The starting point

When I joined Excelerate Consulting as their in-house Marketing Analyst in January 2025, their LinkedIn page had around 3,000 weekly impressions. Not bad for a boutique consulting firm — but not generating leads, not building a recognisable brand, and not compounding.

By October 2025, that number was 60,000+ weekly impressions. The page was generating inbound enquiries, founders were getting LinkedIn connection requests from warm prospects, and the brand had become recognisable in its niche. This is how we did it.

This is not a "post every day and good things happen" story. We posted 3–4 times per week at a consistent quality standard. Volume matters less than format discipline and narrative clarity.

What wasn't working

Before I arrived, the LinkedIn content was a mix of: company news nobody cared about, generic "5 tips for business success" posts, and the occasional case study written for the client's ego rather than the reader's problem. Classic B2B LinkedIn — present but invisible.

The core problem: the content was written about Excelerate, not for the people Excelerate wanted to reach. Nobody follows a LinkedIn page to hear about your culture. They follow because you consistently say something useful about problems they have.

The three content pillars

The first thing I built was a content pillar framework. Every post had to belong to one of three categories — no exceptions:

Pillar 1 — Problem + Insight

These are posts that name a problem the target client recognises, then offer one non-obvious insight about it. Format: 3–6 lines. No header image needed. High engagement because the hook is a problem, not a solution pitch.

Example hook: "Most B2B companies invest in LinkedIn but see flat results. Here's the one thing almost everyone gets wrong about how the algorithm distributes posts."

Pillar 2 — Behind-the-work

These are posts that share a specific, concrete thing we did — a tactic, a result, a mistake, a tool. Not general advice: this specific thing happened in this specific context. These are high-trust because they prove you actually do the work, not just write about it.

Example: "We ran an A/B test on our ebook landing page. Version A had a form above the fold. Version B had 200 words of copy before the form. Version B converted 34% better. Here's why we think that is."

Pillar 3 — Framework posts

These are posts that give the reader a mental model — a way of thinking about a problem that they'll use long after they've forgotten where they learned it. These posts tend to get shared and saved, which sends strong signals to the algorithm.

Example: "There are three levers in any content funnel. Most teams only pull one. Here's how to identify which lever you're ignoring."

Post formats that actually work

Format matters as much as content. Here's what worked for us in order of average reach:

1. The hook-driven text post (no image)

Counterintuitively, our best-performing posts were often plain text. LinkedIn's algorithm treats image and document posts differently. A short, punchy hook on a text post expands to "see more" faster and gets shown to a wider initial audience before LinkedIn decides whether to distribute it broadly.

Key: the first two lines must make you want to click "see more." I tested hundreds of hooks. The ones that worked best either named a specific problem, made a counterintuitive claim, or opened a narrative with an unresolved tension.

2. The document carousel (PDF slides)

These are posts where you upload a PDF and LinkedIn renders it as a swipeable slide deck. They have the highest dwell time of any format — people swipe through slides, which LinkedIn counts as engagement. For frameworks and step-by-step content, these are the highest-reach format when they work.

The mistake most teams make: 20-slide carousels. Keep it to 6–8 slides. If a slide takes more than 5 seconds to read, it's too long.

3. The short video (90 seconds max)

We used HeyGen-generated avatar videos for some of these. Not for everything — AI video has a visible quality ceiling and using it for authentic "behind the scenes" content felt wrong. But for explainer posts and framework walkthroughs, a clean 60–90 second video worked well.

Key: native upload always. YouTube links get suppressed by LinkedIn's algorithm. Upload the video file directly.

4. The poll

Polls generate comments more than any other format. Comments are the highest-weight engagement signal on LinkedIn. We ran one poll per fortnight on a question we genuinely wanted the answer to. The comments became content research for future posts.

The weekly cadence

Here is the exact posting schedule I ran:

That's 3–4 posts per week. We batched content creation on Monday mornings — 2 hours to draft the week's posts, review, schedule. Using Typefully for scheduling on the company page.

Results and what drove them

The 20× growth wasn't linear. The first three months were slow — impressions moved from ~3K to ~8K weekly. Then there was a step-change around month four when two posts went beyond our first-degree network. After that, growth compounded because the page had enough followers and engagement history for LinkedIn to show new posts to a wider cold audience by default.

What drove the step-change: a carousel post about B2B funnel mistakes that got shared by three people with 10K+ followers each. One post, one good share, and the page crossed a threshold.

The lesson: you can't predict which post will break through. You can control the frequency, quality, and consistency that makes it likely to happen — and that makes the benefit compound when it does.

What to steal from this

If you're running LinkedIn for a B2B business or a personal brand, the highest-leverage actions are:

LinkedIn rewards the person who keeps showing up with something useful. It doesn't reward the person who posts when they have something to sell. Separate the two and your growth compounds. Conflate them and you plateau.
Want the calendar version? Download the LinkedIn Content Calendar for the 30-day posting plan, hook bank, format matrix, metrics tracker, and repurposing checklist.