Why most build-in-public efforts stall by month three
Build-in-public is two practices in a trench coat. The first is broadcasting — telling the internet what you are working on, what you decided, what broke, and what the numbers look like. The second is listening — reading what the internet says back, in replies, in quote-posts, in DMs, in screenshots that get reposted somewhere you do not follow. Most operators do the first half exclusively. That is why their build-in-public effort plateaus around month three with warm peer applause and almost no pipeline.
The peer applause is the trap. You ship a thread about a hard decision, three other founders you respect quote it with the rocket emoji, your follower count creeps up by 40, and the dopamine arrives on schedule. None of those people are buyers. They are colleagues with their own products, watching with professional courtesy. The buyers — the people who would pay you — are reading silently from accounts with five followers and avatars of their dogs, and they will never speak unless you give them a reason to. The reason is almost always a reply. They drop a one-line question in the thread, you ignore it for six hours, they sign up for a competitor. The deal you never closed is the deal you never knew was on the table.
The replacement framing is build-in-public is a conversation, not a feed. Every post is a question your audience can answer; every reply is the answer arriving; every reply you ghost is a buyer telling their friend you are unresponsive. The post is 20% of the work. The 80% is the reply, the follow-up DM, the unlinked-mention thank-you, the quote-post you would not have seen if you were not watching for it. This playbook is the operational manual for running both halves with discipline so the program compounds instead of plateauing.
A useful sanity check before you read further. If your last ten build-in-public posts produced more peer reposts than buyer DMs, you are running broadcasting-only. If your last week of replies took less than two hours of your time, you are running broadcasting-only. If you do not know who screenshotted your last revenue post and where it ended up, you are running broadcasting-only. None of those are fatal — but the next ninety days of effort are wasted unless you fix it. The rest of this guide is the operating manual; for the broader startup-stack context the practice slots into, see the ReplySocial-for-startups breakdown where build-in-public is one of four channels in the founder-led playbook.
What counts as “in public” and what is performance
Build-in-public has been diluted into a hashtag, and the hashtag now wraps an enormous amount of content that is not building in public — it is announcing in public, performing in public, or grinding-content in public. The distinction matters because the audience can tell the difference even when the operator cannot, and the conversion math only works on the genuine version. The following table maps the three most common failure modes against the real practice.
Daily “day 247 of building in public” counter posts with no actual update.
Daily reading habit; weekly substantive update with a number, decision, or failure attached.
Why. The counter post is a streak, not a transmission. Streaks broadcast discipline; transmissions broadcast learnings. Buyers are looking for the second.
“Honest moment — I'm exhausted today and considering quitting.”
“Three weeks of failed onboarding A/B tests. Here are the variants we ran, here is the metric that broke each one, and here is what we are trying next.”
Why. The first is performance of vulnerability and produces zero useful information. The second is operational candor and produces a teachable artifact. Buyers convert on the second; peers cheer for the first.
“We hit 10K MRR! Wild journey, thanks everyone.”
“We hit 10K MRR. The unlock was killing the free trial after eight months of optimizing it; here is the conversion data from before and after, and here is the one objection it removed.”
Why. The bare milestone is for the founder's group chat. The annotated milestone tells everyone reading what you actually learned, and the learning is what compounds into category authority over a year.
The shorthand: every build-in-public post should answer the question “what would a smart, busy operator learn from this if they only had thirty seconds?” If the answer is “nothing, just that you exist” — do not post it, or rewrite it until it is. You are not running a journal in public. You are running an open notebook of operating decisions that other operators can learn from and that buyers can use to assess whether you are competent enough to entrust with their problem.
The three audiences any build-in-public lap reaches
Every public update lands on three distinct audiences at once, and each behaves on a different time scale, with a different conversion shape, and with a different reply pattern. Most operators write for the first audience exclusively because the first audience is the only one that responds visibly. That is also why most operators eventually quit the practice — the visible response curve is bounded above by the size of their peer network, and the invisible curve is the one that actually pays.
How they behave. Reply visibly, repost, drop the rocket emoji, sometimes write the canonical reply that defines the discourse around your post for the next 48 hours. They are the first 80% of any reply count.
Conversion shape. Almost zero direct conversion. Peers occasionally turn into customers when their company hits the right size, but the lag is 12-36 months and the trigger is rarely your post itself.
