@RicheloKillian: signal 82/100
Is this your account? Share your reading:
#1 of 190 accounts read, fewer needle hits than 100% of them. The leaderboard
@RicheloKillian is in Find
In Find, waiting for its first search
People search Find by what they want to read, and only accounts under 10K followers show up. Every search that shows @RicheloKillian is counted here.
Richelo Killian, the last 20 posts, read by six needles. A reading of the words, not a verdict on the person, and not a prediction of reach.
Every post, ranked
Best to weakest by overall signal. Pick a needle to rank the posts by that needle alone.
-
1 of 20 Strongest post
16views0likes0reposts0replies
Day 91 for and tonight the suite thesis stopped being a diagram. Big reveal: EmailWorx is live. It is the second SenderWorx product, a deliverability command center for businesses sending through a major email platform, and it went into full production a week ago after being battle tested with consulting clients first. The part that belongs on THIS account: every address check inside EmailWorx runs on IsItDisposable. The product this account has documented for 91 days, the one still reading $0 MRR below, is now infrastructure inside a bigger product. That was the plan from the first day of the suite: each tool feeds the next, the audience compounds, and the platform extraction we did in September is why product two shipped in weeks instead of months. And the distribution lesson, learned the hard way on product one, applied on product two: before promoting anything, I built a small internal tool that identifies companies in the EXACT target market. It found 58,000 of them. Not a persona, not a keyword audience. A list of named businesses with the precise problem this product solves, and the outreach will be slow, personal, one at a time, because that is the only motion that has ever produced a quality lead in 91 days of trying everything else. Product one taught the lessons. Product two inherits them all. $0 MRR on ISID. Day 91. The suite is real now. #buildinpublic
signalVague1%Nothing new30%Secondhand3%Nothing to use9%Bait or brag50%Hype16% -
2 of 20 57views1likes0reposts4replies
Day 88 for Since August I have had AI agents reading X around the clock, hunting for founders complaining about fake signups, trial abuse, and burner emails. Hundreds of leads scored and filed. Tonight, what three months of robot listening actually taught me about my market. 1. The pain is real but it is episodic. A founder tweets about spam signups the week it burns them, maybe twice a year. Across all of X, the genuinely in pain, catchable at the right moment founder appears a handful of times a week. The agents find them, and I reply to the good ones, but it is a trickle by nature, not a channel. 2. Most matches are the right PERSON on the wrong day. SaaS founder, runs signups, will absolutely hit this problem eventually, just not today. You cannot sell an umbrella by following people who will someday be rained on. 3. The real lesson is where the pain goes instead. People do not tweet about fake signups. They SEARCH about them, at 2am, the night the bot wave hits their form. Which is why the compounding bets, the articles and the data reports sitting in search results and AI answers waiting for that exact moment, were always the right channel, and the agents proved it by elimination. Listening at scale did not find me a pipeline. It found me the shape of my market, which turned out to be worth more. $0 MRR. Day 88. #buildinpublic
signalVague3%Nothing new35%Secondhand4%Nothing to use8%Bait or brag32%Hype11% -
3 of 20 48views3likes0reposts2replies
Day 78 for Tomorrow night, the commitment made on the record a week ago comes due: the October report goes to Hacker News. Attempt three, and the first one shaped like a research story instead of a product launch. The first two attempts were Show HN submissions from a new account with a commercial link, and the spam filters ate both within minutes, before any human read a word. Tomorrow's version is just a plain story: the report URL, a title stating the finding (one in 26 disposable email domains dies within 30 days of going live), and the product nowhere in it. If HN wants to argue with the methodology, the author will be in the thread. If the filters eat a third one, then the lesson is about the account, not the work, and the work keeps publishing monthly regardless. The mechanics of keeping this promise, for the record: it was announced publicly with a date, scheduled at 9:30pm with the submission pre-written down to the title, and the trade was named in advance, one hour of Wednesday morning sleep. That is the entire system for doing scary things: decide once, in public, while brave, and leave future you nothing to decide. $0 MRR. Day 78. Tomorrow we knock a third time. #buildinpublic
signalVague1%Nothing new42%Secondhand3%Nothing to use7%Bait or brag29%Hype5% -
4 of 20 25views1likes0reposts1replies
