A signal-first, call-first outbound engine for the DataHub motion.

Seven ISO participant lists, EIA-860M generator inventory, FERC market-based-rate database, regulatory dockets (PUCT Project 58000, AESO Engage, ERCOT board), job boards with title-pattern detection per persona, event rosters, filings. Every signal arrives with its own source link. No third-party intent tools; your existing site-visitor and identity feed stays yours and gets joined to the signal store, not duplicated.

Up to 100 accounts per batch, each row carrying the signal, the date, the persona, the warm path if one exists, and the suppression status against your CRM. About 30 minutes of Adam's time per batch. Nothing is touched without his approval. Active customers, open pipeline, partners and opt-outs are suppressed before the batch even reaches him.

Name, title, organization, market; the signal with its date and source link; a one-line reason the signal predicts a data purchase; what to open the call with; the offer to make; direct dial and mobile where one exists; LinkedIn link (profile checked live); verified email; the warm path if one exists; the batch ID Adam approved. Same card format in both modes.

Accounts routed, calls made, conversations, meetings booked and held, time from signal to first touch, which signals are retired. Week-six readout re-fits the signal weights on measured results and sets the volume mix for the following quarter. Measured, never projected.

Daily: job boards (title patterns per persona), regulatory dockets (PUCT 58000, AESO Engage, ERCOT board, FERC), press and filings. Weekly: participant-list diffs across every ISO market, FERC authorizations, board-packet crawl for incumbent-contract signals, event rosters. Monthly: EIA-860M generator-inventory diff, ERCOT resource adequacy report, AESO connection project list.

Project and SPV names mapped to the parent organization. Parent to its web domain. Domain to your CRM account. Account to the 2 or 3 people who hold the buying roles from your framework. Suppression runs here: active customers, open pipeline, partners and opt-outs never get a cold touch; their signals go to your own senders as expansion or reactivation nudges.

Direct dial and mobile where one exists, LinkedIn profile checked live (we drop people who have left the role before the list reaches Adam), email verified. Warm paths computed from public data (alumni, co-presence on ISO committees, shared intermediaries like Alberta agents or Texas aggregators, investor and board paths).

Batches of up to 100 accounts; each row carries the signal and the persona; 30 minutes of Adam's time per batch. The first batch is the hard-dated set: Alberta REM participants, PJM auction participants, the ERCOT large-load queue, and the 12CP industrials routed to Adam. Subsequent batches pick up new-entrant and lifecycle signals as they fire.

Signal + source link + why it predicts a data purchase + opener + offer + direct dial + LinkedIn + verified email + warm path + batch ID. Hand-raises from your existing site-visitor feed arrive at the top of the queue the same day, with the signal that fired on that account attached, so you never double-touch a visitor who already came in.

Call first (your reported finding: cold email converts poorly with this audience; timely calls work). LinkedIn connect the same day with a note under 300 characters about the signal. Email as the third touch, measured, not assumed. Every touch carries the signal and play it came from. Week-6 readout retires dead signals and re-fits weights on measured data.

From public LinkedIn profile histories. Output: target, the mover, the customer-side colleague who can introduce.

A customer and a target seen on the same roster or filing. Output: target, the shared forum, the customer-side person in it.

The 77 AESO agents and the ~199 Texas registered aggregators from the PUCT directory that represent both a customer and a target. Also the advisors in your channel list.

From SEC, SEDAR+, ASX and similar public filings. Output: target, the shared party.

Detection: daily job-board scans (per-persona title patterns), daily regulatory-docket crawls (PUCT 58000, AESO Engage, ERCOT board, FERC), press and filings. Weekly diffs across PJM, MISO, NYISO, ISO-NE, IESO, SPP Markets+, AESO participant lists. Monthly EIA-860M, ERCOT resource adequacy, AESO connection project list.

Resolution: project and SPV names to parent orgs to web domains to your CRM accounts. Suppression against customers, pipeline, partners, opt-outs.

Enrichment: direct dial + mobile + LinkedIn (checked live) + verified email per person. Warm-path computation across the 4 path types in §05.

