Dwell time is quietly outranking likes on LinkedIn
The question: when LinkedIn decides who sees your post, does a 'like' matter more than how long someone stops to read it?
A reverse-engineering analysis (a study that infers algorithm behavior from observed reach, not internal docs) tracked roughly 3,000 company-page posts and modeled reach against engagement signals. The method: regress final impressions on each signal while holding follower count constant. A correlation (two things moving together) is not causation (one causing the other) — but the pattern was consistent.
Three findings:
— Dwell time (seconds a post stays in view before scroll) predicted reach more strongly than likes.
— A like in the first 60 minutes was worth several times a like at hour six — the 'golden hour' is real but narrow.
— Comments over five words outperformed one-word comments, suggesting the model weights effort, not volume.
Caveats: inferred, not confirmed; single quarter; B2B pages only — treat as directional.
What it means for B2B: write a strong first three lines so the reader pauses before the 'see more' fold. That pause is the signal you can actually influence.
Bottom line: optimize for the stop, not the tap.
The question: when LinkedIn decides who sees your post, does a 'like' matter more than how long someone stops to read it?
A reverse-engineering analysis (a study that infers algorithm behavior from observed reach, not internal docs) tracked roughly 3,000 company-page posts and modeled reach against engagement signals. The method: regress final impressions on each signal while holding follower count constant. A correlation (two things moving together) is not causation (one causing the other) — but the pattern was consistent.
Three findings:
— Dwell time (seconds a post stays in view before scroll) predicted reach more strongly than likes.
— A like in the first 60 minutes was worth several times a like at hour six — the 'golden hour' is real but narrow.
— Comments over five words outperformed one-word comments, suggesting the model weights effort, not volume.
Caveats: inferred, not confirmed; single quarter; B2B pages only — treat as directional.
What it means for B2B: write a strong first three lines so the reader pauses before the 'see more' fold. That pause is the signal you can actually influence.
Bottom line: optimize for the stop, not the tap.
Putting the link in the first comment: tested, mostly a myth
The question: does moving your link out of the post body and into the first comment actually rescue reach?
The folk theory says the algorithm suppresses posts with outbound links, so practitioners hide links in comments. An A/B-style test ran matched posts — same copy, link in body versus link in first comment — across a set of accounts. Method: paired posts, randomized which version each account ran.
Three findings:
— The reach difference between link-in-body and link-in-comment was small and inconsistent — not the dramatic penalty folklore claims.
— Link-in-comment posts got fewer total clicks, because many readers never expand comments.
— Where in-body links did lose reach, the loss was smaller than the click-through loss from hiding them.
Caveats: link suppression may vary by account history and topic; one test window — treat as directional.
What it means for B2B: you may be trading measurable clicks for an unmeasured, possibly tiny reach gain. If clicks are the goal, the 'hide the link' tactic can be net-negative.
Bottom line: test it on your own account before adopting a rule built on someone else's anecdote.
The question: does moving your link out of the post body and into the first comment actually rescue reach?
The folk theory says the algorithm suppresses posts with outbound links, so practitioners hide links in comments. An A/B-style test ran matched posts — same copy, link in body versus link in first comment — across a set of accounts. Method: paired posts, randomized which version each account ran.
Three findings:
— The reach difference between link-in-body and link-in-comment was small and inconsistent — not the dramatic penalty folklore claims.
— Link-in-comment posts got fewer total clicks, because many readers never expand comments.
— Where in-body links did lose reach, the loss was smaller than the click-through loss from hiding them.
Caveats: link suppression may vary by account history and topic; one test window — treat as directional.
What it means for B2B: you may be trading measurable clicks for an unmeasured, possibly tiny reach gain. If clicks are the goal, the 'hide the link' tactic can be net-negative.
Bottom line: test it on your own account before adopting a rule built on someone else's anecdote.
The 60/40 split that most B2B teams invert
The question: what share of B2B marketing effort should go to long-term brand building versus short-term demand capture?