Day 92 for A thing I did not expect about wiring product one into product two: EmailWorx instantly became IsItDisposable's most demanding customer. Every trial user so far poked at the API politely and left. Real production traffic does not poke. It arrives with opinions. Within days of EmailWorx running live, ISID got pushed harder than ninety days of marketing ever managed, and the integration forced a design decision worth sharing. The decision: a trusted mailbox provider layer. When EmailWorx investigates an address, it checks its own trust list first, and the providers that will never be disposable, never reach ISID at all. Every check that DOES arrive is a genuine suspicion. The principle underneath generalizes to any API you build or buy: do not spend a lookup answering a question you already know the answer to. Filtering the obvious before the API call cuts cost, cuts latency, and keeps the interesting signal clean. Most integrations skip this and pay per check to confirm that Gmail is still Gmail. Dogfooding is usually described as using your own product. The stronger version is making your products use each other, because software is a much ruder customer than you are, and it files its complaints immediately. $0 MRR. Day 92. The suite argues with itself now. #buildinpublic
signalVague2%Nothing new44%Secondhand5%Nothing to use5%Bait or brag31%Hype12% -
5 of 20 11views0likes0reposts0replies
Day 89 for Tonight you nearly got a celebration post. Here is what happened instead. This week brought a trial signup that did what no signup in this product's history has done: configured API keys, integrated the endpoint, and made 25 real checks. Our first milestone, first qualified trial, is defined as exactly that, and I had the bell ringing post drafted. ALL CAPS and everything. Then I spent five minutes checking who he was before celebrating, and found his website: an API directory, with a freshly built page documenting our endpoint and pricing. The 25 calls were not a customer integrating. They were a cataloguer testing endpoints to write us up. So the milestone stays unrung, because the definition was always a trial that uses the product for its own purpose, judged by eye, and my eye says this is not that. The scoreboard only works if the referee cannot be bribed by his own hope. The consolation is real though: somewhere out there, a directory site tested our API thoroughly and is building an accurate third party listing of it, unasked. That is distribution, not revenue. I will take it, correctly labeled. Still zero qualified trials. Still telling you the truth first. $0 MRR. Day 89. Check before you ring. #buildinpublic
signalVague1%Nothing new45%Secondhand3%Nothing to use9%Bait or brag29%Hype6% -
6 of 20 92views3likes0reposts5replies
Day 84 for Something happened this weekend that is worth unpacking, because it is the build in public loop actually closing for once. Last night's post about our did_you_mean field, the one that catches jane@gmial.com and suggests is quietly the best received post this account has had. Multiple replies, and more than one person said some version of the same thing: that should be front and center on your sales page, not buried in an API response. They are right, and here is the uncomfortable admission: I have spent three months marketing this product as keeping bad signups OUT, because that is the category's framing. It took a Saturday night post and strangers' replies to surface that the rescue angle, keeping GOOD signups from being lost to typos, might be the half of the story that actually lands with people. Blocking is defensive and abstract. Saving a customer you already paid to acquire is money. So the hero copy is getting rethought this week with both halves in it. Ninety plus days of posting into the void, and the void finally started doing product marketing for me. Ask your audience nothing, show them everything, and occasionally they hand you your own positioning. $0 MRR. Day 84. #buildinpublic
signalVague2%Nothing new48%Secondhand3%Nothing to use8%Bait or brag37%Hype10% -
7 of 20 28views2likes0reposts1replies
Day 85 for Monday numbers, with a twist I have been waiting to use. The dataset crossed 228,763 tracked disposable domains today: 1,234 genuinely new ones in the last 7 days, 56 in the last 24 hours, across 685 distinct TLDs. Here is the twist. Because the October report published a strict 30 day mortality rate, those 1,234 newcomers arrive with a testable prediction attached: roughly 1 in 26 of them, call it 47 domains, should be dead within a month of first being seen alive. That is not a vibe, it is a falsifiable claim our own November edition will check against this exact cohort. That is the quiet difference between publishing statistics and publishing a series. A one off report describes the past. A monthly series makes predictions its own future editions can embarrass. Either the November numbers confirm the mortality rate and the finding hardens into something citable, or they diverge and THAT becomes the story, because a changing death rate in the disposable ecosystem would mean the providers changed behavior. We win either way, which is the only kind of bet research should make. $0 MRR. Day 85. The data is now grading its own homework. #buildinpublic
signalVague1%Nothing new32%Secondhand5%Nothing to use24%Bait or brag35%Hype9% -
8 of 20 29views2likes0reposts2replies
