Serro
Monthly Performance Report
April 1 to April 30, 2026
LinkedIn · Jake Kim

At a glance

Impressions
30.6K
across 9 posts
Engagements
352
likes, comments, shares, saves, sends
New Followers
+67
▲ 1.6% · 4,295 to 4,362
Posts Shipped
9
≈ 2 posts/week
Saves
9
high-intent signal
Read this month as a baseline, not a trend

April is the first tracked month for this account (analytics began April 7, when the profile was connected). Volume was light at 9 posts, and a single off-topic post (the Tim Cook "xApple" joke) accounted for 67% of all impressions. The numbers below are real, but treat April as the establishing baseline. May is the first month with enough on-brand volume to read as a trend.

Performance overview

Follower growth

+67 net followers from when tracking began (Apr 7) to month end. The clearest step came on Apr 22 to 23, right after the Tim Cook post went wide.

Posts by impressions

One post (Tim Cook xApple) drove 20.4K of the month's 30.6K impressions. The rest of the feed clustered between 380 and 2,800.

Top performing posts

What broke through this month

Each list ranks by a different signal: impressions for reach, saves for "I want to come back to this" intent, and engagement rate for resonance on smaller-reach posts. With only 9 posts and a wide spread, the leaderboards are directional rather than definitive.

🏆 By impressions (reach)

Who got seen the most.

1
Dear Tim Cook (xApple)
Off-topicImage · Apr 22
20.4Kimpressions
2
Our AI-first coding challenge
Founder / BTSText · Apr 7
2.8Kimpressions
3
Why you need a TPM at 50 people
FrameworkText · Apr 15
1.7Kimpressions
4
1.6Kimpressions
5
1.2Kimpressions

💾 By saves (high intent)

The strongest B2B signal. Readers bookmarking to revisit.

1
Our AI-first coding challenge
Founder / BTSText · Apr 7
3saves
2
Dear Tim Cook (xApple)
Off-topicImage · Apr 22
2saves
3
Why you need a TPM at 50 people
FrameworkText · Apr 15
2saves
4
Context is the real moat
Hot takeImage · Apr 20
1saves
5
1saves

⚡ By engagement rate

Smaller-reach posts that overperformed for their audience.

1
Welcome Wenxue Zhang
Product/companyImage · Apr 30
2.3%eng. rate
2
LaserData is inevitable
CommunityQuote · Apr 5
2.0%eng. rate
3
2.0%eng. rate
4
Context is the real moat
Hot takeImage · Apr 20
1.6%eng. rate
5
Build on LaserData
CommunityQuote · Apr 14
1.3%eng. rate

Format analysis

Format mix this month

Long-form text led volume. Quote reshares were the lightest format and the weakest performers.

Average impressions by format

Image looks strongest, but only because the Tim Cook post is an image. Strip that one outlier and text is the real workhorse.

Content pattern analysis

What we posted, and what each pattern returned

We read each of the 9 posts and classified them by the job the post was doing. Five patterns appeared (no event promo this month). The breakdown shows where the on-brand engineering thought leadership stands before May scales it up.

Avg impressions per pattern

Off-topic tops the chart entirely on the Tim Cook outlier. Founder/BTS is the strongest on-brand pattern this month.

Total saves per pattern

Saves are scarce at this volume. Founder/BTS and hot takes edged ahead.

Pattern breakdown

Five content patterns observed across 9 published posts.

Pattern Posts Total imp. Avg imp./post Total saves Top performer
Off-topic / community share 3 21,320 7,107 2 Tim Cook xApple (20.4K imp)
Founder / company BTS 1 2,806 2,806 3 AI-first coding challenge
Frameworks 1 1,662 1,662 2 Why you need a TPM
Hot takes vs incumbents 3 3,913 1,304 2 AI adoption review (1.6K imp)
Product/company updates 1 915 915 0 Welcome Wenxue Zhang
📌 The viral post was off-brand, and that is worth naming

The Tim Cook "xApple" image drove 20.4K impressions, 6x the next-best post. It is funny and humanizes Jake's ex-Apple story, but it says nothing about Serro or the problem Serro solves. It pulled reach without pulling the engineering-leader audience the account is built for. Fine as an occasional personality post. Not a template to chase.

📌 The on-brand engine is Jake's engineering thought leadership

Strip the meme and the real signal is clear: the AI-first coding challenge (2.8K, 3 saves), the TPM framework (1.7K, 2 saves), and the AI-adoption hot take (1.6K) all landed with the right crowd. These are the posts that explain how Serro thinks. They are where May should concentrate.

