How to Build an AI Outreach System for Your Coaching Business (In One Evening)
You can build a working AI outreach system for a coaching business in one evening, using a scraper, a LinkedIn connector, and an AI assistant that plans first and builds second. At our weekly AI meetup in Lombok, a fitness coach asked the question that started it: she sells to individuals through Instagram ads, so could she pitch her wellness program to companies instead, without hiring a sales team?
We built the answer live, in about two hours. Here is exactly how, so you can copy it.
Where do you start? With your manual steps
The biggest mistake in automation is starting with the tools. Start with the process. Write down what you would do by hand:
- Find the right companies.
- Find the right person inside each one.
- Send a message that does not feel like spam.
- Follow up.
- Send an offer to whoever raises their hand.
That list is your blueprint. Everything else is plumbing.
Who should you target? One customer profile
Ads cast a wide net. Cold outreach is a sniper shot. For our coach, the target became: HR leads at UK companies with 50 to 100 employees. Small enough to decide fast. Big enough to have a wellness budget.
One insight from the room changed the pitch. Remote companies already pay for wellness perks. Most go unused. So the offer is not "buy my program." It is "give your team a wellness perk they will actually use."
How do you brief the AI? Goal first, tasks second
We opened Claude and described the whole idea in plain language. Who we target. Which channels. What data to save. What the daily routine should look like.
Then the important part: ask for an implementation plan first. Review it. Only then let it build. We split the plan into parallel tasks, one chat per feature. The lead database. The LinkedIn connection. The scraper. The demo generator. Each one focused, each one fast.
What tools do you need?
Three pieces did most of the work.
A scraper. AI assistants cannot browse LinkedIn or Instagram directly. Tools like Apify can. It found the right profiles and pulled their public data. Ten dollars goes a long way.
A connector. An open source LinkedIn connector let Claude send connection requests and read replies without leaving the chat.
An email option. For low volume, your own inbox is fine. If you ever scale to real volume, use a separate domain bought just for outreach. One spam flag on your main domain and every email you send lands in spam.
If you are weighing paying for tools against building your own, we ran that exact math in Build vs Buy: What Replacing $340 a Month of SaaS Actually Cost Me.
What do you say? Lead with a gift, not a pitch
Nobody replies to "buy my stuff." The first message offers something free and personal. A taster session for their team. Or better: a one page wellness report built for their company, with a realistic first 90 days and a booking link at the bottom.
The numbers back this up. Backlinko's study of 12 million outreach emails found an average reply rate of 8.5 percent, and personalizing the message body lifted replies by 32.7 percent. Personalization is the single biggest lever you control.
So the system researches each company before writing a word. Website, LinkedIn, recent posts. The message references what it found. That is the trust signal. It says: a human did their homework here. Even when a human did not.
One warning. AI loves to invent statistics. It will happily claim your program improves retention 20x. Ground every number in real research or delete it.
How do you avoid getting banned? Go slow on purpose
Platforms ban accounts that behave like bots. We learned this the hard way. The fix is boring: start at 3 to 5 connection requests a day. Scale to about 15. Spread them across the day. The system enforces the cap so enthusiasm cannot break it.
This is the counterintuitive part. The automation is not for volume. It is for consistency and personalization. Fifteen deeply personalized messages a day, every day, beats three hundred generic ones that get you banned.
How does it get better? Let it improve itself
The best part came last. A weekly routine reads every conversation. Which openers got replies. Which offers got ignored. What objections came up. Then it rewrites its own messaging based on what won.
A daily digest tells the coach who to reply to, what went out, and what failed. She stays in the loop. The system does the rest.
What this means for coaches
You do not need a sales team to go B2B. You need one clear customer profile, one genuinely useful gift, and a system that shows up every day without getting tired.
The whole build took one evening. The tools cost less than a gym membership. The hard part is not technical anymore. It is knowing your offer and your customer. That part is still yours.
Frequently asked questions
How much does an AI outreach system cost to run?
Around $30 to $60 a month at coaching scale. Roughly $10 to $25 for scraping, $20 for email if you outgrow your own inbox, plus your AI assistant subscription. Paid all-in-one alternatives run $40 to $100 a month.
Do I need to know how to code?
No. You describe the goal in plain language, review the plan the AI proposes, and approve drafts before they send. The coach in this story does not code.
How many messages a day is safe?
Start at 3 to 5 LinkedIn connection requests a day and scale to about 15, spread across the day. Going faster is the most common way accounts get restricted.
If you want the deeper version of this, with an app your clients actually use on the other end of the funnel, that is what we build at Tribed.
Turn your expertise into a branded app business
Your programs, community, payments, content, automations, and an AI coach trained on your method and voice—built into one app for the App Store, Google Play, and web.