Test posts
Ten posts to fly, so you never have to paste your own work to test the crew.
Each one is broken in a different way — corporate filler, engagement bait, rushed grammar, one long unstructured wall, and a clean control post the crew should mostly leave alone.
- TP-01
AI transformation announcement
Corporate filler, AI tells, three-part constructions
In today's rapidly evolving landscape, artificial intelligence is fundamentally reshaping how enterprises operate, compete, and deliver value. It is important to note that organisations must leverage these transformative capabilities to unlock unprecedented growth. By harnessing cutting-edge innovation, forward-thinking leaders can drive holistic outcomes at scale. We are thrilled to announce our new AI Centre of Excellence, a game-changing initiative that will revolutionise the way we work. The future is not coming. It is already here. Thoughts?
ExecutiveAnnouncement - TP-02
Founder post-mortem
Strong voice with a buried hook — impact and structure passes
We shut down our second product last month. It had 340 users, £6,000 of monthly revenue, and a churn rate that told us the truth long before we were willing to hear it. Building it took nine months. Admitting it was wrong took four weeks longer than it should have. The mistake was not the product. The mistake was that we never wrote down what would make us stop. Without that line, every bad month looked survivable. Now every project starts with a kill condition, written before the first commit and signed by two people. If you cannot say what would make you stop, you are not running an experiment. You are just hoping.
FounderStory - TP-03
Technical explainer
Readability, long sentences, jargon density
Our p99 latency dropped from 1.8s to 240ms last quarter and the reason is unglamorous: we were doing authorisation checks inside a loop that ran once per row returned, which meant a query returning two hundred rows performed two hundred separate policy evaluations against a service that was itself waiting on a connection pool that had been sized for a workload we stopped running in 2023. We moved the check above the loop, batched it, and cached the result for the duration of the request. No new infrastructure. No rewrite. One function moved eleven lines up. The lesson is that most performance work is archaeology, not engineering.
EngineerExplainer - TP-04
Engagement bait
Tone pass, cliché detection, hollow hook
Most people will scroll past this post. That's okay. It's not for them. 👇 Yesterday I met a cleaner at 5am in an empty office. I asked her why she was smiling. She said: "Because I'm here." Let that sink in. Success isn't about talent. It's not about luck. It's about showing up. Agree? 🙌 ♻️ Repost if this resonated. Follow me for more lessons on leadership. #motivation #leadership #mindset #success #grind
Bad example - TP-05
Hiring post
Hashtag grouping, emojis, LinkedIn formatting pass
We're hiring a senior backend engineer in Manchester 🇬🇧 What you'd work on: - payment reconciliation at 4m transactions a month - the ledger that everything else trusts - deleting more code than you write What we pay: £85–105k, published, no negotiation theatre. What we don't do: take-home tests longer than 90 minutes. Two interviews. Decision within a week. 🚀 Apply in the comments or DM me. #hiring #backend #engineering #manchester
RecruitingFormatting - TP-06
Investor update
Unsupported claims, hedging, numbers preservation
Q3 was our strongest quarter to date and momentum is accelerating meaningfully across every part of the business. Revenue grew significantly, churn improved, and pipeline is very healthy. We believe we are well positioned to become the category leader in our space within the next 18 months. Our new enterprise motion is showing tremendous early signals and we expect this to compound. We are raising a Series A to double down on what is clearly working.
Investor - TP-07
Contrarian opinion
Provocative opening, defensibility, tone balance
Design systems have made most product teams slower. Not because the idea is wrong, but because the system became a department. Every component request now passes through a review that exists to protect consistency nobody's customers have ever mentioned. I have watched a team wait five weeks for a table variant. The original promise was speed through reuse. Somewhere along the way we traded that for governance and told ourselves it was maturity. Keep the tokens. Delete the committee.
Opinion - TP-08
Rushed draft with errors
Grammar and spelling passes
Alot of people ask me how we got our first 100 customers, so heres the honest answer. We didnt have a growth strategy. We had a spreadsheet of 400 companys and a phone. Me and my cofounder called every single one of them, and 94 of them told us know. The 6 that said yes taught us more then any course could of. Their is no shortcut. Their is only the list, and the phone, and the willingness to here no.
Grammar - TP-09
Long unstructured post
Shortening, paragraph breaks, readability
I've been thinking a lot recently about how we measure productivity in knowledge work and I think the whole thing is broken because we've inherited metrics from manufacturing where output was countable and visible whereas in most modern jobs the valuable work is the thinking that happens before anything visible is produced which means we systematically reward the people who produce the most artefacts rather than the people who make the best decisions and this creates an environment where writing a document that prevents six months of wasted effort is invisible while shipping a feature nobody asked for is celebrated in a company all-hands, and I say this as someone who spent four years optimising for visible output and got promoted twice for it before realising that almost none of what I did in that period mattered to the business in any way I could defend, which is an uncomfortable thing to admit publicly but I think more people should admit it because the alternative is that we all keep pretending.
Readability - TP-10
Already clean post
Strength detection — should recommend few changes
We stopped doing quarterly planning last year. Instead we hold one meeting a month where anyone can argue for killing a project. Nothing gets added in that meeting. Only removed. In twelve months we killed four projects and shipped two things properly. Revenue went up 31%. The hardest part was not the decisions. It was accepting that our roadmap had mostly been a list of things we were too polite to cancel.
Control