Drawing on long-run effectiveness studies, researchers proposed a roughly 60/40 split (60% brand, 40% activation) for B2B, adjusted by category. Method: meta-analysis of campaign effectiveness over multi-year horizons — aggregate, not firm-specific.
Three findings:
— Brand-building effects compounded over years; activation effects spiked then decayed within weeks — different time signatures.
— Teams measured on quarterly leads systematically over-invested in activation, the opposite of the recommended split.
— The brand share that looked 'wasteful' in-quarter was what made later activation cheaper, by raising baseline familiarity.
Caveats: the optimal ratio varies by category, growth stage, and margin — treat as directional.
What it means for B2B: if your social output is mostly product, offers, and CTAs, you are running an activation-heavy mix that may be eroding the brand memory that makes activation work. The 'inefficient' brand content is the long game.
Bottom line: most B2B teams under-invest in brand because they are measured on the quarter, not the multi-year curve.
The question: what share of B2B marketing effort should go to long-term brand building versus short-term demand capture?
Drawing on long-run effectiveness studies, researchers proposed a roughly 60/40 split (60% brand, 40% activation) for B2B, adjusted by category. Method: meta-analysis of campaign effectiveness over multi-year horizons — aggregate, not firm-specific.
Three findings:
— Brand-building effects compounded over years; activation effects spiked then decayed within weeks — different time signatures.
— Teams measured on quarterly leads systematically over-invested in activation, the opposite of the recommended split.
— The brand share that looked 'wasteful' in-quarter was what made later activation cheaper, by raising baseline familiarity.
Caveats: the optimal ratio varies by category, growth stage, and margin — treat as directional.
What it means for B2B: if your social output is mostly product, offers, and CTAs, you are running an activation-heavy mix that may be eroding the brand memory that makes activation work. The 'inefficient' brand content is the long game.
Bottom line: most B2B teams under-invest in brand because they are measured on the quarter, not the multi-year curve.
An agency swapped single images for carousels. Dwell time tripled, demos followed.
The question: does longer dwell time (seconds a viewer spends before scrolling past) actually move pipeline, or just vanity stats?
The case: a B2B marketing agency rebuilt its LinkedIn output over one quarter — replacing single-image posts with 8-to-10-slide PDF carousels on the same topics. Sample: 40 single-image posts (prior quarter) vs 38 carousels (test quarter), same author, same cadence.
Three findings:
— Estimated dwell time rose from ~4s to ~13s per view (LinkedIn shows partial signals; they triangulated with slide-advance taps).
— Reach per post climbed 2.1x, consistent with dwell being an input the algorithm rewards.
— Most striking: inbound demo requests citing 'saw your LinkedIn' went from 2 to 9 across the quarter.
Limitations: the agency also tightened its topics that quarter, so dwell and relevance are 'confounded' (tangled causes) — correlation, not clean causation. n is small.
What it means for B2B: carousels buy attention seconds cheaply, and attention seconds appear upstream of both reach and intent.
Bottom line: format change plausibly drove pipeline, but a topic change rode along — read it as suggestive.
—
Продолжение про campaign objectives — @B2BLever
The question: does longer dwell time (seconds a viewer spends before scrolling past) actually move pipeline, or just vanity stats?
The case: a B2B marketing agency rebuilt its LinkedIn output over one quarter — replacing single-image posts with 8-to-10-slide PDF carousels on the same topics. Sample: 40 single-image posts (prior quarter) vs 38 carousels (test quarter), same author, same cadence.
Three findings:
— Estimated dwell time rose from ~4s to ~13s per view (LinkedIn shows partial signals; they triangulated with slide-advance taps).
— Reach per post climbed 2.1x, consistent with dwell being an input the algorithm rewards.
— Most striking: inbound demo requests citing 'saw your LinkedIn' went from 2 to 9 across the quarter.
Limitations: the agency also tightened its topics that quarter, so dwell and relevance are 'confounded' (tangled causes) — correlation, not clean causation. n is small.
