Quick Answer: Running a LinkedIn outreach campaign to Heads of AI at mid-sized companies requires a built audience, a tailored sequence, and automation. Origami provides the built-in LinkedIn sequencer—you can find leads, enrich them, sequence, send, and track replies all in one platform.

This guide assumes you already have a list of Heads of AI at mid-sized companies. If you don’t, jump to how to build a list of Heads of AI at Mid-Sized Companies to get one in minutes using Origami’s AI agent.

Here, we’re covering the next phase: refining that list, crafting a 3‑touch LinkedIn sequence that actually gets replies, and launching it directly from Origami without switching tools. I’ll give you the exact messages I’ve used, plus what to expect when you press send.

Step 1: Build (or refresh) your list in Origami

Even if you already have a list, it pays to rebuild a fresh batch right before a campaign. Data decays fast—especially email and job changes. Here’s the prompt I’d type into Origami today:

“Find Heads of AI at US-based mid-sized companies (200–2,000 employees). Include their name, title, company, LinkedIn profile, verified email, and phone. Exclude companies that are AI vendors or consultancies. Focus on industries like healthcare, financial services, manufacturing, and logistics—where AI adoption is accelerating but teams are still lean.”

What you get back: a clean table with verified contacts, enriched profiles, and signals like “recently posted about MLOps” or “team size 3‑10.” Origami’s agent chained multiple data sources to confirm emails and phone numbers, so you’re not staring at guesswork.

If you don’t have an account yet, the free plan gives you 1,000 credits (no credit card) — enough to build and enrich a list of 50–100 Heads of AI. Paid plans start at $29/month.

But having a raw list isn’t the same as having an outreach-ready list.

Step 2: Refine and qualify the list

You’re not going to blast everyone. The key to LinkedIn outreach that doesn’t get you ignored is preaching to the choir—people who already feel the pain you solve.

How I qualify a Head of AI list

Filter by company size (employee count). Mid-sized isn’t uniform. A 200‑person fintech and a 1,800‑person manufacturer have very different AI maturity. I create two sub‑segments:
- Early Adopters (200‑500 employees): Often the Head of AI is the first hire, wearing multiple hats. Their pain: proving value fast, building a team, choosing the right stack.
- Scaling (500‑2,000 employees): They likely have a small team and some production models. Pain: technical debt, governance at scale, showing ROI to the CFO.
Check for AI/ML tool mentions in their LinkedIn profiles. Origami enriches each contact with “tools used” or “technologies mentioned.” If I see Kubernetes, MLflow, Hugging Face, or specific cloud AI services, I know they’re hands-on. If they only list “strategic AI leadership,” they’re more buyer than builder—different messaging needed.
Scan recent LinkedIn activity. Origami’s enriched data flags people who posted about AI in the last 90 days. Those are hotter leads. Someone griping about model drift or celebrating a production deployment is far more likely to reply.
Exclude obvious misfits. If the Head of AI title belongs to someone in a 200-person company that’s actually a digital marketing agency, skip. Same for people who’ve been in the role for less than 3 months—they’re in survival mode, not buying mode.

What “qualified” looks like

A qualified Head of AI at a mid-sized company:

Owns budget or strongly influences it (even if they’re not the final signer).
Is actively building or scaling a team.
Has talked publicly (or in their profile) about challenges like deploying models, data pipelines, or measuring business impact.
Operates in a sector where AI is a competitive advantage, not a buzzword.

Once you’ve got a shortlist of 50‑150 people who tick those boxes, it’s time to write the sequence.

Step 3: Create the LinkedIn 3‑touch sequence

In Origami, you have two ways to build your sequence:

Paste your own templates. Write a 3‑touch sequence yourself (connection request, follow‑up, final message), set delays between each touch (Day 1 → Day 3 → Day 7, or any cadence you prefer), and launch. You retain full control over copy.
Let the agent write it. Ask Origami’s AI to generate a personalized 3‑day LinkedIn sequence for all leads automatically. It pulls in each lead’s profile data—title, company, industry, tech stack—and writes unique messages per person. This scales personalization without you staring at a blank page.

Below, I’m giving you the templates I’d use if I were crafting the sequence myself. Steal them, tweak them, test them.

The 3‑touch LinkedIn sequence

Target audience: Heads of AI at mid-sized companies (200–2,000 employees), across industries like healthcare, fintech, manufacturing, logistics. Their core challenges: going from prototype to production, proving ROI, managing technical debt while keeping the team lean.

Touch 1: Connection request note (Day 1)

Sent with a personalized connection invite.

