
Last Saturday was session six of The Independent Actuary founding cohort. Two to go.
Two weeks ago I shared the niche scorecard from week one. This issue gives away the exercise from week five, which is the week the work changes.
Weeks one to four build a position: a niche, a priced offer, a profile, a content habit. All of that makes you findable. But findable is passive. Week five is where you stop waiting to be found and start contacting people.
It is also the week people most want to postpone. The line I hear most is “I’ll reach out once I’m ready.” This exercise is my answer to that, and it works whether or not you ever leave your job.
Start With 30 Names
Your first clients will almost certainly come from people who already know you. The trust is already there. Cold outreach is a numbers game. Warm outreach is a relationship game, and you start it with a lead.
So write down 30 people you know professionally. Former bosses and colleagues. Auditors. Reinsurer contacts. Software vendors. Recruiters. Exam-mates. People who comment on your posts.
Most actuaries believe they have no network until they write the list. If you stall at 15, export your LinkedIn connections, scan old project folders, and think through committees and conference lists.

Sort Them Into Three Tiers
Next to each name, write which of these three they are.

The last column is the ask that fits each tier. An advice ask is “I’d value your take on this.” A referral ask is “if anyone in your world is wrestling with this, I’d be grateful for an introduction.” A direct ask is “worth 20 minutes to see if this is useful for your team?”

The Tier Everyone Underrates
Most people put all their attention on buyers and ignore bridges. That is backwards.
A buyer has one company’s problems. A bridge sees many. An auditor might see twenty carriers a year. A software vendor hears from every client who is struggling with the model. A retiring actuary has clients who will need someone next.
My own bridges, without names: a CEO I work with forwarded my note to an advisory firm that is now a referral partner. A peer at a boutique firm agreed to send me conflict and overflow work. The honest status is more commitments than signed deals so far. Bridges compound slowly, but they do compound.
One more check before you send anything. Put your list of target companies next to your 30 names. For each company, ask who you know there, or who knows someone there. Those paths are your first five messages.

The Research Ask
This is the part that matters if you are employed and can’t sell anything yet.
You don’t need a finished offer to start. You need a question. A research ask is a message asking how someone handles a problem in your area, to learn and not to sell. “How is your team handling the quarterly close these days?” “What’s the hardest part of experience studies for a company your size?” No offer, no rates, no “I’m available.”
The message is four lines and under 100 words: a personal first line, one line on what you’re working on, one ask, and an easy out.
Hi [Name], [a shared memory or their recent news]. I’ve been going deep on [your topic] lately, and I’d really value your take on [one specific question]. Would you be open to a 20-minute call in the next couple of weeks? Completely understand if things are busy.
If the first line could go to anyone, rewrite it. A tip that came out of the room: read the person’s recent posts first and open with the topic they care about. People reply to their own passions.
Two rules if you’re employed. Never approach your employer’s clients. And read your outside-work policy before you take any paid work.
These conversations reconnect you with people you haven’t spoken to in years, and they teach you what buyers actually struggle with. Then, on the day you do go independent, your announcement goes to people you spoke with last quarter, not people you last saw five years ago.

Five Messages This Week
The homework was simple. Tier the 30 names. Send five messages. Log each one in a spreadsheet. Book one 20-minute conversation.
Five a week means 30 names is six weeks of outreach. If someone doesn’t reply, send one gentle bump after five to seven days, then leave it for a month. Silence isn’t no. It’s “not now.”
Not everyone in the room sent five, and that’s fine. But messages went out, replies came back, and some members now have calls booked.
Set your expectations honestly. This is a pipeline, not a vending machine. One real opportunity in 60 to 90 days is a good start.
The first message is the hardest one. After that, it’s just Tuesday.

Cohort Two
The founding cohort has two sessions left. Cohort two runs in spring 2027, same format: eight live Saturday sessions, 10 to 15 seats, a private 1:1, and the book included. The price is $1,497, with a two-payment option. The waitlist gets first access to seats.
If you want on the list, hit reply and tell me one thing: where you are in the journey. Still thinking about it, building on the side, or already out and tired of the recruiter channel. I read every reply and answer each one myself.
If you’d rather not reply, the waitlist form is here: https://form.typeform.com/to/bGYLfTPw

P.S. Never going independent? Write the 30 names anyway. The same three tiers decide who thinks of you when a role opens up.

Looking for clarity on consulting, income, or next steps?

Last week we covered The Drugs Work. Do People Keep Taking Them? Why Persistence Decides What GLP-1s Mean for Insurers.
👉 If you missed the last week’s issue, you can find it here.

💼 Sponsor Us
Get your business or product in front of thousands of engaged actuarial professional every week.
💥 AI Prompt Of The Week
Automate My Close
Helps you find the automation wins hiding in a recurring valuation or reporting process. Ranks each step by time saved and risk, so you start with the safe, high-payoff pieces.
The Prompt:
Here are the steps in my [monthly/quarterly] [valuation/reporting] process, with rough time spent on each: [paste steps]. For each step, tell me whether it is a good candidate for automation, the simplest tool that would do it (Excel/Power Query, Python, SQL, or an AI agent), and the control I would need so an automated error doesn’t slip through. Rank the steps by time saved versus risk, and tell me which one to automate first.