What it means for B2B: carousels buy attention seconds cheaply, and attention seconds appear upstream of both reach and intent.
Bottom line: format change plausibly drove pipeline, but a topic change rode along — read it as suggestive.
—
Продолжение про campaign objectives — @B2BLever
The 70% of B2B sharing you never see
The question: how much of your content travels through channels analytics cannot track, and does it matter for pipeline?
Dark social (sharing via DMs, Slack, email forwards, and private groups where no referrer tag survives) is structurally invisible. A vendor study reconstructed the gap by comparing self-reported 'how did you hear about us' survey answers against tracked referral data across several hundred B2B buyers. Method: triangulation — survey ground truth versus analytics attribution.
Three findings:
— Roughly 70% of content shares left no trackable referrer; they showed up in analytics as 'direct traffic'.
— Buyers who arrived via dark social converted at a higher rate than tracked social clicks.
— The biggest dark channel was not consumer messaging but internal Slack and email forwards inside the buyer's own org.
Caveats: self-report is noisy; survey samples skew toward respondents who remember — treat as directional.
What it means for B2B: your 'direct traffic' spike after a good post is probably earned reach, not brand luck. Add a 'where did you hear about us' field to demo forms; it is the cheapest dark-social meter you have.
Bottom line: untracked does not mean unimportant.
The question: how much of your content travels through channels analytics cannot track, and does it matter for pipeline?
Dark social (sharing via DMs, Slack, email forwards, and private groups where no referrer tag survives) is structurally invisible. A vendor study reconstructed the gap by comparing self-reported 'how did you hear about us' survey answers against tracked referral data across several hundred B2B buyers. Method: triangulation — survey ground truth versus analytics attribution.
Three findings:
— Roughly 70% of content shares left no trackable referrer; they showed up in analytics as 'direct traffic'.
— Buyers who arrived via dark social converted at a higher rate than tracked social clicks.
— The biggest dark channel was not consumer messaging but internal Slack and email forwards inside the buyer's own org.
Caveats: self-report is noisy; survey samples skew toward respondents who remember — treat as directional.
What it means for B2B: your 'direct traffic' spike after a good post is probably earned reach, not brand luck. Add a 'where did you hear about us' field to demo forms; it is the cheapest dark-social meter you have.
Bottom line: untracked does not mean unimportant.
Why the LinkedIn carousel stopped working
The question: did document carousels lose reach because the audience tired of them, or because the algorithm reweighted the format?
An analysis compared carousel (PDF document) reach across two time windows on the same set of accounts, isolating format from author. Method: same posters, same cadence, different quarters — a quasi-natural experiment that controls for who is posting.
Three findings:
— Median carousel reach fell sharply between the windows even as follower counts grew.
— The decline tracked a rise in carousel supply: as more accounts adopted the format, per-post reach compressed — a saturation effect, not pure fatigue.
— Carousels that front-loaded the payoff on slide one held reach better than 'swipe to the end for the answer' decks.
Caveats: reach is platform-reported and noisy; two windows cannot rule out a single algorithm change — treat as directional.
What it means for B2B: format edges are temporary arbitrage. The early adopters captured outsized reach; the format then mean-reverted as everyone copied it.
Bottom line: ride a format early, but do not build a content strategy on one. The advantage decays as adoption rises.
The question: did document carousels lose reach because the audience tired of them, or because the algorithm reweighted the format?
An analysis compared carousel (PDF document) reach across two time windows on the same set of accounts, isolating format from author. Method: same posters, same cadence, different quarters — a quasi-natural experiment that controls for who is posting.
Three findings:
— Median carousel reach fell sharply between the windows even as follower counts grew.
— The decline tracked a rise in carousel supply: as more accounts adopted the format, per-post reach compressed — a saturation effect, not pure fatigue.
— Carousels that front-loaded the payoff on slide one held reach better than 'swipe to the end for the answer' decks.
