LinkedIn Post Ideas for Inbound Marketers

Our blog traffic fell 50% year over year. Our demo requests didn't move at all. The drop started right when AI Overviews rolled out more broadly in our category. Informational keywords — "what is," "how to," definitional content — got hit hardest, down closer to 65%. Bottom-funnel comparison and use-case pages barely moved, down maybe 8%. Here's what that split told us: roughly half our historical traffic was empty-calorie visits — people getting their question answered in the search snippet or the AI summary and never clicking through at all. We'd been reporting that traffic as a content win for three years. It was never pipeline. It was search engines using our content as raw material and giving the answer away before anyone saw our brand. What we changed: we stopped measuring content success by sessions and started measuring it by assisted pipeline touches and branded search lift, tracked monthly. We also shifted roughly 30% of our content budget away from top-of-funnel definitional posts — the kind most likely to get fully summarized — toward comparison pages, original research, and customer-story content that AI Overviews can't fully answer without sending someone to the source. Six months into the new mix, total sessions are still down from peak, but demo requests from organic are up 11%, and our content team stopped panicking every time a traffic graph dipped. If your traffic dropped and your pipeline didn't, you already have your answer about which content was ever doing real work.

The pillar-and-cluster model built most of our decade's inbound strategy. I think it's aging badly, and I'll show you the alternative we're actually running now. The old model: one pillar page, ten to fifteen cluster posts linking up to it, built to signal topical authority to Google's crawler. It worked when ranking was mostly about comprehensive coverage and internal link equity. The problem now: AI answer engines don't crawl your cluster architecture to understand your authority — they synthesize from whichever individual page answers the question best, regardless of what it links to. A brilliant, well-linked cluster of mediocre pages loses to one excellent standalone page from a competitor with no cluster structure at all. What replaces it, in our content plan: fewer pillar-and-cluster builds, more "answer density" pages — single pages engineered to be the single best, most citable answer to one specific question, with original data or a genuinely distinct point of view baked in, rather than comprehensive coverage of ten related subtopics. Mapped example: instead of a 12-post cluster around "content marketing strategy," we built one page around "how much should a B2B company spend on content marketing" with our own benchmark data from 40 customer accounts. It now gets cited by three AI tools we've checked and outranks cluster-built competitor pages that have 15x the page count on the topic. The cluster model isn't dead everywhere — for genuinely broad topics it still helps navigation. But as a ranking strategy, depth-per-page is beating breadth-of-cluster in what we're seeing. What's replaced your old cluster strategy, if anything has?

One 45-minute customer interview. Eight pieces of content. Here's the exact path, with time estimates. Step 1 (0 min, happens during the call): Record and transcribe via Otter. Zero extra time — this is just how we run every interview now. Step 2 (30 min): Full case study, written from the transcript, structured around the specific outcome — the number they gave us, the timeline, the "before" pain point in their words. Step 3 (10 min): Three LinkedIn posts pulled directly from quotable moments in the transcript, lightly edited for readability, always credited. Step 4 (15 min): One FAQ entry added to our product page, phrased exactly as the customer asked their original question — real buyer language beats our internal phrasing every time. Step 5 (20 min): A sales enablement one-pager, the objection they raised early in the call and the specific thing that changed their mind, handed straight to the sales team as a talk track. Step 6 (15 min): One data point extracted for our quarterly benchmark report, anonymized and aggregated with other interviews. Step 7 (10 min): A short video clip, if they were on camera, cut to under 90 seconds for the website testimonials page. Step 8 (10 min): One cold outreach line for our sales team, referencing the use case for similar-profile prospects. Total: about 110 minutes of repurposing work for one interview, spread across a week, not done in one sitting. We run six of these a quarter now. It's become our highest-leverage content workflow by a wide margin.