Operating tell. If your replies are 90% peers and your DMs are sparse, you are over-indexed on peer-resonant content (decisions other founders relate to) and under-indexed on buyer-relevant content (decisions that show your category competence).
How they behave. Read everything; rarely reply in public. When they do reply, the reply is a one-line technical question, often without context. They DM more than they post. They screenshot your numbers and forward them to a colleague who is not on the platform.
Conversion shape. The dominant conversion channel for build-in-public, but invisible at the surface. Lag of 2-12 weeks from a post to a signup, often via a path that does not include direct attribution. The only reliable measurement is brand-search lift and self-reported source on signup.
Operating tell. If your DMs include questions that betray category-specific knowledge — “does it support webhooks?”, “how does pricing scale past 50 seats?” — your buyers are reading. Ignoring those DMs for six hours is the most expensive procrastination in your week.
How they behave. Critics write specific objections to specific claims, often citing data you did not include. Trolls write generic dismissals (“mid”, “cope”, “this guy thinks…”) and disappear.
Conversion shape. High-information critics flip into the strongest advocates when their objection is engaged honestly — see vignette 02. Low-information trolls never convert and engaging with them is a tax on your time and a poison on your replies; mute, do not reply.
Operating tell. The critic test is whether the objection cites a number, a workflow, or a specific failure mode. If yes, treat it as the most important reply you will write that week. If the objection is vibes, mute the account and move on. The cost of confusing the two is high in both directions.
The operating implication: write for audience 02 (prospects) explicitly while audience 01 (peers) reads over their shoulders. A post that names a real workflow, cites a real number, and admits a real limit will convert prospects and entertain peers; a post that names a vague decision, cites no numbers, and claims everything is going great will entertain peers and convert no one. The two are not in tension; they are aligned along the axis of operating specificity. Write specifically and the right people read.
The asymmetric content mix: what compounds and what wastes effort
A working build-in-public mix is not balanced. Roughly 70% of the value comes from four post shapes, and the other 30% is the connective tissue (announcements, questions, light-touch updates) that holds the cadence together. Operators who try to balance their cadence equally across all post types end up writing more announcements than analysis and watching their compounding stall. The rule of thumb is to over-invest in the four shapes below and let everything else fill in around them.
A decision you made, the alternatives you considered, the constraint that decided it, and what you expect to learn. These compound because they are recyclable: when the decision pays off, you cite the original post; when it backfires, you write a follow-up that cites it; either way the post becomes part of an arc.
A specific thing that broke, the timeline, the metric you watched, the fix or the abandonment. These compound because they are the rarest content category in the genre — most operators publish triumph and bury failure. The post-mortem reads as proof of operating maturity and is disproportionately bookmarked, screenshotted, and quoted in podcasts months later.
MRR, churn, conversion rates — but always with the shape of the curve and the operating decision attached. “MRR is up 40% MoM” is noise; “MRR is up 40% MoM, the unlock was killing the free trial, here is the conversion data” is signal. Numbers without shape compound nothing because they are uninterpretable.
An A/B test, a pricing change, a positioning shift — with the hypothesis stated up front, the metric defined, the result reported even if the experiment failed. These compound because they create a reputation for actually running the experiment loop, which is rarer in the wild than it looks on Twitter.
“Day 247.” The streak is for you, not for your readers. After day three the marginal information content is zero and the post takes up a slot a real update could have filled. Run streaks privately if they help your discipline; do not publish them as their own content.
“Tired today but pushing through.” “Big day ahead!” They perform vulnerability or determination without transmitting any operating information, and they train your audience to skim past you. The audience that bookmarks your decision posts is the same audience that mutes accounts running 50% mood posts.
“Keep building.” “Consistency wins.” The motivation industry already covers this beat with people whose job is producing it; you are not adding marginal value, and the posts dilute the signal of the substantive ones around them. Buyers who land on your profile and see motivational quotes scroll on.
The temptation to weigh in on the AI debate, the framework war, the pricing discourse — all platform-natural content that drives short-term engagement and long-term distortion. Your account becomes legible as a generic commentator rather than the operator of a specific product, and the prospects who would have converted scroll past assuming you are a content account.