Caveats: reach is platform-reported and noisy; two windows cannot rule out a single algorithm change — treat as directional.
What it means for B2B: format edges are temporary arbitrage. The early adopters captured outsized reach; the format then mean-reverted as everyone copied it.
Bottom line: ride a format early, but do not build a content strategy on one. The advantage decays as adoption rises.
Thought leadership: read by few, trusted by buyers
The question: does executive thought-leadership content actually influence buying, or is it vanity publishing?
A recurring B2B decision-maker survey (a panel sample of senior buyers polled annually) asked how thought leadership affected vendor selection. Method: attitudinal survey, self-reported behavior — useful for direction, weak on hard causation.
Three findings:
— A majority of decision-makers said strong thought leadership directly led them to invite a vendor they had not considered.
— A similar share said weak or generic thought leadership made them lower their respect for a brand — the downside is symmetric.
— Time spent did not correlate with reach: the most influential pieces were often read by small, senior audiences, not viral ones.
Caveats: respondents may overstate rational influence; the publisher commissioned it — treat as directional.
What it means for B2B: thought leadership is a depth play, not a reach play. A piece read by 400 of the right buyers can outperform a post seen by 40,000. Measure who reads, not how many.
Bottom line: bad thought leadership is worse than none. Publish a strong view or stay quiet.
The question: does executive thought-leadership content actually influence buying, or is it vanity publishing?
A recurring B2B decision-maker survey (a panel sample of senior buyers polled annually) asked how thought leadership affected vendor selection. Method: attitudinal survey, self-reported behavior — useful for direction, weak on hard causation.
Three findings:
— A majority of decision-makers said strong thought leadership directly led them to invite a vendor they had not considered.
— A similar share said weak or generic thought leadership made them lower their respect for a brand — the downside is symmetric.
— Time spent did not correlate with reach: the most influential pieces were often read by small, senior audiences, not viral ones.
Caveats: respondents may overstate rational influence; the publisher commissioned it — treat as directional.
What it means for B2B: thought leadership is a depth play, not a reach play. A piece read by 400 of the right buyers can outperform a post seen by 40,000. Measure who reads, not how many.
Bottom line: bad thought leadership is worse than none. Publish a strong view or stay quiet.
Pairs well with this channel
@BrandYourselfDaily — Your daily personal-branding Q&A: we answer the questions creators and pros keep… Quietly one of the better feeds in the space.
@BrandYourselfDaily — Your daily personal-branding Q&A: we answer the questions creators and pros keep… Quietly one of the better feeds in the space.
Native video underperforms text for B2B — with one exception
The question: should B2B accounts shift to video because the platform pushes it, or does text still win for this audience?
A cross-format analysis measured reach and engagement-rate across text, image, and native video on B2B pages, normalized per follower. Method: per-format medians across thousands of posts — descriptive, not causal.
Three findings:
— Plain text and single-image posts had higher median engagement-rate than native video on B2B pages.
— Video reach was bimodal: most underperformed, a minority massively overperformed — averages hid the spread.
— The overperforming videos shared a trait: a strong on-screen text hook in the first two seconds, watchable with sound off.
Caveats: engagement-rate favors low-follower accounts; format choice is not random — treat as directional.
What it means for B2B: 'the platform favors video' is not the same as 'video favors you'. For most B2B pages, text is the higher-floor bet. Video is a high-variance play worth running only with a captions-first, sound-off design.
Bottom line: do not abandon text because the algorithm flirts with video. Match format to your variance tolerance.
The question: should B2B accounts shift to video because the platform pushes it, or does text still win for this audience?
A cross-format analysis measured reach and engagement-rate across text, image, and native video on B2B pages, normalized per follower. Method: per-format medians across thousands of posts — descriptive, not causal.
Three findings:
— Plain text and single-image posts had higher median engagement-rate than native video on B2B pages.
— Video reach was bimodal: most underperformed, a minority massively overperformed — averages hid the spread.