I audited 120 of our oldest blog posts. We deleted 60 of them. Total organic traffic rose 17% over the following quarter. The decision criteria, applied post by post: — Zero organic clicks in the last 12 months AND no backlinks pointing to it: delete, no exceptions. 34 posts fell here. — Covering a topic that a newer, better post on our own site already ranked for: consolidate, redirect to the stronger page. 19 posts. — Thin, under 500 words, published purely to hit a publishing cadence quota years ago, no real answer to any actual query: delete. 7 posts. What we kept even with low traffic: anything ranking for a genuinely valuable long-tail term regardless of volume, and anything still earning referring domains even years later. Why traffic went up after deleting content, which feels backwards: Google was spending crawl budget indexing 120 pages, a third of which were actively hurting our average content quality signal for the domain. Removing the weak third let the remaining, stronger 60 posts consolidate ranking signal instead of competing against our own thin content for the same queries. The redirect map took the longest part of the project — three days of matching old URLs to the closest live equivalent, checking each one manually rather than trusting a bulk redirect rule. Content pruning still feels heretical to defend internally — "we're deleting work we paid for" is a hard sentence to say in a planning meeting. But the graph doesn't lie, and I have it ready for the next time someone questions the plan.

A prospect quoted a line from our ebook, verbatim, on their very first sales call. That ebook was published two years and four months earlier. Our AE didn't recognize the reference at first — had to ask which piece they meant. The prospect had downloaded it 14 months before ever filling out a demo form, according to our CRM timestamp. Fourteen months of silence, then a call, already sold on our point of view before the AE said a word. This is the part of inbound that a 90-day attribution window will never show you. If you're judging content ROI by whether it converted within a quarter, that ebook looked like a dead asset for over a year. By any reasonable last-touch or even multi-touch model with a standard lookback window, it contributed exactly zero to that deal. It contributed everything to that deal. What this changed in how we plan content: we stopped killing pieces just because they weren't showing recent form-fill activity, and started tracking a longer-tail metric instead — content mentioned unprompted on sales calls, logged manually by AEs in a simple CRM field. It's messy data, self-reported, imperfect. It's also the only signal that would have caught this ebook's real value at all. The dark funnel is real, and it's longer than any dashboard we have accounts for. Some of your best-performing content from three years ago is quietly working right now, invisible to every report you're looking at. What's the oldest piece of content you've had referenced back to you unprompted?

Three lead magnets we ran that filled our CRM with junk, and what they had in common. 1. The "ultimate guide to [broad topic]" PDF. Downloaded by students, job seekers researching the field, and competitors doing content research, far more than actual buyers. It converted at a rate that looked great in a monthly report and terrible in a pipeline report — less than 2% of those leads ever became a qualified opportunity. 2. The generic template pack. High download volume, almost no correlation with buying intent, because a free template is useful to anyone in the role regardless of whether their company can afford or needs our product. 3. The industry benchmark report gated behind a full form — name, company, job title, phone number, company size. We over-gated a piece that should have earned trust, and lost roughly 60% of would-be downloads at the phone number field alone, according to our form analytics, while the leads who did push through were often just curious, not evaluating anything. What replaced all three: lead magnets scoped narrowly enough that only someone actually evaluating our category would want them — a competitor comparison worksheet, a specific ROI calculator tied to our own product's mechanics, an implementation checklist assuming you'd already bought something like us. Narrower magnets, smaller download numbers, dramatically higher opportunity conversion rate — we went from roughly 2% lead-to-opportunity on the broad guide to 14% on the narrow ROI calculator. The lesson: a lead magnet's job isn't maximum downloads. It's self-selecting for buyers, even if that means a smaller, uglier-looking top-of-funnel number.

Someone told me they'd read our entire pricing comparison inside a ChatGPT conversation and never once visited our site. That was the moment our content strategy actually changed, not just our talking points about it. What we changed: — More original data. AI assistants synthesize existing content well but can't invent our own numbers. Every new pillar piece now includes at least one proprietary stat pulled from our own customer base or product usage data, something no summary can replicate without citing us directly. — Quotable, self-contained stats. We rewrote key sentences so they read as complete, citable claims on their own — "companies using X see a 23% reduction in Y" instead of a sentence that only makes sense in the paragraph around it. Assistants pull sentences out of context; we now write for that extraction. — Structured, direct answers near the top. Less scene-setting before the point, more front-loaded direct answers, because that's the section most likely to get lifted whole into a generated summary. — Author and source attribution built into the content itself, not just the byline. If an assistant is going to summarize us, we want our name attached to the claim inside the summary, not stripped out. We can't fully measure the payoff yet — there's no clean "cited in ChatGPT" report in GA4. But branded search volume is up 22% over the last two quarters, and I don't have a better explanation than: people are encountering us inside AI answers and searching our name afterward. How are you adjusting for a reader who may never actually land on your site?