One operating constraint to add to the mix: never share things that betray a customer or a teammate. Customer-specific anecdotes need explicit permission and are usually better published as a case study; internal disputes are never public-facing, even when you are right. The rule is “public about my decisions, private about other people’s situations.” Crossing it once permanently changes how your team and your customers relate to your account, and the change does not reverse.
Six illustrative vignettes
The six vignettes below are illustrative composites — names and numbers are invented to convey shape, not data. They cover the full distribution of build-in-public outcomes from quiet compounding wins to loud public failures, because the failure cases are the lessons most build-in-public posts pretend do not exist.
A solo developer ships a single-purpose SaaS in a niche category and posts twice a week — one decision-with-rationale, one numbers-with-shape. They reply to every DM within forty minutes for the first six months. By month nine they pass $8K MRR. The follower count is only 1,400, but the conversion rate from follower to trial is roughly 4%, an order of magnitude above the platform average.
What surprised them. The single highest-converting post of the entire run was a 70-word failure post-mortem about a botched migration. It drove 31 trial signups in the week after publication; the celebratory milestone posts on the same account drove 2-3 each.
Lesson. Followers are a vanity metric; conversion-per-follower is the metric that pays. A small audience of buyers beats a large audience of peers by an order of magnitude. The way you build the buyer-skewed audience is by writing for buyers — operating specificity, not motivational vagueness.
An open-source maintainer publishes an architectural decision thread on X. A senior engineer at a competing project replies with a detailed objection citing two failure modes the post had not addressed. The maintainer rewrites the post into a long-form blog reply, links to it, and credits the critic by handle. The critic shares the response post.
What surprised them. The critic introduced the maintainer to the engineering team at a Fortune 500 nine months later. That introduction became the first $40K enterprise contract for the open-source company. The thread had ~80 likes; the conversion path was invisible to any standard attribution model.
Lesson. High-information critics are the highest-leverage audience to engage. The cost is one careful long-form reply; the upside, sometimes, is a relationship that closes deals years later. Never mute the specific objector. Always mute the generic dismisser.
A funded founder posts daily streak counters and weekly milestone announcements for fourteen months. Follower count climbs from 1,200 to 18,000. Trial signups attributable to the practice in the same period: roughly 60. The team kills the program after a board meeting where the build-in-public ROI is challenged.
What surprised them. A post-mortem audit of the period showed the founder had ignored ~340 DMs, including 28 that were direct buyer questions about pricing or specific features. The dominant cause of the low conversion was not the content; it was the unanswered inbox.
Lesson. A build-in-public program that does not staff the reply layer is a marketing-spend hole. Either commit to thirty-minute reply discipline, or do not run the program at all. The middle option — broadcasting daily and replying when convenient — is strictly worse than not posting.
A two-person team posts a candid revenue chart with a one-line note about a positioning change that drove the inflection. A larger creator screenshots it and posts the screenshot to their own audience. The screenshot has no link back to the team's account.
What surprised them. The team only learned about the screenshot four days later when a new customer mentioned it on an onboarding call. By then the screenshot had been quoted into two newsletters and reposted on LinkedIn twice. The total impressions of the unlinked variant exceeded the original post by roughly 40x.
Lesson. The most viral version of your post often does not include your handle. If you are not running unlinked-mention monitoring, you are missing the largest amplification events of your build-in-public program in real time. The first reply window on those events drives an outsized share of the resulting signups.
An indie founder publishes a candid thread describing a frustrating support interaction with a specific customer. The thread is detailed enough that other customers identify the subject. Two churned within the week; the original customer posted a public response that ranked above the founder's account in search for the brand name for several months.
What surprised them. The most-engaged reply on the original thread came from a competitor's CEO offering empathy in a way that quietly redirected three prospects in the founder's pipeline into a competitor evaluation. The empathy was real; the strategic effect was not.
Lesson. Public about your decisions, private about other people's situations. There is no version of leaking customer specifics that does not compound against you over a year, even when the venting feels justified in the moment. If you need to vent about a customer, vent in your group chat. The internet remembers.
A pre-launch founder spends six months posting decision threads about the problem space they are working on, with no link to anything because there is nothing to ship yet. They reply to every DM within thirty minutes. By launch day they have ~1,800 followers and a list of 240 self-tagged early-access requesters they harvested from replies and DMs.
What surprised them. On launch day, 67% of the early-access list converted to paid within seven days — a conversion rate roughly five times higher than the same founder's post-launch acquisition channels delivered for the next year. The pre-launch reply work, not the launch post, did almost all of the conversion.