— The overperforming videos shared a trait: a strong on-screen text hook in the first two seconds, watchable with sound off.
Caveats: engagement-rate favors low-follower accounts; format choice is not random — treat as directional.
What it means for B2B: 'the platform favors video' is not the same as 'video favors you'. For most B2B pages, text is the higher-floor bet. Video is a high-variance play worth running only with a captions-first, sound-off design.
Bottom line: do not abandon text because the algorithm flirts with video. Match format to your variance tolerance.
95% of your buyers aren't buying — and that's the point
The question: if only a sliver of your market is in-market right now, who is your social content actually for?
The '95-5 rule' comes from B2B advertising research arguing that at any moment roughly 95% of buyers are out-market (not currently buying) and only 5% are in-market. Method: aggregated buying-cycle data across categories — a modeled estimate, not a precise census.
Three findings:
— Most content optimized for the in-market 5% (demos, pricing, comparisons) reaches people who will not act for months or years.
— Brands remembered before the buying trigger entered consideration sets at much higher rates — memory beats timing.
— The mechanism is 'mental availability': being recalled at the moment a need appears, which is unpredictable.
Caveats: ratios vary widely by category and deal size — treat as directional.
What it means for B2B: most of your social posts should build memory in the out-market 95%, not chase conversions from the 5%. That reframes 'this post got no leads' as possibly fine.
Bottom line: you are advertising to a future buyer who has no idea they will buy. Be memorable now.
The question: if only a sliver of your market is in-market right now, who is your social content actually for?
The '95-5 rule' comes from B2B advertising research arguing that at any moment roughly 95% of buyers are out-market (not currently buying) and only 5% are in-market. Method: aggregated buying-cycle data across categories — a modeled estimate, not a precise census.
Three findings:
— Most content optimized for the in-market 5% (demos, pricing, comparisons) reaches people who will not act for months or years.
— Brands remembered before the buying trigger entered consideration sets at much higher rates — memory beats timing.
— The mechanism is 'mental availability': being recalled at the moment a need appears, which is unpredictable.
Caveats: ratios vary widely by category and deal size — treat as directional.
What it means for B2B: most of your social posts should build memory in the out-market 95%, not chase conversions from the 5%. That reframes 'this post got no leads' as possibly fine.
Bottom line: you are advertising to a future buyer who has no idea they will buy. Be memorable now.
The engagement signal hierarchy nobody publishes
The question: not all engagement is equal — so which actions actually move reach, in what order?
An analysis ranked engagement signals by their association with downstream reach, controlling for post size. Method: partial correlations — measuring each signal's link to reach while holding the others constant, so credit is not double-counted.
Three findings:
— Saves and shares (effortful, intent-signaling actions) associated most strongly with extended reach.
— Comments sat in the middle; likes near the bottom — likes are cheap and the model appears to price them accordingly.
— Reshares with added commentary outperformed bare reshares, implying the platform rewards new content creation, not pass-along.
Caveats: signal weights likely shift over time and by surface; correlational — treat as directional.
What it means for B2B: a call-to-action of 'save this for your next planning cycle' may do more for reach than 'agree?'. Design the ask around the high-value action, not the easy one.
Bottom line: chase saves and qualified shares. Likes are applause; saves are intent.
The question: not all engagement is equal — so which actions actually move reach, in what order?
An analysis ranked engagement signals by their association with downstream reach, controlling for post size. Method: partial correlations — measuring each signal's link to reach while holding the others constant, so credit is not double-counted.
Three findings:
— Saves and shares (effortful, intent-signaling actions) associated most strongly with extended reach.
— Comments sat in the middle; likes near the bottom — likes are cheap and the model appears to price them accordingly.
— Reshares with added commentary outperformed bare reshares, implying the platform rewards new content creation, not pass-along.
Caveats: signal weights likely shift over time and by surface; correlational — treat as directional.
What it means for B2B: a call-to-action of 'save this for your next planning cycle' may do more for reach than 'agree?'. Design the ask around the high-value action, not the easy one.
