How We’ve Learned To Pick Winners

Over the last 8 years, we’ve hired well into the double digits.

If you’ve ever had a bad staff member, you know the cost of hiring the wrong person.

My hit rate for a good employee hire was about 50% when I started.

That number has gone up to 80% - here’s the process I used to improve this output.

1. Start with a very accurate job description of what you want someone to do in your business.
   
   Repeat this step until you are confident you’ve got this right - DO NOT SKIP THIS (IF IT’S VAGUE YOU’LL GET VAGUE OUTCOMES)
   

2. Now take a specific task from this job description - here is an example:
   
   “Objective: Get more people using Whop for our company
   

   We have 43 people in our Whop community, with only 7 having created content for us. We’d love to see at least half our community using Whop to create content for us.
   

   How would you:
   

   • Activate more of our current users to create at least one piece of content for us per month

   • Recruit other creators into our community to make content for us at least once a month

   • How would you keep this community engaged with our company mission - what type of posts and content would you share with them?”
     

3. Once the person has done this, ask them why they chose this method of execution
   

4. Set a time limit they can spend on this process of no more than 3 hours, ask for it in 48 hours after and pay them the hourly rate they would get if they took the role

It’s amazing how quickly this process weeds people out.

I’ve bet on hires who I thought would nail this task, but didn’t get back to me within the 48 hours.

Or they nailed the task at first glance, but when you dig deeper into their reasoning, they used ChatGPT and copied that output verbatim.

Take the extra time to write a good task and test it; it’ll save you tens of thousands.

Being Sensible

I remember telling Kale at the end-of-year drinks for Compound Gym that we'd hit $1M in ARR at Gravy by the end of 2025.

That was December '23, and we'd just done $12,698 revenue in one month.

It took us till March '25 to break $20,000.

And only just now, October 2025, we have crossed $30,000 in a month for the first time.

A far cry from the million dollar a year run rate I thought we'd be on by now.

An idiot call really for someone whose been doing this for a decade. Only .001% (made up stat) of businesses actually have revenue curves that only go up and to the right like that.

And even for them, it doesn't last long.

What I did right this time, though (this isn't my first big incorrect projection), was not overcapitalising our business.

I didn't hire, take on software, accountants or anything in between until I had dollars in the bank.

We still don't even have a logo. Although I am tempted by our latest milestone.

Surely $30k = get logo.

I kept patient and have only committed to new spending at the very last minute.

The amount of stress I've saved myself by doing this is hard to describe.

I've been in overcommitted positions for my businesses, and while working in others.

Spending on projections rather than money in the bank creates unnecessary chaos and is entirely avoidable, and it is expensive in many intangible ways.

Silly projections are fine - I'm going to make another right now and say this time next year we'll definitely be doing $1M ARR.

But it doesn't really matter how wrong I am, as long as I stick to spending only when we already have the dollars to pay for it.