Most B2B accounts post about the company. The ones that work post about the decision the buyer is trying to make.

Open almost any B2B social account and the pattern is the same: a new hire, a stand at a show, an anniversary, a certificate. All true, all uninteresting to anyone outside the building. The account is a noticeboard for the company rather than a service to the reader, and it performs exactly as a noticeboard would.
The reason it fails is not the format. It is that the reader is not in buying mode when they see it. At any moment a very small share of your market is actively looking to buy; the rest are working. Content that only makes sense to someone already shopping reaches almost nobody, and content that assumes they care about your internal milestones reaches nobody at all.
The alternative is not to post more often. It is to post about the thing the reader is dealing with today, which is almost never you.
The specification detail. One dimension, tolerance, material or standard, explained properly. Why a particular grade fails at a particular temperature. What an IP rating actually covers. These posts are short, unglamorous and are saved and forwarded more than anything else you will publish, because they solve a problem someone has open in another window.
The failure mode. What goes wrong, why, and what it looks like before it goes wrong. Nobody else in your category will publish this, because it feels like admitting that things fail. It is the single most credible thing a technical supplier can say, and it positions you as the person who has seen the problem before.
The decision framework. How to choose between two approaches, with the trade-off stated honestly, including the case where your product is the wrong answer. A buyer who uses your framework to make a decision has adopted your way of thinking about the category, which is worth more than a click.
The work itself. Not a case study with a result attached, but the thing you made: a page, a drawing, a stand, a test rig, a film. Process is more interesting than outcome, and it is the only proof that survives being scrolled past at speed.
Social media is not a lead source for a considered purchase, and treating it as one is why most B2B programmes get cancelled. On a decision that runs eleven months and involves five people, the job of social is to be the reason your name is already familiar when the search starts.
That is a memory job, not a conversion job. It works by accumulation: the same reader seeing something useful from you four or five times over a year, in a context where they were not being sold to. When the requirement finally lands on their desk, you are the supplier they think of first and the one they trust enough to shortlist without a referral.
Judged as lead generation it will always look like a failure. Judged as the reason your name gets typed into a search bar, it is often the cheapest thing you do.
Impressions and follower count tell you almost nothing, because a large audience of the wrong people is worse than a small one of the right people. Both numbers are also the easiest to inflate and the hardest to connect to revenue.
Measure instead: saves and shares, which indicate the content was useful rather than merely seen; profile and website visits from companies on your target list; direct messages that start a technical conversation; and the share of inbound enquiries that mention having read something of yours. That last one is a single question added to your enquiry form and it will tell you more than any dashboard.
Set the baseline before you start, review it at six months rather than monthly, and accept that the curve is flat for a long time and then is not.
Two useful posts a week, sustained for a year, beats daily posting that stops in March. The constraint is rarely ideas and almost always approval, so agree the boundaries once — what can be said about products, what needs review, who signs off — and then let the person writing get on with it.
The material already exists. It is in the answers your application engineers give on the phone, the questions that come back on every quotation, and the drawings on the desk. A short weekly conversation with the technical team will produce a quarter of content in an hour.
Key takeaways
For most technical businesses, LinkedIn carries the buying committee and is also where people check you after meeting you. YouTube matters if your product is demonstrated rather than described. Instagram is worth it only when the work is visually strong. Pick one and do it properly rather than syndicating everywhere.
Twice a week, indefinitely, is a better target than daily for a month. Consistency matters more than volume because the mechanism is accumulated familiarity, not reach in any single week.
Both, but weight it towards people. Individual accounts consistently out-reach company pages, and a technical specialist explaining something in their own words is more credible than the same words under a logo. The company page is where someone checks you are real.