LinkedIn, Intent-Led Outbound & AI GTM Operating Manual

Three things worth your time this week:

Try GoHighLevel free for 30 days. If you're still stitching together five tools to run outbound, nurture, and CRM, you're paying for complexity you don't need. GHL is the platform we run our own client infrastructure on. Get 30 days free here: gohighlevel.com/highlevel-bootcamp

🤖 Get $200 in free credits on Viktor. Viktor is worth testing if you're building AI-run workflows into your GTM stack. Claim your credits: app.viktor.com/signin

Want to talk through your GTM setup directly? Grab time on my calendar: calendly.com/scalematic/kaizenconsulting

Most people think outbound is a volume problem. Send more, get more. That's backwards.

Modern outbound works when four systems reinforce each other:

The goal isn't more messages. It's identifying the right people, catching them at the right moment, starting a natural conversation, and staying useful in their feed while you do it.

A modern GTM engine combines Sales Navigator, LinkedIn content, signal-based prospecting, Clay, enrichment, cold email, DMs, AI agents, calls, CRM, nurture, and founder-led media into one system. Not because more tools equals more results, but because each layer covers a gap the others leave open.

Your profile gets inspected every time someone receives a connection request from you, reads your comment, sees your post, gets your DM, or hears your name somewhere else. It needs to build credibility in the three seconds someone glances at it.

Photo. High resolution, good lighting, clean background, face clearly visible. Skip the low-quality selfie. Your photo and headline follow you across the platform, so they're doing work far beyond your profile page.

Headline. Most people waste it on a job title. "Founder @ Acme" tells someone who you are, not why they should care. Better:

Formula: Role + Result + What You Share. Example: Founder @ Acme • Built $10M Pipeline • Sharing Outbound Systems. Front-load the value before LinkedIn truncates the line.

Banner. Treat it like a billboard: who you help, what they achieve, proof, a simple CTA. Not a decorative graphic that says nothing.

About section. Don't write a resume. Use: Problem → Solution → Proof → Credibility → CTA. Short paragraphs. People scan, they don't read walls of text.

Featured section. Testimonials, case studies, top-performing posts, demos, media hits, lead magnets, results. Choose for business value, not likes.

Experience and recommendations. Write outcomes, not duties. Ask good clients for recommendations when it's natural to.

Most list-building fails because it stops at job title, company size, location, and industry. That gets you thousands of people who could buy. It doesn't get you people who have a reason to care now.

50 highly relevant prospects will outperform 500 generic ones, especially on LinkedIn where your message capacity is limited compared to email.

Signals worth building lists around: Job changes, recent LinkedIn activity, mutual connections, group membership, profile visits, job postings, funding, tech adoption, competitor engagement, content engagement.

Sales Navigator filter: Changed Jobs. New leaders spend their first months evaluating systems, launching initiatives, reconsidering vendors, and looking for early wins. That makes them more open to change.

Sales Navigator filter: Posted Recently. Active users are more likely to see your connection request, your message, and actually respond.

Connections of connections. Warm network proximity beats cold targeting almost every time.

Groups. Reveal shared interests and give you a natural opener.

Viewed your profile. A warm signal. Curiosity has already happened. Prioritize follow-up here.

Tenure. New leaders are open to change but still learning the org. Long-tenured operators understand the problem deeply and often hold budget authority. A useful middle range to test is 2–7 years, depending on your offer.

Boolean search. OR expands title variations, AND requires multiple traits, NOT excludes noise.

Example targeting B2B SaaS founders and CEOs:

Use OR for title variance, AND sparingly for category requirements. Don't stack five AND clauses and wonder why your list is empty.

Let AI build the string for you. Prompt: "I'm targeting [ROLE] at [COMPANY TYPE]. Ideal prospect: Job Title [titles], Company Type [category], Company Size [size], Must Have [criteria], Exclude [criteria]. Create a Boolean search string and explain the operators."

The best list often doesn't come from a standard database. Consider: podcast guests, conference speakers, job postings, Google Maps, competitor followers, content engagers, industry groups, company blog contributors, recent company announcements.

