One Year on LinkedIn

The first thing the data tells you is that this is not a steady drip -- it is a series of pulses. Across 365 days the profile shows a clear rhythm: every few days a spike rises out of a baseline that hovers between 500 and 1,500 impressions. The largest spike is January 28, 2026, at 31,654 impressions, an algorithmic outlier nearly 2× the second-highest day. The pulses correspond to publishing events -- when content lands, the audience shows up; when it doesn't, the residual decay of prior posts keeps things ticking over. Two periods stand out as visibly compressed: July 2025 and February 2026, where the pulses thin out and the baseline dominates. Both align with US summer and mid-winter dips on the platform broadly, but the unusually flat lines suggest a posting-cadence drop rather than a reach problem.

Smoothing the daily noise reveals what the eye misses on the raw chart. The 14-day rolling average peaks in late May 2025 at roughly 4,500 impressions per day, drops sharply through June and bottoms out in July around 1,000. From August it climbs back to a stable plateau of 1,800–2,400, with secondary peaks in November–December and again in April. The takeaway is that the launch month set a peak that has not been matched, but the floor has risen -- the post-July baseline is consistently higher than the pre-July floor. Audience size grew across the year even as the high-water mark drifted lower; this is the canonical signature of a creator transitioning from "novelty reach" to "compound audience reach."

Plotting impressions cumulatively flattens the noise into something more honest: the total reach trajectory is roughly linear, with subtle slope changes in early summer and late winter. By the 100-day mark the profile had accumulated about 200,000 impressions; by 200 days, ~390,000; by 300 days, ~580,000. The line never goes negative -- there is always reach happening. Notable inflection points are visible around day 60 (a steepening from the May launch) and day 180 (a flattening through the summer). The slope of the second half (roughly 1,800/day) is meaningfully shallower than the slope of the first half (roughly 2,100/day), which formalizes the H1-vs-H2 reach decline addressed in figure 16.

Ranking the months produces a clear hierarchy. May 2025 leads with 102,415 impressions -- 36% above the second-place month, December 2025 (75,045). The bottom three months are February 2026 (29,924), July 2025 (35,274), and September 2025 (50,508). The pattern matches calendar-driven attention dips: 4th of July week, mid-winter, post-summer reset. December outperforming might surprise -- but it tracks with platform research showing a December-pre-Christmas content spike before the holiday silence. The standard deviation across months is high: σ = ~21,000 impressions, indicating that monthly performance is volatile and dependent on a few hits. The chart does not yet say what drove May; it says only that May was extraordinary and uneven.

Where total impressions tell you how loud you were, engagement rate tells you whether anyone leaned in. Three months break above 2.30%: May 2025 (3.21%), November 2025 (2.46%), and April 2026 (2.39%). May was the launch effect -- content was novel. November and April are more interesting: they are recovery months after slumps, suggesting that engagement quality rises when posting cadence reduces and selection improves. The two clearly weak months on quality are October 2025 (1.34%) and June 2025 (1.66%). October's collapse is the most striking anomaly in the dataset because impressions held that month -- only the engagement rate fell. Whatever the topic mix or format mix was that October, it warrants a forensic review.

The LinkedIn export does not list every post -- only the top 50 by engagement and the top 50 by impressions. To estimate posting cadence, this analysis counts daily impression spikes (any day exceeding 2,000 impressions, well above the 1,400-impression median baseline). On that basis, May 2025 saw an estimated 22 posting days, December 2025 saw 17, and July 2025 collapsed to 4. The total inferred posting days for the year is 122, suggesting roughly 2 to 3 posts per week on average, with strong months reaching 4–5 per week and weak months below 1. This figure is a lower bound: days with multiple posts share a single spike. The spike-day count tracks closely with top-60 publication count, validating the inference.

Aggregated to a weekly view, posting cadence is bimodal. The single most common week has 2 estimated posts; the next-most-common week has 0. About 25% of weeks had zero posting spikes, meaning the audience went a full seven days without a publishing event from this account. Weeks with 4+ posts are rare (under 10% of the year) and tend to cluster in May and December. The takeaway is that the publishing cadence is inconsistent: there is no observable steady 3-per-week rhythm. A regular cadence would show as a tight cluster around the median; this distribution is wide and flat. Consistency is a lever that has not been pulled.

