House of Pain is a gym in Chesterfield, MO. One location, a local customer base, and a paid-per-member business model. They came to us in August 2023, and the brief was refreshingly unglamorous: they didn't need viral, they needed the nearest fitness-minded people to know their name.
So that's what we built. Geo-locked targeting to Chesterfield and the St. Louis metro. Targeting pointed at followers of other local gyms, local personal trainers, and local fitness influencers, which is to say people who have already proven they'll pay money to work out. A welcome DM to every new follower with a free trial session attached. The result on the case study page is 929 new local followers, 100% of them inside the service area, and an immense increase in membership signups attributable to the Instagram channel.
That's a good outcome and I'll defend every piece of it. But I want to be honest about what it does and doesn't settle, because the gym is the client on our case studies page whose economics look least like the rest.
Why a gym is not a roofer
Look at the other work on the site. Zykan Exteriors sells roofs. Rocket Tint sells tint jobs, with a lifetime average ticket that climbed past $300. Seatmate sells a chair. In all three, the revenue event happens once. You win the customer, you collect, and the transaction closes. Marketing's job in those businesses is genuinely mostly acquisition, and the scoreboard is cost per acquired job against the value of that job.
A membership business inverts that. House of Pain doesn't get paid when someone signs up. It gets paid when someone signs up, and then again next month, and again the month after. The sale is not an event, it's a subscription that quietly re-decides itself every thirty days, and the thing casting that vote is whether the member walked through the door in the previous four weeks.
What that does to the value of a lead
Our own book on Magnet Pro says a qualified lead costs us $19, off $5.05 a site visit and a 21% visit-to-lead rate, verified through the Meta Marketing API. Those numbers are stable. What's not stable, and what no ad platform will ever report to you, is what the lead is worth once it converts.
Run the arithmetic on the ratio alone, because I'm not going to publish a membership price or a churn curve I didn't measure. A member who stays twelve months is worth four times a member who stays three months. Identical acquisition cost. Identical ad account. Identical creative. Four times the revenue, decided entirely by things that happen after marketing's job is over: the front desk, the first two weeks, the class schedule, whether anyone noticed the member stopped coming.
That ratio is the whole game, and it means the return on the ad account isn't set by the ad account. If members leave faster than the funnel replaces them, better targeting just makes the treadmill go faster. You can buy your way to a bigger top of funnel forever and the membership count sits flat, and the ad spend looks like it stopped working when what actually happened is that it never was the constraint.
The line in the case study that hides a second conversion
Our own case study says that converting even 5 to 10% of those 929 followers to trial sessions, then members, pays for Magnet Pro many times over. That's true. It's also doing something sneaky in four words: "to trial sessions, then members."
There are two conversions in that phrase and only one of them is a marketing conversion. Follower to trial is ours. We built the DM, we wrote the offer, we put the right person on the other end of it. Trial to member is the gym's, and it's won on the floor with a coach who learns the person's name. And then there's a third conversion nobody counts at all, which is member to member-again next month, and that one repeats for the life of the account.
The case study reports what we measured: local followers delivered, targeting precision, signups up. It does not publish a retention curve, and I'm not going to invent one to make a cleaner story. What I'll say is that the retention curve is the number I'd want on the wall if this were my gym, and it's the one most owners cannot produce when I ask.
What retention machinery actually looks like
The uncomfortable part for a marketing firm is that most of the fix isn't marketing. But some of it is, and it uses the exact same equipment we already installed for acquisition, pointed in the other direction.
- The welcome DM that converts a follower into a trial has an obvious sibling nobody builds: the day-14 message to a member who came twice and then stopped. Same channel, same tooling, dramatically different economics.
- The CRM that everything funnels into by day 90 of our cadence should be holding attendance and tenure, not just leads. A lead ledger tells you the top of the funnel. A tenure ledger tells you whether the business is actually compounding.
- Owned audience data is a retention asset, not just an acquisition one. Every visit fires your pixels and joins your retargeting pool and stays yours forever, which means a lapsed member is someone you can reach for pennies instead of paying full price to acquire a stranger who is statistically less likely to stick than the person who already liked you.
None of that is exotic. It's the same four pillars we run everywhere, with the recognition that for a subscription business the funnel doesn't end at signup. It ends when the member leaves, and every month before that is revenue you already paid to acquire.
What I ask a gym owner on the Zoom now
I don't open with reach anymore. I open with: how many of the people who were members a year ago are still members today? If the owner can answer that from a system rather than a feeling, we're going to have a productive engagement, because we can put a real number on what each of those 929 locals was worth and price the ad spend against it.
If the owner can't answer it, that's the first thing we build, and I'll say so on the call instead of selling a bigger ad budget into a leak. Marketing can put the right 929 people in front of a gym in Chesterfield. It cannot make them come back in February. The gym does that, and until it does, more leads are just a more expensive way to stay the same size.