The closer the source sits to actual intent, the stronger the list.

Podcast guest prospecting. Find shows your buyers appear on. Extract guests, identify companies, segment by fit, research the actual episode, then reach out with context:

Job posting intelligence. Postings tell you what a company is investing in. Selling Laravel development? Monitor listings mentioning Laravel, PHP, Drupal. Then:

Google Maps. Strong for local-market prospecting — schools, restaurants, dental practices, contractors. Search category + geography, pull the company, find the site, find the decision-maker, enrich. Often outperforms broad LinkedIn industry filters for local businesses.

Segment into tiers.

A useful signal indicates a possible immediate need, is verifiable, and creates a natural conversation starter.

Strong signals: job changes, funding, hiring, product launches, market expansion, tech adoption, content engagement, competitor engagement, company page engagement, website visits, profile views.

Stack signals for higher priority. One signal is useful. Multiple concurrent signals are much stronger: funding + hiring, new VP Sales + Salesforce adoption, job change + LinkedIn engagement, product launch + market expansion.

Score intent based on what actually converts for you. Example model:

Your business will weight these differently. That's the point — build the model around your own conversion data, not someone else's.

Don't pitch the signal. The signal creates context, it isn't the message.

Timing decays. A signal from six months ago is far weaker than one from last week. Build systems that surface events fast.

Clay functions as a GTM workflow layer: combine data providers, enrich contacts, detect signals, run AI research, score prospects, export qualified leads. Think of it as an intent-signal factory.

Multi-source enrichment:

One source misses events. Combined sources close the gap.

AI signal qualification. Finding a signal isn't the win. Knowing whether it matters is. Example output:

AI can run that analysis at scale, across your whole list, continuously.

Custom scoring beats generic scoring. Example weighting:

Job change 25 pts · Relevant post 25 pts · Funding 15 pts · Tech adoption 15 pts · Hiring 10 pts · ICP fit 10 pts

Thresholds: 80+ immediate outreach, 60–79 priority, 40–59 nurture, below 40 monitor.

Once configured, a Clay workflow can continuously refresh account data, detect new events, score prospects, notify reps, export high-intent leads, remove existing customers, and suppress already-contacted leads. The output is a daily flow of fresh prospects rather than a static list you burn through once.

Instead of starting from a fixed list, AI agents can monitor competitors, influencers, job changes, funding, LinkedIn engagement, and company activity — surfacing people already showing intent. Those leads flow into LinkedIn campaigns, your CRM, or cold email.

A 30-day intent motion:

LinkedIn DMs are not cold emails. The first goal is a response, not a full pitch.

Initial DM.

Short. No pitch. Just conversation.

Progression, once they reveal a real problem.

The arc: Question → Pain → Proof → Value → Next step.

Backup follow-up.

The 80/20 of DM-setting. Strong operators don't pitch immediately, use corporate language, ask for calls with no context, or interrogate. They write naturally, ask smart questions, follow up with value, and let the prospect's own pain create the reason for the call.

Fix the interrogation problem. Question → Answer → Question feels like an interview. Better: Acknowledge → Add a statement → Ask.

Helpful beats needy.

Prioritize warm signals first: followed you, connected with you, liked a post, commented, downloaded a resource, viewed your profile. They already know you exist — treat them differently than a cold contact.

Weak follow-ups restate that you're waiting: "Bumping this," "Thoughts?," "Did you see my message?"

Better follow-ups bring something new: a case study, a short video, a resource, a relevant observation, a specific insight. Some systems run 5–10 follow-ups, but only when each one carries enough new value to justify sending it. Never repeat the same message.

Text isn't the only medium available. Voice and video notes create a pattern interrupt when kept short.

Three-line formula: Context → Insight → Question.

Contrarian hook:

Social proof hook:

Double voice note strategy. Send two short notes close together. The first opens a loop:

The second closes it:

The open loop creates curiosity that a single message doesn't.