📌 Quote reshares are dead weight

The two LaserData quote-reshares averaged 438 impressions, the lowest of any format. A one-line caption on someone else's post gives the algorithm nothing to expand and the reader nothing to save. Cut these or replace them with an original take.

Hook analysis

What the best hooks had in common

The first sentence decides reach on LinkedIn. Across April's on-brand winners, the same ingredients show up. The single most repeatable asset is Jake's "I left Apple / 5 years at Apple" credential paired with a contrarian claim.

1
A credential or number up front

"I left Apple 2 years ago to build real-time context infra", "If your startup has hit around 50 people", "More than 90% of our code at Serro is AI written". Concrete numbers and the Apple credential establish authority before the argument starts.

2
A named target to push against

"Google is tying performance reviews to AI adoption. I think that's going to backfire." Naming Google (or a specific practice) gives the reader a clear position to react to. The posts with no named tension (quote reshares) went nowhere.

3
A contrarian or counterintuitive claim

"Context over everything", "the title should reflect that", "that's a feature, not a bug". Each hook promises a take the reader has not heard, which is what earns the comment and the save.

4
An explicit promise of payoff

"Here's my answer", "Let me explain", "Here's what that looks like". The strongest openers tell the reader the explainer is coming, which buys the dwell time the algorithm rewards.

5
First person, founder voice

"I've been thinking about", "I left Apple", "At Serro, we". The account works hardest when it sounds like Jake talking, not a company broadcasting. The product post (Wenxue welcome) was warm but had no argument, and pulled the lowest reach of the on-brand posts.

What didn't work

Bottom performers and why

The two lowest-reach posts were both quote reshares of a partner (LaserData). At 384 and 492 impressions they were roughly 5x below the on-brand text posts.

9 of 9
Build on LaserData (reshare)
384 impressions · 0 saves · Apr 14
Hypothesis: bare reshare, no original argument

"If you're building real-time agents, build on LaserData!" is a one-line endorsement on someone else's post. There is no hook, no number, no take, and nothing to save. Reshares like this reach the resharer's most loyal followers and stop there. If a partner is worth amplifying, write an original post explaining why it matters to Serro's audience.

8 of 9
LaserData is inevitable (reshare)
492 impressions · 0 saves · Apr 5
Hypothesis: same failure mode, second instance

Another quote-reshare with a short caption. Two of these in April, both at the bottom of the table. The pattern is consistent enough to retire. The engagement rate looks acceptable (2.0%) only because the denominator is tiny.

🛑 Failure modes to avoid in May

1. Bare quote-reshares with a one-line caption (lowest reach of the month, twice).
2. Chasing off-topic virality: the Tim Cook post won reach but not the ICP audience.
3. Product and hire posts with no argument attached (warm, but lowest on-brand reach).

What to do next month

Playbook for May

Five moves to build on April's signal and fix its misses.

1
Lead with Jake's ex-Apple engineering thought leadership

The on-brand posts that worked all carried Jake's operator credibility and a contrarian claim. Make this the spine of the calendar: aim for 2 to 3 thought-leadership posts per week on how elite engineering orgs actually work and where AI changes the job. This is the content that pulls the right audience.

Evidence: coding challenge (2.8K, 3 saves), TPM framework (1.7K), AI-adoption take (1.6K) were the on-brand leaders.
2
Turn theses into named frameworks

The "you need a TPM at 50 people" post earned saves because it gave readers a rule they could apply. Package more takes as frameworks with a clear trigger and a short list (symptoms, layers, steps). Frameworks are the format most likely to be saved and reshared by engineering leaders.

Evidence: the one framework post (TPM) drove 2 of the month's 9 saves on a single post.
3
Cut quote-reshares; replace with original takes

The two LaserData reshares were the bottom two posts. If a partner or a piece of news is worth Jake's attention, write an original post that makes an argument about why it matters to engineering orgs. Keep the substance, drop the bare reshare.

Evidence: reshares averaged 438 impressions vs ~1,800 for on-brand text.
4
Pair product and hire posts with substance

Hire announcements (Wenxue) build culture but carry little reach on their own. When announcing a hire or feature, anchor it to the problem Serro solves, the way the Wenxue post tied her to "work signals ingestion." Better still, follow a product post with a thought-leadership post on the same theme within a few days.

Evidence: the product/hire post pulled 915 impressions, the lowest of the on-brand posts.
5
Enable lead scraping to get a real ICP read

We cannot yet measure who is engaging by role and company because lead scraping is off for this profile. Turning it on in Ordinal unlocks the ICP-fit metric and the named senior-engager list that make these reports actionable for sales. Worth doing before the May report.

Evidence: no lead-level audience data was available for April.