count is the purest vanity metric there is.
The CRM deserves a word too, because I run a marketing agency and ours cannot prove social works either. Most of the deals in our pipeline carry no source at all, and the few ever marked as social came in before any of this measurement existed. When nothing is tagged, “social produces nothing” and “social is not tracked” look identical. There is no single fix. Tagged links catch the minority who click. The mention list catches the silent watchers, but only if whoever takes the call writes it on the opportunity record, not in a notebook. A “how did you hear about us” form field asks the buyer to do your attribution for you, and your salespeople already ask that question. The gap is almost never that nobody asked. It is that the answer never made it onto the deal.
What it costs
A sales-and-marketing team of one, asked why the company’s social had gone quiet, once gave me a one-word answer. “Time.” That is the real cost and it is worth pricing honestly.
In my program it runs like this. About an hour a week on a walk, being interviewed by a voice chat that pulls the ideas out of me and turns them into briefs. About an hour to shoot, which produces five to seven videos. Another hour or two editing with Claude, which does the cut, the captions and the weekly reads, and which does not pick the topic, write the sentence, or decide whether a card is true. Then about ten minutes a day to post natively to each platform, by hand, because I want to see what goes out. Call it four to five hours of my own time a week for a video every weekday.
The delegated model does not make that work disappear. It moves it. Someone still has to get the owner’s thinking out of the owner, which is an interview and a brief, not a request for “some content ideas.” Someone shoots or captures, edits, writes the captions, posts, answers the comments, keeps the mention list, and reads the numbers against the sentence. The owner gives the hour on camera and a review pass before anything goes out under their name. Add it up and it is a part-time job for somebody, whoever employs them, and it competes directly with the same person’s time on email, paid search, and the website. If that trade is not obviously worth it for your sentence, that is useful information. If you want a number for the agency version of that job, our social media management pricing is public.
The first version of my own program, a year earlier, was scripted and generic by design, and it reached roughly a thirtieth of what the first-person work has. That lesson cost me a year, and it is the cheapest thing in this article.
The posting rules you have heard of, and where they fit
If you have searched this topic you have met the rules of thumb. The 5-3-2 rule says that out of every ten posts, five should be content from others, three your own, and two personal. The 5-5-5 rule is either five posts each of educational, shared and promotional content, or a daily habit of five likes, five comments and five follows. The rule of 7 says a buyer needs about seven exposures before acting, and its origins are disputed.
They are cadence and mix rules, which is the layer underneath the sentence. None of them is a strategy, and following one before you have written the sentence is a disciplined way of staying busy. Once the sentence exists, pick whichever cadence you can actually keep.
When the sentence says no
A strategy that can only say yes is a sales pitch, so this is the no.
Social is a poor investment when your buyers are few and known by name, because direct outreach reaches them sooner and cheaper. When the sale is fully referral-gated and the buyer never researches in public. When nobody in the company can be the face and the work itself does not film. When the same hours in paid search or outbound would return inside a quarter and the business needs the quarter. And when compliance strips out the specificity that makes the content credible in the first place. One of our own account managers, asked by a professional-services client what else we could do for them, told them plainly that he would not recommend social, and we did not sell it.
The opposite error is just as common. A founder once told me on a pricing call to “just go get a TikTok specialist, you’ll probably blow up lead gen just by being on TikTok.” Platform first, outcome never named. That is the posting calendar with a goal on top, and it is where this article started.
Invest now, fix the foundations first, or hand it to someone
My own program is three weeks old as I write this, fifteen videos posted on every weekday but one, and I will not be able to read my follower trends honestly until mid-October. The mention list is the one signal I trust so far, the review date is on the calendar, and the honest weak spots in the content are evidence and specificity, which is what I am working on. That is enough to keep going until the date.
So, the decision, in the shape I use for every marketing bet. Set a date, let it play out, then keep going, pivot, or stop.
Invest now if the sentence names a job no cheaper channel is already doing, someone in the company can give the hour, and you have written down what evidence counts, what result justifies continuing, and when you will look.
Fix the foundations first if you cannot say plainly why you win, if the problem might be the brand rather than the marketing, or if the website leaks the confidence social sends to it. Social amplifies whatever is already true, which is no help at all if the true thing is a weak position.
Hand it to someone if the sentence is sound and the team is not there. That is the shape of an outsourced marketing department done well, where the owner keeps the hour and the sentence and the team carries the rest, or of social media management on its own if social is the only piece you need carried. If you want a straight read on which of the three you are, talk to us and we will tell you, including when the answer is to wait.