A textbook power law. Two-thirds of days (243 of 365) sit below 2,000 impressions, and another 15% sit between 2,000 and 3,000. Only 23 days -- 6.3% of the year -- exceed 5,000 impressions, and exactly one day exceeds 10,000 (the Jan 28, 2026 outlier at 31,654). When the top 6.3% of days produce roughly 31% of total impressions, the implication is that a handful of breakthrough posts carries the year. This is consistent with creator-economy research on LinkedIn specifically; the platform's algorithm rewards a small number of pieces with disproportionate reach. The strategy implication is that protecting the conditions for breakthrough posts (timing, topic, format) matters more than maximizing total post count.

Thursday delivers the highest average impressions (2,326), followed by Monday (2,144). Saturday is the clear outlier at 1,191 -- 38% below Thursday and 49% below the highest weekday. The other five weekdays cluster between 1,840 (Wednesday) and 1,979 (Tuesday), suggesting the platform algorithm distributes weekday content roughly evenly. The volume story is straightforward: post on a weekday and you'll be seen; post on Saturday and you won't. Sunday is the surprise -- at 1,949 it is comparable to Tuesday and Wednesday, which contradicts the conventional "weekend is dead" advice. The volume chart is one half of the story, however; figure 10 shows the other half.

Where Thursday wins on volume, Sunday wins on quality. Sunday's 2.36% engagement rate is the highest of the week, followed by Friday (2.26%) and Wednesday (2.24%). The lowest engagement rate is Saturday (1.98%). The Sunday finding is the most actionable insight in this paper because it inverts the standard creator advice. The mechanism is plausible -- weekend posts face less competition for the feed slot, and the audience that scrolls Sunday morning is the audience that opted in. Combined with figure 9 (Sunday's volume is comparable to weekdays), Sunday becomes the highest-leverage publishing day in the dataset: comparable reach with the year's best engagement rate. It is also the most underused -- the next figure shows where top posts actually land.

Of the 60 best-performing posts of the year, 13 (22%) were published on Sunday, more than any other day. Tuesday is second with 12 posts; Wednesday and Thursday tie at 9. Saturday is again the bottom with only 4 top posts (6.7%) -- meaning Saturday content is not just structurally weaker, it almost never produces a hit. The Sunday + Tuesday + Wednesday + Thursday cluster accounts for 43 of 60 top posts, or 71.7%. Friday is the surprise underperformer: at 8 top posts it lags despite figure 10 showing Friday as the second-best engagement-rate day. This may reflect lower friday posting frequency rather than weak Friday performance. The combined evidence from figures 9, 10, and 11 is unanimous: Sunday is the under-utilized lane.

A separate signal entirely. Tuesday acquires 15.0 new followers per day on average -- the highest of the week. Thursday (14.0) and Wednesday (13.0) follow. Sunday and Saturday are tied at the bottom (8.8 and 8.2). The Tuesday spike almost certainly reflects the Talent Trends newsletter cadence -- Tuesday is the company-wide email day, and the newsletter halo extends to LinkedIn. This decouples cleanly from impressions and engagement. Tuesday comes third for reach and third for engagement, and still first for the audience actually committing. The strategic implication is that Tuesday and Sunday play different roles -- Tuesday converts existing readers into followers; Sunday deepens the relationship with people already following.

The curve is monotonic -- followers never decline -- and the slope steepens in the second half. The starting estimate is 2,782 followers; the ending count is 7,087. The first half added 1,787 followers; the second half added 2,518. There is no plateau visible in the back third, suggesting the audience is still in the steep-acquisition phase of the S-curve. Visually, the curve has two acceleration zones: around day 30 (early June 2025) as the launch month converted, and around day 200 (early November 2025) when a sequence of strong engagement months pulled followers in. There is no observable contraction zone -- even during the July and February slumps, the cumulative line continues climbing.

The monthly bar chart for new followers reveals a different ranking than impressions. May 2025 leads at 606 new followers, but the second and third positions are March 2026 (501) and November 2025 (487) -- both months with relatively modest impressions. April 2026 (432), January 2026 (408), and September 2025 (386) round out the top six. The bottom three are the predictable July 2025 (108), August 2025 (164), and December 2025 (365 -- surprisingly low for a high-impression month). The lesson: follower acquisition does not track impressions linearly. Months that combine moderate volume with high engagement quality -- November, March -- outperform months that combine high volume with low quality -- December.