Four people. Twenty content assets shipped a month. Here's the actual system, not the idealized version. Our kanban has five columns: Idea, Briefed, In Progress, Review, Published. Nothing skips a column, even when we're behind — that discipline is the whole reason this works at our size. Monday standup, 20 minutes: each person states what's moving from one column to the next this week, not a status update on everything they're touching. If something's been stuck in "In Progress" for more than two weeks, it gets discussed live, not just noted. Definition of done, the same five items on every single piece: SEO brief followed, internal links added, one original data point or quote included, CTA matched to funnel stage, and a Slack post drafted for distribution. Miss any one, it doesn't move to Published, full stop. The quality gate that slows us down on purpose: every piece gets one "does this actually help someone" read before publishing, done by whoever didn't write it. It adds a day to our timeline. It's also caught at least three genuinely thin drafts this year that would have hurt our overall site quality signal. What makes 20 assets a month possible with four people: about 60% of that count isn't new writing — it's repurposed customer interviews, refreshed old posts, and format variations of existing pillar pieces. Pure net-new long-form content is closer to six pieces a month. The rest is the system working the assets we already have harder. The standup is boring. The kanban is unglamorous. Twenty assets a month is the result of both, every week, without exception.

Seven content formats, ranked by pipeline influence, pulled straight from our CRM's content-touch field over the last four quarters. 1. Comparison pages ("us vs. [competitor]") — highest opportunity-influence rate of anything we publish, by a wide margin. Bottom-funnel intent, no ambiguity about why someone's reading it. 2. Customer case studies — second highest, especially when the featured customer matches the prospect's industry or size. Sales reps request these by name more than any other asset. 3. Original research and benchmark reports — strong influence, and the longest half-life of anything on this list; a report from 18 months ago is still getting referenced in deals today. 4. Webinars, specifically the recorded, gated replay — surprisingly strong, particularly for larger deal sizes where multiple stakeholders watch it independently before a group call. 5. Thought leadership / opinion posts — moderate direct pipeline influence but strong correlation with branded search lift, which shows up in pipeline a step removed. 6. How-to / educational blog posts — high traffic, genuinely low direct pipeline influence in our data. Useful for top-of-funnel awareness, rarely the thing a rep points to in a deal. 7. Definitional / glossary-style content — lowest pipeline influence of the seven, despite often being high-traffic. This was the surprise that stung: our highest-traffic content category is our least commercially useful one. The practical takeaway we took from this: we shifted next quarter's calendar to roughly 40% comparison and case-study content, down from about 15%, and cut glossary content by half. Where would your CRM data put webinars, if you're tracking this?

Would you trade half your blog traffic for double your newsletter reply rate? I'm asking because our team is actually facing this tradeoff right now, not as a hypothetical. Our top-of-funnel blog content drives sessions Google rewards. Our newsletter, sent to about 9,000 people, gets maybe 40 replies a month — but those replies convert to sales conversations at a rate our blog traffic has never come close to matching. The honest tension: growing blog traffic is measurable, reportable, and defensible in a monthly meeting. Growing newsletter reply rate feels smaller and slower, even though I increasingly believe it's the better long-term asset. A reply is someone choosing to talk to us. A session is someone the algorithm chose to show us to. If I had to pick: I'd take double the newsletter replies, without much hesitation. Owned-audience engagement doesn't evaporate with the next algorithm or AI Overview update — the list is ours, the relationship compounds, and a reply is about as close to sales-qualified intent as unpaid content gets. But I recognize that answer is easy for me to give because our blog was never our primary lead source to begin with — it's a company where relationships close deals, not search traffic volume. This question forces a real answer about what you actually believe converts, not what's easiest to report upward. Half your blog traffic for double your reply rate — which side are you on, and what does your business actually reward?