Lesson. Build-in-public is a relationship factory more than a content factory. The replies and DMs in months 1-6 build a list of pre-qualified, pre-warmed prospects that no paid channel can match for cost. Operators who launch before they have done the relationship work are launching to strangers.
The pattern across the wins and the failures is the same one section 1 named: the operators who win run both halves — broadcasting and listening — with discipline. The operators who fail run only the broadcasting half, get tired, and conclude the practice does not work. It works. They were running half of it.
The reply discipline most operators skip
The reply layer is where build-in-public converts, and it is the layer almost nobody runs at a professional level. The operating shape is a thirty-minute first-reply window after every post, a daily inbox triage, and a weekly mention sweep that catches the conversations you would otherwise miss. Each of the three has a specific failure mode if skipped, and skipping any one of them silently caps your conversion rate at a fraction of what the program could deliver.
The thirty-minute first-reply window
The first thirty minutes after a post is when the early replies arrive, the algorithm decides whether the post is worth amplifying, and the prospects who saw it in their feed make a one-tap decision about whether to engage further. Replying inside that window — to every substantive comment, with at least one full sentence — does three things at once. It signals to the algorithm that the post is generating discussion (which earns more reach), it gives early commenters a reason to keep watching the thread (which earns more replies), and it gives prospects a low-friction proof point that the operator behind the account is responsive (which earns trust). Skip the window and the post hits roughly half the reach it could have, permanently, on a curve you will not see in the analytics.
The daily inbox triage
Once a build-in-public account passes ~1,000 followers, the DM and mention volume begins to exceed what a casual scroll-and-respond habit can handle. The replacement is a single 20-30 minute block per weekday where you triage the day's inbound — buyer-shaped DMs first, peer DMs second, mentions third, the rest into mute or follow-up. A unified inbox across X, LinkedIn, and any other channel where you publish is the only sustainable form of this; the alternative is forty open tabs and a missed deal a week. The operating shape is exactly the one in the social inbox use-case — replies, mentions, and DMs in one queue, triaged in one pass — and the failure mode is obvious only in retrospect, when you discover three buyer DMs from last week with no answer and they have already signed up for a competitor.
The weekly mention sweep
Mentions on social platforms come in two flavors: the linked variety (where someone tags your handle and the platform notifies you) and the unlinked variety (where someone names your product, screenshots your post, or quotes you without the tag — and the platform notifies nobody). The unlinked variety is where a meaningful share of the high-leverage moments live, because the people who screenshot your charts, the journalists who quote you in a piece, and the creators who repost your numbers to a different audience often do not bother to tag the source. Catching the unlinked variant requires keyword monitoring on your product name, the unlinked-mention variants of your name, and the URL of your most-shared post — set as standing monitors that surface a hit any time your name lands in front of a new audience. Pair that with mention tracking for the linked variant, and you stop missing the conversations that drive most of the uncaptured signups in your program. A reasonable cadence is a fifteen-minute sweep every Friday plus instant alerts for a small number of high-priority phrases (your founder name, your category-defining phrase, any URL you have published in the last 30 days). For the LLM and viral-thread era specifically, expect a rising share of replies to your posts to be generated by bots and reply-farms — filter them ruthlessly so the signal in your inbox stays human.
A reply taxonomy that scales
Once volume is real, you need a small set of reply patterns you can deploy in seconds. Not templated text — that reads like a marketer pretending to be a peer — but a mental taxonomy of three or four shapes that cover ~80% of the inbound. Buyer-shaped technical question? Answer in one paragraph and offer a DM if it goes deeper. High-information critic? Engage in long form, in public. Generic peer applause? Heart-react and move on. Generic dismissal or troll? Mute, do not reply. The reply-template tool is useful as a starting library, but the goal is to internalize the patterns until you can execute them at speed without looking. A practiced operator can clear thirty replies in twenty minutes; an unpracticed one takes ninety minutes for the same volume and gets fewer of them right.
Tooling: what to use and what to avoid
The build-in-public tooling stack splits into four jobs: writing, monitoring, replying, and measuring. Each has a small set of credible options and a longer list of tools that are sold to indie hackers and that quietly degrade the program when adopted. The list below is the working stack in 2026 plus the categories of tool to refuse on sight.
