6 weeks to outbound that runs on real buying signals
Twelve live sessions, two hours each, taught by Ishan Vats. You do not learn tools. You build one multichannel system across email, LinkedIn and voice, triggered by signals you can defend, then run it live against a real account in the final two weeks. Recordings go up the same day if you miss one.
Application only. The application scores you as you fill it in, so you will know where you stand before you submit.
Email
Day 0
LinkedIn
Day 2
Voice
Day 5
written back
Most outbound fails for a reason no course covers.
It is not the copy. It is that the list had no reason to be contacted this week, the enrichment quietly cost more than the meeting was worth, the domains were never built as infrastructure, and four channels hit the same account in two days because nothing was suppressing anything.
Every one of those is an engineering problem. The Outbound Build teaches them as engineering problems, in the order you actually hit them, and makes you run the result against a live account before you are allowed to call it finished. That is the whole intent: you leave with a system in production, not notes about one.
Four things, all of them real.
Taught by Ishan Vats.
Ishan Vats
Founder of IV Consulting, an AI and automation consultancy in Delhi serving clients worldwide. Ten years and 150 plus clients. I build the GTM and AI systems teams run on, which means the failure modes in this curriculum are ones I have had to fix in production, on someone's real pipeline, with their money on the line. I teach every session myself.
Twelve sessions. Two phases. One system.
Two live sessions a week, two hours each, for six weeks. Every session ships a working component that feeds the next one. Sessions are recorded and posted the same day, so a missed session costs you the live question time, not the build.
The ICP contract and the account universe
Every broken outbound system is a broken ICP wearing automation. We write the ICP down as a contract with exclusion rules, not a persona slide, then build the universe it implies.
Signal taxonomy: what earns the word pinpoint
Most teams collect the signals that are easy to get, then wonder why replies do not move. We rank signal families by correlation to revenue, then attach source, freshness window and decay to each. A funding round is a different asset on day 3 and day 60.
Signal capture I: people
Job changes, hiring patterns, promotions and role gaps. The highest intent signals in B2B are about humans moving, not companies announcing. We instrument them on a schedule so they arrive fresh.
Signal capture II: company, stack and web
Funding, tech stack changes, competitor moves, site visitors and content engagement. Then the part everyone skips: composite scoring with decay, so one account with four weak signals does not outrank one with a single strong fresh one.
Enrichment that survives an audit
The session nobody else teaches. Waterfalls, validation gates, bounce ceilings, and the number that decides whether this is a business or a hobby: cost per verified contact. We set a spend ceiling and make the table respect it.
From a signal to a sentence you can defend
Above the line versus below the line messaging, the six personalisation buckets, and one test applied to every generated line: would you say this out loud on a call? AI writes the variants. A human keeps the veto.
Deliverability as an engineering problem
Domains, mailbox estate, warmup schedules, SPF, DKIM, DMARC, inbox rotation, reply routing and volume maths against a stated bounce ceiling. Treated as infrastructure with failure modes, not as a list of tips.
Multichannel orchestration: one account, many doors
The spine of the programme. Email, LinkedIn, voice and retargeting sequenced off a single signal, with cross channel suppression so the channels never collide on one account, and CRM writeback so the system has a memory.
Intake, baseline and the compliance gate
Your account is scoped and signed off. Baseline numbers are recorded before anything sends, because a system with no baseline can claim anything. Consent, suppression lists and jurisdiction checks are cleared here or you do not send.
Production, at small volume
The system goes live. Low volume on purpose, real prospects, real replies. Session 10 is a desk review, not a lecture: your tables and workflows reviewed against a published rubric. Most people find their session 8 build breaks on contact with reality. That is the point of having two weeks left.
Read the instrument, then kill something
Reply rate segmented by signal, meetings per thousand touches, cost per meeting, deliverability drift. Then the discipline: kill one signal that is not earning its enrichment cost, and double down on one that is.
Handover and defence
Documentation, runbook, named failure modes with recovery steps, and a cost model the account owner can read. Then a twenty minute live defence to the cohort and one invited guest, who asks the hard question.
Ten tools in the estate.
Working access is included for the six weeks. Smartlead and Lemlist are both here on purpose: Smartlead is the sending infrastructure, Lemlist is the multichannel orchestration.
Clay
Orchestration, waterfall enrichment, AI columns.
Smartlead
Sending estate, deliverability and volume.
Lemlist
Multichannel sequencing and personalisation.
Prosp
Contact and phone discovery in the waterfall.
n8n
Schedulers, glue, retries and error handling.
Claude Code
Skills, MCP servers, research and copy agents.
HeyReach
LinkedIn signal capture and compliant sending.
Apify
Scraping for signals that have no API.
HubSpot
CRM writeback, so the system has a memory.
Retell AI
Signal triggered voice, a channel nobody else teaches.
This is what you get, prebuilt.
Ten skills you run rather than reinvent, each tied to the session that teaches it. They stay yours after the cohort ends.
Eight tools, built for this programme.
Not a swipe file. Each one is the working artifact from the session that teaches it, so you leave with the thing itself rather than notes about it. Yours to keep and reuse on every account after.
Want the thinking behind them first? , free.
This is not for everyone, and we would rather say so now.
Apply if
- You can point at a real account to run this against: your employer, your company, or a client.
- You have tried outbound and watched it underperform, and you want to know why.
- You can commit roughly eight hours a week: four live, four building.
- You are a founder, GTM or sales lead, RevOps, agency owner, or a technical marketer.
Do not apply if
- You want a job guarantee. There is not one, and anyone offering one is selling you something else.
- You will not do the 90 minute bench test. It is the filter, and it is not optional.
- You want to become an expert in every tool. This teaches a system, not a tool tour.
- You cannot make the live sessions and will not watch recordings within 48 hours.
One price. Everything included.
- 12 live sessions, two hours each, with Ishan
- Recordings posted the same day, yours to keep
- Tool access for the programme window
- The Signal Library, the eight tools, and every template used in class
- Two one to one desk reviews on your live account
- Your recorded week 6 defence, as a portfolio artifact
Before you apply.
What if I miss a live session?+
Six weeks sounds short. Is that enough?+
Do I need to know how to code?+
What exactly is the bench test?+
Why is the cohort so small?+
Do I need to buy all the tools?+
What do I actually walk away with?+
Is there a certificate?+
What if I apply and get declined?+
Cohort 2. Six weeks. One system.
If you have an account, a reason to contact the people on it, and eight hours a week, the application takes about eight minutes and tells you where you stand before you submit.
Not sure it is for you? Book the call and we will tell you honestly, whether or not you apply.