Normalizing for reach reveals the most important trend in this dataset. The first six months averaged 4.77 followers per 1K impressions; the last six months averaged 7.76 -- a 62.7% improvement. The monthly chart shows the climb is steady, not lumpy: the metric improves quarter over quarter with only one mild reversal (January 2026). November 2025 sets the absolute peak at 8.92 followers per 1K. This is the central counterintuitive finding of the case study: impressions declined, but each impression became more valuable. An audience converting at 7.76/1K is qualified -- they are seeing the content and self-selecting in. The mechanism is most likely audience refinement (the wrong followers from the May launch churned out, and the right ones stayed) plus content fit (the topic mix has narrowed toward what resonates).

The H1-vs-H2 comparison condenses the entire argument into a single chart. Impressions fell 13.4% (374,415 → 324,380). Engagements fell 13.8% (8,196 → 7,066). Engagement rate held flat (2.19% → 2.18%). But new followers grew 40.9% (1,787 → 2,518) and follower-conversion efficiency grew 62.7% (4.77 → 7.76 per 1K impressions). The two reach metrics declined; the two business metrics improved. This is the inversion that the rest of the paper has been building toward. If the assessment of the year ended after figure 4, it would read as a story of decline. After figure 15, it reads as a story of audience refinement. The vanity metrics fell while the metric that actually matters improved.

Each cell shows the average daily impressions for that day-of-week within that month -- 84 cells total. The heatmap consolidates the three preceding figures into a single visual. May 2025 glows uniformly -- every weekday is hot. July 2025 and February 2026 are uniformly dim. Sunday holds value through the year better than other days; even during slumps, Sunday cells are darker than their Saturday neighbors. October 2025 is the most surprising row: Tuesday and Wednesday are bright (a heavy posting cadence) but the engagement-rate map shows October collapsed on quality (figure 5 -- 1.34% ER). This is the visual signature of high reach plus low resonance -- the algorithm pushed out content that the audience did not lean into.

Among the 40 top posts where both metrics are available, engagement rate ranges from 1.01% to 4.41% -- a 4× spread. The median is 2.24%, the mean is 2.42%, and the 90th percentile is 4.06%. The 3% threshold separates "broadcast" posts from "resonance" posts: posts above 3% are the ones where the audience genuinely leaned in, and there are 11 of these in the top-60. The single highest-ER post is May 23, 2025, at 4.25% (530 engagements on 12,480 impressions); the single lowest-ER top post is January 28, 2026, at 1.29% (the broadcast outlier). The histogram skews right with a long tail of high-quality posts -- this is the shape you want. A left-skewed histogram (most top posts at low ER) would suggest engagement-bait or algorithmic flukes; this distribution suggests genuine content-market fit.

Plotting impressions against engagements with reference lines makes the post-quality structure visible. Most posts cluster near the 2% diagonal -- the median performance band. The cluster above the 3% line contains the genuine connection posts, ten of them sitting between 5K and 17K impressions with 200–700 engagements. The single dot far to the right (31K impressions, ~400 engagements) is the January 28, 2026 broadcast post -- high reach but well below the 2% line, indicating the algorithm pushed it out widely without a corresponding spike in engagement. There is no observable correlation between impressions and engagement rate within the top set; the highest-ER post (May 23) and the highest-impression post (Jan 28) are on opposite ends of the chart. Reach and resonance are independent variables in this dataset -- chasing one will not produce the other.

The audience is bicontinental and senior. Greater Dublin alone accounts for 24.4% of followers -- a single city carrying nearly a quarter of the audience, reflecting Irish background and university network. Grand Rapids (8.9%) and South Bend (7.7%) account for the U.S. Midwest concentration tied to current work and Notre Dame. Industry is broad -- Financial Services leads at 7.8% but no single sector exceeds 8%, indicating a diversified follower base. Seniority is the cleanest signal: 51.3% of followers are at decision-maker level (Senior + Director + CXO + Owner combined). For an executive-search adjacent personal brand, this is exactly the right altitude. The unresolved tension is geographic: the buyer (Midwest exec-search market) and the audience (heavily Dublin) are partially but not fully aligned. The strategic question for year two is whether to bridge the two halves or specialize.