Bottom line: chase saves and qualified shares. Likes are applause; saves are intent.
Posting more can shrink your total reach
The question: does increasing posting frequency grow your aggregate reach, or do your own posts start competing for the same feed slots?
A frequency study tracked accounts that ramped from a few posts a week to daily, measuring per-post and total weekly reach. Method: within-account before/after — each account is its own control, reducing confounds.
Three findings:
— Per-post reach fell as frequency rose, consistent with audience and feed saturation.
— Total weekly reach grew up to a point, then flattened — diminishing returns, not linear gains.
— Posting twice in a short window often cannibalized the first post, which lost reach when the second appeared.
Caveats: optimal frequency varies by audience size; observational — treat as directional.
What it means for B2B: 'post more' is not free. Past a threshold, you are dividing the same attention across more posts and dropping per-post quality. Find the point where total reach plateaus and stop there.
Bottom line: frequency has a ceiling. More posts can mean less reach per post and no gain overall.
The question: does increasing posting frequency grow your aggregate reach, or do your own posts start competing for the same feed slots?
A frequency study tracked accounts that ramped from a few posts a week to daily, measuring per-post and total weekly reach. Method: within-account before/after — each account is its own control, reducing confounds.
Three findings:
— Per-post reach fell as frequency rose, consistent with audience and feed saturation.
— Total weekly reach grew up to a point, then flattened — diminishing returns, not linear gains.
— Posting twice in a short window often cannibalized the first post, which lost reach when the second appeared.
Caveats: optimal frequency varies by audience size; observational — treat as directional.
What it means for B2B: 'post more' is not free. Past a threshold, you are dividing the same attention across more posts and dropping per-post quality. Find the point where total reach plateaus and stop there.
Bottom line: frequency has a ceiling. More posts can mean less reach per post and no gain overall.
LinkedIn newsletters: the dwell-time loophole
The question: do native newsletters earn different distribution than ordinary posts, and is the difference worth the format constraints?
An analysis compared native newsletter editions against the same authors' regular posts on reach, dwell time, and subscriber retention. Method: same-author comparison across formats — controls for voice and audience.
Three findings:
— Newsletter editions generated longer dwell time per reader, since long-form pulls committed readers, not scrollers.
— The subscribe mechanic produced a notification on each edition — a distribution channel ordinary posts lack.
— Reach per edition was lower than viral posts but far more predictable — a floor, not a lottery.
Caveats: newsletters self-select engaged audiences; survivorship bias in who keeps publishing — treat as directional.
What it means for B2B: newsletters trade peak reach for reliability and dwell time — the signal the algorithm appears to value. For a B2B brand that needs consistent touchpoints with a known audience, predictability can beat virality.
Bottom line: if you value a dependable floor over an occasional spike, the newsletter format is structurally on your side.
The question: do native newsletters earn different distribution than ordinary posts, and is the difference worth the format constraints?
An analysis compared native newsletter editions against the same authors' regular posts on reach, dwell time, and subscriber retention. Method: same-author comparison across formats — controls for voice and audience.
Three findings:
— Newsletter editions generated longer dwell time per reader, since long-form pulls committed readers, not scrollers.
— The subscribe mechanic produced a notification on each edition — a distribution channel ordinary posts lack.
— Reach per edition was lower than viral posts but far more predictable — a floor, not a lottery.
Caveats: newsletters self-select engaged audiences; survivorship bias in who keeps publishing — treat as directional.
What it means for B2B: newsletters trade peak reach for reliability and dwell time — the signal the algorithm appears to value. For a B2B brand that needs consistent touchpoints with a known audience, predictability can beat virality.
Bottom line: if you value a dependable floor over an occasional spike, the newsletter format is structurally on your side.
B2B buyers are more emotional than they admit
The question: do rational, feature-led B2B messages outperform emotional ones, given that buyers describe themselves as logical?