Approval workflow: batches of 100 accounts delivered to Adam with signal + persona per row. ~30 min of his time per batch.

Slack-card delivery: every approved lead lands in your channel as a self-contained card with everything the caller needs.

LinkedIn + email touches: we run connect requests under 300 chars tied to the signal; email as the measured third touch.

Weekly readout: Monday note + live sheet. Per signal: accounts routed, calls made (self-reported from your caller or Mode B logs), conversations, meetings. Retired signals named.

Everything in Mode A, plus:

Managed caller: trained English-speaking BDR on a recorded dialer, dial volume paced to signal cadence (tens per week, matching the signal-timed universe, not blind outbound).

Call hygiene: every call logged with outcome against the source signal. Recorded calls (per applicable state law). Direct-dial and mobile numbers worked first; LinkedIn connect + email follow the same day.

Meeting booking: qualified conversations booked straight onto Adam's calendar with the full signal context in the invite. "Qualified" means: fits a buyer row in §03, has a live dated or lifecycle signal, person holds a buying role from your framework, person has agreed to the conversation knowing what it is about.

Reply flow: every interaction surfaces in your Slack channel same day with the outcome tagged. Adam has full visibility.

Mode-switch: you can move from Mode A to Mode B (or back) at any readout. The build is identical.

Roster harvest: speakers + sponsors + exhibitors from the public organizer pages 4-6 weeks out. Where an attendee app provides a roster to registered attendees, your registered attendee exports it and we prioritize and enrich it (we never log in on anyone's behalf).

Enrichment + ICP scoring: every rostered organization and person scored against the 6 buyer rows in §03. You walk in with a ranked shortlist, not a directory.

Pre-event outreach: personalized email + LinkedIn (+ calls under Mode B) to the top-scored contacts in the 2-3 weeks before, referencing the event by name and offering a booth slot or coffee.

Meetings pre-booked: confirmed slots dropped into Adam's or your caller's calendar before anyone is in the room.

Post-event: follow-up sequence to every meeting-held and every no-show within 72 hours of event close.

Priced per event, billed on confirmation of the event date. Sponsorship decisions wait on the roster harvest and are yours.

ISO lists normalized and de-duplicated across markets. EIA entities mapped to parents. Alberta assets to participants to agents. Your CRM export matched. Suppression loaded (customers, pipeline, partners, opt-outs, inactive desks tagged for your senders). Slack channel and lead-card format agreed. Daily and weekly jobs live. Enrichment and live profile checks done for batch one (the hard-dated set: Alberta REM + PJM auction + ERCOT large-load + 12CP industrials routed to Adam). Warm paths computed for the top 50.

Mode A: your caller works batch one; we run the LinkedIn and email touches around the calls. Mode B: we run all three touches. AESO session attendance logged as a signal. ERCOT Power Next roster worked if the January event was added. Batches 2 and 3 approved by Adam: new-entrant signals (L1-L2) and first-merchant-year owners (L3). Monthly jobs live (EIA-860M diff, ERCOT resource adequacy, AESO connection project list). Readout one at day 30: signals observed, calls made, conversations, meetings by signal; dead signals retired.

Signal cadence runs weekly. Capability signals (C1-C4) layer in. Incumbent/renewal signals (R1-R3) start firing once board-packet crawls catch the first contract windows. Reactivation signals (X1-X2) handed to your own senders with our signal attached. Week-6 readout: measured per-signal conversion, dial-to-conversation rates, time-from-signal-to-first-touch. Planning estimates in the strategy are replaced by measured values; dead signals retired; signal weights re-fit.

Engine running on the winning signals from week 6. Referral-path layer stable, warm-path asks going through Adam and the CEO where relationships support it. ERCOT Power Next follow-ups (if the event was added) feed back into the queue. Day 85: POC report delivered with the full measured dataset, per-signal conversion, mode comparison, warm-path outcomes, and a scale-up or re-scope recommendation based on observed data. You decide to continue month to month in the same scope, expand, switch mode, or stop. Every asset (lists, suppression, signal store, enrichment, warm paths, Slack history) stays yours regardless.