Hi [FirstName], saw you’re leading AI at [Company]. Most mid‑size teams I speak with wrestle with taking models from PoC to production without blowing up costs. I share practical insights on that—thought it’d be worth connecting.

Why it works: Acknowledges their world immediately (PoC to production), frames you as someone who understands the constraint (costs), and gives a low‑friction reason to accept. No pitch, no compliment fishing.

Touch 2: Follow‑up message (Day 3)

Sent after they’ve accepted, as a direct message.

Thanks for connecting, [FirstName]. I’ve been researching how mid‑market AI leaders manage model deployment with lean teams. Would you be open to a 10‑minute call to hear how you’re handling it? Not selling anything—just gathering perspectives. If not, no worries.

Why it works: Asks for their experience rather than talking about yourself. “Not selling anything” lowers guard. “If not, no worries” removes pressure.

Touch 3: Final message (Day 7)

Soft close, sent if no reply to Touch 2.

Hi [FirstName], circling back. I’ve been mapping how peers at mid‑sized companies cut AI infrastructure overhead by ~30% without adding headcount. If you’re open to a 15‑min chat this week, I’ll share what’s working. Otherwise, I’ll leave you be. Cheers.

Why it works: Provides a concrete, quantified value prop (“30%”) that’s directly relevant to their resource constraints. Clear ask, clear off‑ramp.

These three messages are short (50‑100 words each), avoid generic flattery, and respect the recipient’s time. They speak to the specific pain of a Head of AI at a mid‑sized org: doing more with less, justifying spend, scaling without breaking things.

If you’re using Origami‘s AI‑written option, the agent will produce variations that pull in details like their recent post on model monitoring or their mention of a particular cloud service, making each message feel less templated. But the structure remains the same—connection, curiosity, value.

Step 4: Send the sequence directly from Origami

This is where Origami saves you hours. You’ve already built the list and enriched it. Now you launch the sequence inside the same platform—no export, no spreadsheet gymnastics, no gluing three tools together.

Attach the sequence. In Origami’s sequencer, you select the list you refined in Step 2, pick your 3‑touch sequence (either the AI‑generated one or your custom templates), and set the delays: Day 1 connection request, Day 3 follow‑up, Day 7 final message. You can tweak delays if you want a slower or faster cadence.
Launch. The sequencer sends connection requests and follow‑up messages automatically. It respects LinkedIn’s realistic sending patterns—there’s no spammy burst—and operates within the platform’s guidelines.
Monitor everything in the dashboard. Opens, clicks, replies—all visible in the same feed where you built your list. Click on any contact, and you’ll still see their enriched profile (title, company, tools used, recent activity) so you instantly recall why you reached out. No mental context‑switching.
Automatic un‑enrollment. If someone replies, they’re immediately removed from the sequence. You won’t accidentally follow up with a “just checking in” message after they’ve already agreed to a call. That’s the kind of detail that keeps your reputation clean.
Pure end‑to‑end workflow. Find leads → enrich → sequence → send → track replies. All from one platform. The sequencer itself is free on all paid plans—you only pay for credits used to enrich leads. Paid plans start at $29/month.

What response rates should you expect?

Heads of AI are busy, but they’re also selective connectors. Based on campaigns I’ve run for similar technical leadership roles in mid‑market:

Connection acceptance rate: 30–45% (higher if your profile is credible, you share relevant content, and your note resonates).
Reply rate to follow‑up: 8–15% of those who accepted.
Meeting booked rate: 3–7% of the total list.

Your mileage depends heavily on how well you segment and how specific your messaging is. If you see acceptance rates below 20%, revisit your connection note or your targeting (maybe you’ve accidentally included people at large enterprises where the Head of AI is a different beast). If replies are low but acceptances are fine, iterate on the follow‑up angle. Test a different problem statement, or try shorter/longer messages.

When to tweak messaging vs. when to tweak the list

Low acceptance? Your list might have title mismatches (e.g., someone who’s actually a data science lead but titled Head of AI) or your connection note is too generic. Check if they’re a real fit; then rewrite the note.
Acceptances but no replies? The follow‑up message isn’t sparking curiosity. Try leading with a specific stat or a provocative question about their industry.
High replies but no meetings? Your final message or your offer isn’t concrete enough. “Chat about AI” is weak. “Share how teams of ≤5 are shipping models weekly” is stronger.

Rotate Your Device

How to Run a LinkedIn Outreach Campaign Targeting Heads of AI at Mid-Sized Companies (2026)