A multi-step sequence can combine profile view, like, follow, connection, message, comment, follow-up, profile revisit, CTA, and break-up — automated end to end. Keep it conservative. Account safety comes first.

Example warm-up sequence: view profile → like recent content → follow → send connection → after acceptance, send value message → engage with a post → send personalized follow-up → revisit profile → send CTA → close the loop respectfully.

Platform limits shift over time, so don't anchor to a specific number. The operating principle holds regardless: don't behave like a bot.

At scale, managing hundreds of LinkedIn conversations manually breaks down. AI can remember context, draft replies, categorize intent, handle objections, and schedule.

Workflow: prospect responds → AI receives conversation context → AI checks the offer and SOP → AI drafts or sends the next response → once interest becomes explicit, it moves the conversation toward the meeting.

What the agent needs to do this well: the Goal (what outcome the conversation should reach), the Offer (what's being sold), the Profile (who's speaking), and the Campaign (why the prospect is in this conversation at all). Without that context, the replies come out generic.

AI brings consistent memory, fast response, emotional neutrality, and personalization at scale. Humans still win on high-value accounts, complex negotiation, sensitive conversations, and strategic opportunities.

Context prompt structure:

Never let the AI invent personal facts about your reps or the prospect. The operating rule: use only what you can actually support.

First-message prompt:

Message prompt:

Follow-up prompt:

AI can't sell an offer it doesn't understand. Build a complete offer briefing: problem, solution, differentiation, proof, ICP, common objections, customer outcomes.

One offer, one knowledge base. Don't mix unrelated products into a single AI brain. Separate offerings for different products, different ICPs, different use cases — even one product may need different messaging for SMB, mid-market, and enterprise.

Specificity wins.

Give it proof — metrics, case studies, testimonials, before/after — or the agent sounds generic no matter how well it's prompted.

Test before you trust it. Run simulated conversations against an interested prospect, a skeptical one, a no-budget one, a busy one, and a competitor's current user. Check whether it sounds human, stays accurate, handles objections, and stays on message.

A positive reply isn't the win. What happens after determines whether it turns into revenue.

The math that gets missed. Send 30,000 emails a month at one positive reply per 600 sends, and you get roughly 50 leads. If your response to those 50 is slow, most of them disappear before you ever talk to them. Faster, more coordinated response dramatically improves meeting conversion — often more than any change you'd make to the outreach itself.

Workflow: reply received → workflow automation → AI categorizes → enrichment → Slack alert → CRM. The goal is speed, measured in minutes, not days.

Enrich every qualified lead — phone, LinkedIn, email, whatever's appropriate — so a human rep can respond through multiple channels immediately.

Automate the follow-up tasks. Reps forget. Use CRM automation to force reminders, next actions, sequences, and escalations so every positive lead has a next step assigned to someone.

Context over content.

Value before ask. Sequence: insight → case study → meeting request. Don't ask in every message.

Persistent without annoying. Space your touches, change the angle, add something new. Skip the guilt trips and the daily nudges.

Acknowledge first. "Totally understand why timing would be a concern."

Ask questions to find the real issue. "When you say budget is tight, is that cash flow or simply not a current priority?"

Feel, Felt, Found. "I understand how you feel. Others felt the same. What they found was…" — use it naturally, not mechanically.

Turn objections into discovery. Price → explore value. Timing → explore priority. Authority → identify the actual decision-maker.

"Let me think about it." Don't disappear. Ask: "What specifically do you want to think through?" That usually surfaces the real objection underneath.

Use stories over statistics. A case study lets the prospect see themselves in the outcome.

Know when to walk away. Not every prospect is qualified, and forcing every conversation into a meeting wastes both sides' time.

Three layers make up a real GTM system:

A lean example stack:

Tools change constantly. This is a snapshot, not a requirement — the architecture matters more than any single logo.

Inbound tools: Trigify, PhantomBuster, AuthoredUp, Calendly, RB2B, scraping tools, Kleo, Zapier — for scraping engagement, tracking content, identifying visitors, automating notifications.