A brand-research study compared business buyers' emotional connection to brands against consumer benchmarks, and linked it to willingness to pay and consider. Method: large attitudinal survey scored on emotional-connection scales — self-report, so directionally indicative.
Three findings:
— B2B buyers reported a higher emotional connection to vendors than consumer buyers did to consumer brands, contradicting the 'rational B2B' assumption.
— Personal value (career risk, looking smart, avoiding blame) drove consideration more than business value (efficiency, price).
— Fear of making a costly visible mistake was a stronger motivator than the upside of a good decision — loss aversion at work.
Caveats: stated attitudes are not purchases; benchmarks differ in method — treat as directional.
What it means for B2B: content that addresses the buyer's personal risk ('how to not get blamed if this fails') can outperform feature lists. You are de-risking a human, not optimizing a spreadsheet.
Bottom line: B2B is personal. Speak to the career on the line, not just the line item.
The question: do rational, feature-led B2B messages outperform emotional ones, given that buyers describe themselves as logical?
A brand-research study compared business buyers' emotional connection to brands against consumer benchmarks, and linked it to willingness to pay and consider. Method: large attitudinal survey scored on emotional-connection scales — self-report, so directionally indicative.
Three findings:
— B2B buyers reported a higher emotional connection to vendors than consumer buyers did to consumer brands, contradicting the 'rational B2B' assumption.
— Personal value (career risk, looking smart, avoiding blame) drove consideration more than business value (efficiency, price).
— Fear of making a costly visible mistake was a stronger motivator than the upside of a good decision — loss aversion at work.
Caveats: stated attitudes are not purchases; benchmarks differ in method — treat as directional.
What it means for B2B: content that addresses the buyer's personal risk ('how to not get blamed if this fails') can outperform feature lists. You are de-risking a human, not optimizing a spreadsheet.
Bottom line: B2B is personal. Speak to the career on the line, not just the line item.
The 'see more' fold is your real headline
The question: how much does the visible portion above the 'see more' cutoff determine whether a post gets read at all?
An analysis correlated the first-line content of posts with expand-rate (the share of viewers who tapped 'see more') and downstream dwell time. Method: text-feature analysis against engagement — descriptive, correlational.
Three findings:
— Posts whose first two lines posed a specific tension or number had markedly higher expand-rates than posts that opened with context-setting.
— Front-loading the conclusion did not kill curiosity; specificity did the work, not withholding.
— Expand-rate predicted dwell time, which (per other work) predicts reach — a chain of cheap-to-measure signals.
Caveats: first-line effects entangle with topic and author; correlational — treat as directional.
What it means for B2B: the ~140 characters above the fold are doing the job of a headline. 'Here are three lessons from our Q3 campaign' wastes it; 'Our best-performing campaign had the worst click-through rate — here's why' earns the expand.
Bottom line: write the visible lines like a headline, because functionally that is what they are.
The question: how much does the visible portion above the 'see more' cutoff determine whether a post gets read at all?
An analysis correlated the first-line content of posts with expand-rate (the share of viewers who tapped 'see more') and downstream dwell time. Method: text-feature analysis against engagement — descriptive, correlational.
Three findings:
— Posts whose first two lines posed a specific tension or number had markedly higher expand-rates than posts that opened with context-setting.
— Front-loading the conclusion did not kill curiosity; specificity did the work, not withholding.
— Expand-rate predicted dwell time, which (per other work) predicts reach — a chain of cheap-to-measure signals.
Caveats: first-line effects entangle with topic and author; correlational — treat as directional.
What it means for B2B: the ~140 characters above the fold are doing the job of a headline. 'Here are three lessons from our Q3 campaign' wastes it; 'Our best-performing campaign had the worst click-through rate — here's why' earns the expand.
Bottom line: write the visible lines like a headline, because functionally that is what they are.
Last-touch attribution is overcrediting your bottom-funnel
The question: when last-touch attribution gives a branded search or demo form the credit, what earlier touches are being erased?