Day 80 for The Hacker News debrief, as promised, with the real numbers. The submission survived. After two attempts eaten by spam filters in minutes, the research-shaped third attempt sat on HN for a full day with human eyes on it. That alone took seven weeks and a complete reframing of what we were submitting. What it did: our best organic traffic day on record, 33 visitors in 24 hours against a baseline of about 3. What it did not do: catch a ranking wave, draw a single comment, or produce a signup. Thirty three strangers read the mortality research for about a minute each and moved on. Honest scoring: that is neither the front page dream nor the flag nightmare. It is the most common HN outcome nobody writes threads about, the quiet pass through. The report got read, the domain earned an HN story page that is permanent, and the methodology stood unchallenged, which I choose to take as a compliment nobody intended. The real lesson after three attempts: on HN, surviving is a prerequisite, not a result. Results need either luck or an audience that arrives on its own, and we are still building the second one, one monthly edition at a time. November's report will knock again. $0 MRR. Day 80. Doors open slowly, then all at once. Allegedly. #buildinpublic
signalVague1%Nothing new40%Secondhand2%Nothing to use20%Bait or brag42%Hype12% -
9 of 20 45views2likes0reposts2replies
Day 87 for Tonight, the design decision that costs us features and buys us sleep: we only ever log domains. Never full email addresses. When a check comes in for jane@mailinator.com, the local part is used in memory for normalization and the role account signal, then discarded. What persists is only that someone, sometime, checked That is the most sensitive fact our logs contain about anyone's traffic. The trade is real. Logging full addresses would let us offer per address history, dedupe analytics, retry tracking. Competitors sell some of that. We decided the price was wrong, because a database of every email address that ever crossed your customers' signup forms is not an asset. It is a subpoena magnet, a breach headline waiting for a date, and a trust liability that compounds silently while you sleep. The principle generalizes to any data you are tempted to keep: the only data that can never leak, never be demanded, and never embarrass you is the data you refused to store. Privacy by policy is a promise. Privacy by architecture is a fact. Every vendor in your stack has made this choice somewhere, and most will not volunteer which way. Ask them what their logs would show about your users. The pause tells you plenty. $0 MRR. Day 87. #buildinpublic
signalVague2%Nothing new55%Secondhand7%Nothing to use4%Bait or brag27%Hype8% -
10 of 20 12views0likes0reposts0replies
Day 77 for The sprint ended yesterday. Here is what happened this morning. I opened my task manager and the nightly post task was gone. So was the lead check, the evening check, the weekly review. For a second I thought something broke. Then I saw it: back in July, every recurring task was created with an end date of October 17. Day 90. The scaffolding was built to dismantle itself when the clock ran out, and at midnight it quietly did. Which turns today into a genuinely clean question. For ninety days the system made me show up. Nothing makes me show up tonight. The task that prompted this post did not exist until I recreated it this afternoon, by hand, with no end date this time. That is the difference between a sprint and a practice. A sprint borrows its discipline from a deadline. A practice is what you rebuild on purpose the day the deadline stops doing the work. The accounting thread from last night covers what the ninety days produced. This post covers what they installed. $0 MRR. Day 1 of no particular number. Still here. #buildinpublic
signalVague2%Nothing new51%Secondhand2%Nothing to use13%Bait or brag18%Hype4% -
11 of 20 47views2likes0reposts0replies
Day 90 for Not of the sprint, that clock ran out two weeks ago. Today is day 90 of the restart, which is the anniversary that actually matters. Ninety days ago I confessed, in public, that I had been faking progress ticks to my own accountability system. I quarantined every number, restarted the clock at zero, and made one promise: real numbers, every night, whatever they said. The ninety days since: every single night posted, one of them eight hours late after the day job emptied the tank, and that miss reported as honestly as the wins. Three launches attempted, two eaten by robots, one quiet survival. A revenue counter that has read zero dollars in all ninety posts and never once got dressed up as anything else. A milestone celebration drafted on Thursday and withdrawn before posting because five minutes of checking proved it premature. Here is what I know now that I did not know in July: the faking was never about the numbers being bad. It was about believing I could not be seen with bad numbers. Ninety days of being seen with them, and the audience that stayed is larger, warmer, and more useful than the one I was lying to. Appropriately for the date, the costume came off and the thing underneath turned out to be fine. $0 MRR. Ninety days honest. The streak is the product I am proudest of. #buildinpublic
signalVague1%Nothing new53%Secondhand2%Nothing to use13%Bait or brag38%Hype4% -
12 of 20 50views2likes0reposts6replies