Outbound tools: Clay, Prospeo, Apollo, CompanyEnrich, LeadMagic, TheirStack, BetterContact, Smartlead, AI models, HubSpot, CloudTalk. Give every tool a clear job. Avoid shiny-tool syndrome.

Instantly, Smartlead, Saleshandy, Lemlist, Woodpecker, QuickMail, Reply.io, Mailshake, Snov, Apollo, GMass, PlusVibe, EmailBison, Mails.ai — common features include inbox rotation, warm-up, sender limits, sequencing, multichannel tasks, AI writing, and verification.

Don't run cold outreach through marketing platforms like Mailchimp or HubSpot marketing email. Cold outreach and opt-in marketing have different policies, deliverability assumptions, infrastructure, and sending patterns. Use tools built for the job you're actually doing.

LinkedIn generates inbound demand too: Content → Engagement → Lead magnet → DM → Email list → Call.

Lead magnets should be genuinely valuable — a framework, checklist, template, playbook, audit, calculator. Post: "Comment X and I'll send it." That drives comments, distribution, and warm leads. The quality of the magnet is what makes this work, not the mechanic.

PIER framework: Problem (open with pain) → Insight (something useful or counterintuitive) → Evidence (proof) → Resolution (the takeaway or action).

Break-a-misconception posts. "Most founders think more leads are the answer. I increased sales 28% by fixing follow-up." The contradiction creates attention — keep it truthful.

Case study posts. "Josh automated 70% of onboarding in three days. Here's how." Let the story sell, mention the product naturally.

Direct call-grab posts. Teach something useful, then: "I'm doing 10 free setup calls this week. Comment DM if you want one." Frame the call as help, not a sales meeting — and only offer what you can actually deliver.

A founder can become distribution. A three-layer system: founder-led storytelling (weekly updates, fundraising stories, feature launches, customer stories), paid distribution (ads mirroring what already works organically), and outbound sales (land-and-expand). The layers reinforce each other.

Story arc: Earn (the experience that qualifies you) → Build (how the product/company came together) → Prove (results, metrics, customer stories) → Scale (team growth, expansion, new markets).

Recurring series make content sustainable: Weekly Growth Update, Building the Company, Feature Drop, Customer Rollout, Founder POV.

Define your enemy. Strong positioning names what you oppose — administrative drag, tool sprawl, generic AI, slow incumbents. It clarifies why the product exists.

Pick an archetype deliberately. Example: the Guide — calm, helpful, proactive, keeps the customer in control. It shapes every piece of content you put out.

Funnel: Attention (stories) → Retention (recurring series) → Monetization (one clear CTA).

Templates worth stealing:

Show the mess, not just the win. Public success next to messy reality builds more trust than highlight reels alone.

Make the team a channel. Shared banners, repeated positioning, similar CTAs, coordinated themes across the whole company.

Reddit punishes anything that looks like advertising. Find relevant communities (SaaS, startup, product, industry-specific, competitor discussion) and give more than you take before you ever promote anything — answer questions, share real experience, build account credibility.

Listen for competitor mentions and category terms. When someone describes the problem you solve, contribute something genuinely useful. Promotion stays secondary. The principle: distribute insight quietly, don't behave like an advertiser.

A broad program can include lead research, LinkedIn outreach, cold email, field marketing, content, LinkedIn ads, influencer relationships, search ads, SEO, and analytics. You don't need all ten running at once — choose based on stage, resources, ICP, and sales cycle.

A few worth flagging:

Fifteen steps, in order:

Modern GTM is a closed-loop data system. Signals tell you who may care. AI helps explain why. Content builds familiarity before you ever reach out. Outbound starts the conversation. Human discovery finds the real problem. Automation preserves the context so nothing gets lost. CRM tracks the opportunity. Revenue data teaches the system what actually worked. Then the cycle improves.

The question worth asking isn't how many people can we message.

It's: how accurately can we identify the right person, at the right moment, with the right reason to talk?

That's the whole game.

Got a signal source or a DM framework that's working for you right now? Reply and tell me — the best ones make it into the next issue.