In-niche resident. A blue-tick X Premium account with two to five years of posting history in a single crypto sub-vertical (L1/L2, DeFi, DePIN, AI subnets, stables, RWA, memecoins). Not a fresh handle spun up for the launch. The age plus the niche pattern is the part that lets the resident's reply read as a normal slice of a normal feed, instead of as a coordinated drop. We do not run single-purpose accounts because they get flagged by listing scoring teams in the first pass and they do not survive the next account-quality sweep on X.
Pre-TGE warm-up. Three weeks is the practical floor for a clean warm-up before a TGE. The first week the pool moves into your sub-vertical and starts daily takes on your category. The second week residents cross-engage and the cluster signal starts reading as organic. The third week the warming converts into seeded conversation density that an algorithm and a holder both read as real category momentum. Anything shorter than three weeks is news-cycle work, not launch work, and we say so on the call.
Listing review. Most CEXs and launchpads score a project's socials during their review window. They look at four signals: follower count and growth rate, engagement rate across posts, mention density across category threads, and recent posting cadence. Thin signals on any of the four become a reason to delay the application, downgrade tier or ask for stronger market-making concessions. A 14 to 21 day pool cycle moves all four metrics into the band reviewers expect. We hand over an audit log the team attaches to the listing application instead of reverse-engineering it later.
The 5-minute SLA on news beats. When a category news beat lands (exchange listing for a competitor, audit drop, regulatory shift, hack post-mortem), the standing pool drafts and queues 8 to 18 touches inside five minutes. Each touch frames the news through your seeded narrative instead of through whatever generic take is winning the timeline that hour. The point is not to flood the feed. The point is that the first reframe a reader sees on a category news drop is yours, not someone else's, and that the reframe carries a recognisable voice.
Touch vs reach. A touch is one visible appearance of your project in a live conversation: an original post from a resident, a quote-tweet of a token announcement, a reply under a category influencer, a mention inside a trending thread. Reach is the cumulative impressions across all touches plus cross-engagements. We report both because touches measure the work and reach measures the surface, and a campaign that ships 4,000 touches into a dead niche is different from one that ships 4,000 touches inside a live category cluster. Both numbers go on the weekly review.
The Review loop. Every reply a resident drafts in the first week lands in a queue for your team with full thread context, the trigger that fired and the resident's proposed copy. Most teams approve over 90% of items once the brand card is calibrated, usually inside 7 to 10 days. The loop is not there to slow the pool down. It is there to make brand safety on thousands of monthly touches operationally achievable without a 24/7 review desk on your side. After week two, teams usually move to a weekly batch review of edge-case items only.
Brand card. The single document that gates resident output. It defines tone, on-brand talking points, do-not-touch topics (price predictions, KOL beef, named competitors, regulatory speculation), how to handle FUD inside a thread, and the legal posture for your sub-vertical. Updated monthly with the team. The brand card is the difference between a pool that reads as a coordinated drop and a pool that reads as a normal slice of category conversation, and most onboarding time goes into getting it right rather than into ramping touch volume.
The first 72 hours of a cycle. Setup happens fast. Day one is brand-card intake with the team: tone, banned topics, current narrative posture, legal redlines for your sub-vertical. Day two is resident matching against your category cluster; we pull the right 30 to 50 voices out of the pool and confirm none of them have run for an adjacent competitor in the prior 90 days. Day three the pool starts daily takes on your category without naming you yet, so the residents read as natural slices of the niche when you do enter on day four or five.
Reporting by beat, not by total. We break weekly reach and touches into three beats so the team can see where the volume actually landed. Warm beats are seeded category takes ahead of the date. Launch beats are direct project mentions inside the TGE window. Catch beats are news-cycle responses inside that same window. A campaign that landed 4,000 touches all on launch beats and zero on warm beats looks healthy in total but tells you the room was never warmed. Two beats with under-performing volume usually mean the brand card or the resident match needs tuning before the next cycle.
Vertical-specific calibration. A DeFi protocol launch and a memecoin launch use the same pool but very different brand cards. DeFi cards skew toward mechanism explanation, risk framing and on-chain receipts; the residents trade in technical language and link out to docs. Memecoin cards skew toward narrative density, cashtag rhythm and meme-native responses; technical detail gets ignored. AI subnet and DePIN launches sit closer to DeFi but with hardware or model-quality vocabulary layered in. We have a template per vertical that the team tunes on day one, instead of trying to write a generic launch script that lands nowhere.