Day 83 for Back on schedule. Saturday, so here is a small feature story nobody ever markets. Our API response has a field called did_you_mean. Send it jane@gmial.com and it answers with A typo detector. It gets no landing page, no launch tweet, and it might quietly be the highest ROI field in the response. Here is why. A typo of a real domain is the opposite failure to a burner. The burner is a fake person you accidentally let in. The typo is a REAL customer you are about to lose: their confirmation email bounces, they never activate, and your ESP logs a hard bounce that counts against you exactly like a dead disposable domain does. Same bounce, opposite tragedy. So the same one API call does both jobs at the form: catches the throwaway pretending to be a customer, and rescues the customer about to be lost to a fat thumb on a phone keyboard. One suggests a block. The other suggests a gentle did you mean under the input box, which costs nothing and saves a signup you already paid to acquire. Detection is not only about keeping the wrong people out. Half the value is keeping the right ones in. $0 MRR. Day 83. The chain resumes. #buildinpublic
signalVague2%Nothing new44%Secondhand21%Nothing to use6%Bait or brag36%Hype18% -
13 of 20 96views1likes0reposts2replies
Day 82 for posted about eight hours late, and that is the story. Last night, for the first time since August 3, I simply forgot the nightly post. Not a crisis, not a collapse, no dramatic reason. The acting executive role at the day job is in its fourth week, Thursday emptied the tank, and the one link in this whole machine that never got automated, me remembering to show up and say what happened, quietly failed. Worth being precise about what that reveals. The posts draft themselves, the scheduler ships them, the floors count themselves from the database, the sweep has not missed a day since August. But the trigger for all of it is still a tired human sending one message at 8pm. Eighty one days it held. Night eighty two it did not. I am not restarting any counters over it. The fake ticks in July were a lie; this was a nap. The difference matters, and an accountability practice that cannot tell the two apart teaches you to hide instead of to rest. So: Day 82, late, honest, and the chain continues tonight at the usual time. $0 MRR. Still here. Occasionally on a delay. #buildinpublic
signalVague1%Nothing new55%Secondhand2%Nothing to use18%Bait or brag17%Hype2% -
14 of 20 28views0likes0reposts4replies
Day 93 for Monthly instrument check on the SEO bet, published because nobody ever shows you this part. October's search numbers: 14 Google impressions. Zero clicks. Average position improved to 6.1 from 7.3, meaning the few queries we appear for rank page one, but the queries are so low volume the impressions barely register. Bing's window reads effectively zero. Four months of publishing, 15 plus articles, 3 data report editions, 250 programmatic pages, and organic search is still a rounding error. I am posting this flat month for the same reason I posted the flopped launches: the compounding content strategy was always priced in months, not weeks. Every SEO case study you have ever read starts the graph at month six because months one through five look exactly like this, a flat line with a slightly improving position average, while the index slowly digests a young domain. If the line is still flat at month nine, THAT is a verdict. At month four it is just the boring part of the curve, and the discipline is continuing to feed it without staring at it. Thursday the November data report publishes, the first edition that scores its own previous prediction. The flywheel gets fed on schedule, flat line or not. $0 MRR. Day 93. The boring middle, documented. #buildinpublic
signalVague1%Nothing new62%Secondhand3%Nothing to use13%Bait or brag20%Hype6% -
15 of 20 36views3likes0reposts4replies
Day 86 for Quiet product day, so here is the crossover report nobody asked for: what four weeks in an acting executive seat at the day job has taught me about running a one person SaaS. 1. Delegation and automation are the same skill. At work I hand things to people and trust the process. At night I hand things to pipelines and agents. The failure mode is identical: holding on because nobody does it exactly like you. The fix is identical too: define done, check the output, let go. 2. The weekly wrap up I send upward and the nightly post I send here are the same document. Plain numbers, what moved, what stalled, what I need. If you cannot write that in ten minutes, you do not know your own operation. 3. Budgets taught me the deliberate no long before the sprint did. Every dollar not spent on a bad channel is a dollar still available for a good one, and nobody audits the disasters you declined. 4. The big org runs on fail open thinking too. Every resilient system I manage degrades gracefully instead of breaking loudly. I knew this professionally for years before I made it my product's first principle. The day job and the side business keep teaching each other. Neither knows. $0 MRR. Day 86. #buildinpublic
signalVague3%Nothing new68%Secondhand5%Nothing to use8%Bait or brag46%Hype7% -
16 of 20 12views0likes0reposts1replies
What I learned, compressed: Every quality lead came from an artifact or a one to one conversation. Zero from broadcast. Floors beat goals. Dawn beats evening. Public dates beat private intentions. Instruments lie until they read the source of truth. And a 90 day revenue goal was the wrong frame for a compounding business, which I can prove, because the compounding is now visible and the revenue is not. Yet.