What to use
Writing — drafts and threading
A lightweight thread composer is plenty. The X-native tools are fine; if you publish long threads regularly, the dedicated thread editors save real time. We have a breakdown on Typefully as a writing tool if you are evaluating that category specifically. Most builders over-invest here — the writing is 20% of the work — and the marginal hour is better spent on the reply layer than on a fancier composer.
Monitoring — your name, your post URLs, your category
A unified inbox across X, LinkedIn, Reddit, and Facebook that watches for keyword variants of your product name, founder name, and the URL of your last few important posts. This is the single highest-ROI tool in the build-in-public stack because it surfaces the unlinked-mention moments you would otherwise miss. Pair keyword monitoring with LinkedIn inbox monitoring because LinkedIn build-in-public is now growing faster than X build-in-public on a per-impression basis and most operators are not watching it yet.
Replying — a single triage queue across channels
The unified-inbox shape from the social-inbox playbook is the right shape: replies, mentions, and DMs from every channel queued for triage in one pass. The alternative — switching tabs across native apps — burns more attention than the replies themselves and is the operating reason most build-in-public reply discipline collapses by month three.
Measuring — analytics that show shape, not just totals
X and LinkedIn natively report impressions and engagement; the limit of those reports is that they show totals without the curve. A purpose-built X analytics view that surfaces the long-tail compounding of decision posts (which keep accumulating impressions for weeks) versus the immediate spike of milestone posts (which exhaust in 48 hours) is the more useful artifact for tuning your content mix. Pair it with the profile audit early on so the bio, pinned post, and link strategy are not silently dragging your conversion rate down before any of the content work even runs.
What NOT to use
Each of the four categories below is sold specifically to the indie-hacker and build-in-public crowd, and each one corrodes the practice in a specific way. Refuse on sight.
AI thread generators that promise “viral X threads in seconds”
The output is detectably AI within two sentences and reads as the same generic thread that 5,000 other accounts published the same week. The deeper failure is structural: the entire reason build-in-public works is that you are publishing your specific operating decisions, and an AI cannot generate decisions you have not actually made. Use AI for outlines and edits; write the actual content yourself.
Reply-bot services that auto-reply to your followers' posts
The promise is “maintain engagement at scale”; the reality is that the replies are generic, off-topic, and visibly bot-generated within a week of going live. The cost is that your account becomes legible as a bot account to the very prospects you are trying to convert. Replies are the thing you cannot outsource without destroying the practice. Hire a part-time community person before you adopt a reply bot.
Fake-screenshot and clout-farming tools
Tools that generate fake X profile screenshots, fake DM screenshots, fake revenue dashboards. The screenshot moves on the platform for a week and gets dunked on for a year when someone notices. Build-in-public works because of the implicit credibility contract; faking the artifacts breaks the contract once and you do not get it back. The downside is asymmetric and there is no version where the upside is worth it.
Generic publishing-first schedulers as the central tool
A scheduler-first stack treats build-in-public as broadcasting, which is the failure pattern this whole playbook is correcting. We have detailed comparisons of the major schedulers (Buffer, Hootsuite, Later) on the alternatives directory if you need one — see the Buffer alternative breakdown in particular for the publishing-first vs. monitoring-first distinction. Use a scheduler if you must, but do not anchor your stack around it; anchor around the monitoring and reply layer instead.
The 30-day starter plan + measurement
A working build-in-public habit can be installed in thirty days. The plan below assumes one operator (founder, head of marketing, or solo developer) and zero new headcount. Each week is roughly four hours of focused work plus a daily 30-minute reply window starting in week 1. The plan is intentionally front-loaded with infrastructure so the content work in weeks 2-4 runs on rails rather than fighting tooling.
Infrastructure: monitoring, inbox, and the bio
Set up the monitoring layer first. Connect the channels you actually publish on, set keyword monitors for your product name (and obvious misspellings), your founder name, your category-defining phrase, and the URLs of any posts you have published in the last 30 days. Audit the bio, pinned post, and primary link with the profile audit — the conversion rate from a curious profile visit to a trial sign-up is gated by the profile, not the content, and most build-in-public accounts have not touched their bio in a year. Establish a recurring 30-minute end-of-day inbox triage block on the calendar.