A modeling exercise compared last-touch attribution (all credit to the final interaction) against a multi-touch model on the same B2B deals. Method: re-running both models over identical conversion paths — isolating the methodology effect.
Three findings:
— Last-touch concentrated credit on branded search and direct visits — channels that capture demand others created.
— Top-of-funnel social and content, which rarely sat in the final touch, were systematically undervalued.
— Deals with more touchpoints closed at higher rates, suggesting breadth of exposure matters even when no single touch gets credit.
Caveats: multi-touch models embed their own assumptions; both are estimates, not truth — treat as directional.
What it means for B2B: if you cut the channels last-touch ignores, you may quietly starve the demand that feeds your 'high-performing' branded search. The model that looks efficient can be hiding the engine.
Bottom line: the channel that gets the credit is often not the channel that did the work.
The question: when last-touch attribution gives a branded search or demo form the credit, what earlier touches are being erased?
A modeling exercise compared last-touch attribution (all credit to the final interaction) against a multi-touch model on the same B2B deals. Method: re-running both models over identical conversion paths — isolating the methodology effect.
Three findings:
— Last-touch concentrated credit on branded search and direct visits — channels that capture demand others created.
— Top-of-funnel social and content, which rarely sat in the final touch, were systematically undervalued.
— Deals with more touchpoints closed at higher rates, suggesting breadth of exposure matters even when no single touch gets credit.
Caveats: multi-touch models embed their own assumptions; both are estimates, not truth — treat as directional.
What it means for B2B: if you cut the channels last-touch ignores, you may quietly starve the demand that feeds your 'high-performing' branded search. The model that looks efficient can be hiding the engine.
Bottom line: the channel that gets the credit is often not the channel that did the work.
Employee posts out-reach the company page — but the math has a catch
The question: do employee-shared posts genuinely extend brand reach, or do they mostly recirculate to the same overlapping networks?
A reach study compared identical content posted from a company page versus from employees' personal profiles, measuring unique reach and network overlap. Method: same content, different distributors, deduplicated audience — the overlap measure is the important part.
Three findings:
— Personal profiles out-reached the company page per post, consistent with the platform favoring people over brands.
— But employees' networks overlapped heavily with each other, so aggregate unique reach grew far less than the sum of individual reach implied.
— Reach added by an employee depended on network diversity, not follower count — a connected outsider added more than a high-follower insider.
Caveats: overlap is hard to measure precisely; samples skew to willing sharers — treat as directional.
What it means for B2B: 'we have 500 employees, that's 500x reach' is the overlap fallacy. Real incremental reach comes from employees whose networks differ from each other's, not from sheer headcount.
Bottom line: advocacy scales with network diversity, not employee count.
The question: do employee-shared posts genuinely extend brand reach, or do they mostly recirculate to the same overlapping networks?
A reach study compared identical content posted from a company page versus from employees' personal profiles, measuring unique reach and network overlap. Method: same content, different distributors, deduplicated audience — the overlap measure is the important part.
Three findings:
— Personal profiles out-reached the company page per post, consistent with the platform favoring people over brands.
— But employees' networks overlapped heavily with each other, so aggregate unique reach grew far less than the sum of individual reach implied.
— Reach added by an employee depended on network diversity, not follower count — a connected outsider added more than a high-follower insider.
Caveats: overlap is hard to measure precisely; samples skew to willing sharers — treat as directional.
What it means for B2B: 'we have 500 employees, that's 500x reach' is the overlap fallacy. Real incremental reach comes from employees whose networks differ from each other's, not from sheer headcount.
Bottom line: advocacy scales with network diversity, not employee count.
Pairs well with this channel
@ReachLabReports — Instagram growth decoded through numbers: follower-velocity benchmarks, reach-rate… Quietly one of the better feeds in the space.
@ReachLabReports — Instagram growth decoded through numbers: follower-velocity benchmarks, reach-rate… Quietly one of the better feeds in the space.