signalVague8%Nothing new54%Secondhand11%Nothing to use11%Bait or brag50%Hype21% -
17 of 20 34views0likes0reposts2replies
Day 81 for Quiet day, so here is the design decision I would defend in front of a firing squad: the API fails open. If our service is slow, down, over quota, or overloaded, the response is still HTTP 200 with action allow. Your signup form never breaks because of us. Not once, not for a second, by architecture. This sounds obvious until you price the alternative. A validation service that fails closed turns its own outage into YOUR outage, and a signup form that errors is worse than one that lets a burner through. The burner costs you a fraction of a cent in list hygiene. The broken form costs you every real customer who tried to sign up during the incident, silently, with no error report, forever. The uncomfortable part nobody says out loud: fail open means admitting your product is not important enough to break someone's business over. That is exactly right. Detection is a layer, not a gate. The moment your vendor's availability becomes your conversion rate, you bought a liability, not a tool. Ask every vendor in your signup path what happens when they go down. The ones who answer slowly are the ones to replace. $0 MRR. Day 81. #buildinpublic
signalVague3%Nothing new52%Secondhand25%Nothing to use5%Bait or brag26%Hype13% -
18 of 20 15views0likes0reposts2replies
Day 79 for Short one tonight, by design. The Hacker News submission announced last week went in this evening as planned. That is all I am saying about it until tomorrow: no link, no play by play, not even whether it survived the first hour. Partly because HN rightly penalizes submissions that arrive with an audience, and partly because I have learned not to narrate a result before it exists. The commitment is kept either way, and that was the part I controlled. Full details tomorrow, whichever way it went. $0 MRR. Day 79. The knock is knocked. #buildinpublic
signalVague3%Nothing new74%Secondhand4%Nothing to use19%Bait or brag52%Hype3% -
19 of 20 29views0likes0reposts2replies
So: $15K target, $0 result, and the honest verdict is that the sprint failed at its stated goal and succeeded at its actual function: it forced a complete go to market machine into existence and taught one builder to stop hiding. The sprint ends today. The business does not. Monthly reports continue, the HN submission lands next week, the marketplace door is one approval away, and the first paying customer is still out there, about to meet a very patient machine. Thank you for watching the honest version. It continues. #buildinpublic
signalVague3%Nothing new64%Secondhand4%Nothing to use29%Bait or brag47%Hype11% -
20 of 20 Weakest post
44views2likes0reposts3replies
Day 76 of the restart, Day 90 of the original clock, for The full accounting went up on the timeline an hour ago: nine posts, every number verified against the systems that produced it, including the zero that headlines it. If you have followed this account for one post or all seventy six, that thread is what it was all building toward. I will not summarize it here; it does not deserve summarizing, it deserves reading. What I will say here is the part that did not fit in an accounting. Ninety days ago this was a product with no marketing and a founder who hid from the uncomfortable parts. Tonight it is a machine that publishes research the industry never had, a founder who schedules courage for dawn because he learned he had to, and an audience that watched every flop land and stayed anyway. You cannot put that on a revenue chart. It is still the most valuable thing the ninety days made. The nightly posts continue, because the chain is the asset and the business did not end, only the clock did. $0 MRR. Day 90 of 90. See you tomorrow anyway. #buildinpublic
signalVague2%Nothing new66%Secondhand4%Nothing to use45%Bait or brag62%Hype37%
Read on 2026-11-04. This is your account and you want it gone? Write to hello@postneedle.com. The leaderboard