Monitoring live with at least 5 active queries. Bio, pinned post, and primary link audited and updated. Inbox triage block on the calendar daily. ~4 hours of work.
If you set the queries but no hits arrive in 48 hours, the queries are too narrow. Add unlinked-mention variants of your product name, your category-defining phrase without the brand, and the URL of your most-screenshotted post. Most builders under-spec the query set on the first pass.
Cadence: ship five substantive posts and reply within thirty minutes
Ship one substantive post per weekday — at least three from the four compounding shapes (decision-with-rationale, failure post-mortem, numbers-with-shape, experiment writeup), no more than two announcements or questions. Reply to every substantive comment within thirty minutes of publication. Daily 30-minute end-of-day inbox triage; do not skip a day. By end of week, you should have 5 posts shipped, ~30-50 replies written, and a baseline of which post shapes drove the most prospect-shaped DMs.
5 substantive posts shipped, 80%+ of substantive replies answered inside the 30-min window, daily inbox triage running, baseline of which post shapes drove the most prospect DMs.
If reply latency is slipping past 60 minutes, the issue is calendar — not motivation. Block the 30 minutes immediately after each post on the calendar before you publish, and treat it as a meeting you cannot skip. The reply window is non-negotiable; everything downstream depends on it.
Substance: first failure post and first decision-rationale post
Many build-in-public accounts never publish a real failure post. This is the week to do it. Pick one specific thing that broke (an experiment that failed, a feature that was abandoned, a positioning attempt that did not land), write a 200-400 word post-mortem with the timeline, the metric you watched, and what you are doing next. Pair with one decision-with-rationale post for the next decision on your roadmap. Use the post formatter to keep the structure tight on threads. Continue the daily reply discipline.
1 failure post and 1 decision post shipped. Both with at least 3 substantive replies that turned into a follow-up exchange. Continued daily reply window.
If the failure post feels too risky to publish, pre-share it with one trusted operator and ask them to flag anything that crosses the “public about decisions, private about people’s situations” line. The risk almost always reads smaller from outside than from inside, and the upside on a real failure post is the largest of any post shape in the genre.
Measure, prune, and decide what graduates to month two
Pull the numbers. Posts shipped per shape; replies written; mentions caught (linked vs. unlinked); DMs received from prospect-shaped accounts; brand-search lift in Google Trends or Search Console; trial signups attributable to the channel (referrer + the self-reported “how did you hear” field). The two metrics that matter most are prospect-DMs-per-week (the leading indicator of pipeline) and the share of unlinked mentions you caught (the leading indicator of program health). Compare your share-of-voice to the top three accounts in your category with the share-of-voice tool so you have a baseline for month two. Decide which post shapes graduate to ongoing cadence and which to drop.
Month-1 metrics in a sheet. Share-of-voice baseline captured. Decision made on which 2-3 post shapes graduate to ongoing cadence. Working hypothesis on which post shapes drove the most prospect DMs.
If month-1 attribution is ~zero, do not bail. Build-in-public returns compound on a 6-12 month horizon; the vignette 06 founder did six months of pre-launch work to convert 67% of an early-access list on day one. Tighten your queries, tighten the post mix toward the four compounding shapes, and stay in. Quitting at month 1 forfeits the largest part of the return.
What to measure beyond the 30 days
Direct attribution understates build-in-public the same way it understates Reddit and podcasts — the dominant conversion path is dark, with a long lag and a self-reported ground truth. The reliable signals on a 90-180 day horizon are brand-search lift, the ratio of prospect-DMs to peer-DMs over time, the count of unprompted product mentions by third parties, and the conversion rate from new follower to trial signup. None of those appear in a one-month window with statistical significance, and chasing them weekly is how programs get killed prematurely. Set a 90-day review on the calendar, write the retrospective in public, and let the practice compound. The audience you are building is the one that will buy from you in year two; year one is the deposit.
One last frame to leave with. Build-in-public is not a marketing channel; it is the operating practice of working with the lid off. The marketing returns are a side effect of doing the work in a way that makes the work legible. If the practice is working internally — decisions are getting better because you have to defend them in public, failures are getting smaller because you have to publish them, the team is calibrating faster because everything is on the record — the marketing returns will arrive. Skip the operating side and the marketing side eventually flatlines no matter how many threads you ship. Show up weekly, write specifically, reply within thirty minutes, and let the year do its work.
