Using Search Theory to find winning angles that convert cold traffic at scale.
They hand your research to an intern running ChatGPT deep research. A few Reddit threads, some social comments, your reviews pasted into the AI, and that is where it stops. They have no real idea who they are actually talking to, because they are trusting the AI to know. So when it gets things wrong, and it often does, nobody catches it.
Influencers, UGC, a booked-out podcast studio. All paid for before anyone has worked out what to say, how to say it, and who to say it to.
Nobody knows how many ads have been made for each angle, which are most profitable, which are quietly burning money, or where to go next. No data. No feedback loop. No growth.
(This is why 'just make more ads' is not fixing your problem.)
The US Navy had the same problem as you. In 1942, German U-boats were sinking Allied ships across the Atlantic, and the Navy had to find the submarines to stop them. The trouble was scale. A U-boat could be anywhere in millions of square miles of ocean, and there were only a handful of planes and ships to go and look. Search the whole map evenly and you cover all of it badly, and find nothing.
So they handed the problem to a mathematician, Bernard Koopman. His answer became search theory: stop spreading your effort evenly. Work out where the target is most likely to be, send everything there first, then redraw the map as each sweep teaches you more. Finding the one message that turns a stranger into a buyer is the exact same problem. A near-infinite ocean of angles, a limited budget to search it with. So we point the same maths at your ad account, and stop guessing.
The original search problem. We point the same maths at your ad account.
This is our solution.
We get on the phone with 10 to 20 of your actual customers and let them write the copy for us. Their words, their reasons, the exact moment they decided to buy. By the time we spend a single penny, the ocean is already a fraction of the size.
Cheap, message-isolating creative across the shortlist. Same format every time, so the only thing moving is the message. We track every angle: ads per angle, CPA per angle, ROAS per angle. Nothing left to a hunch.
We scale the proven winners on a 70/20/10 split and keep searching with the rest. New angles queued up before the current ones fatigue, then carried into fresh channels once they are proven and profitable.
Every winner teaches us something, and so does every dead angle. Run it our way, and every round costs less and finds more than the one before.
10 to 20 calls with your actual buyers. Their words become your angles, not an intern's ChatGPT scrape.
Every ad script written, all video edited, statics and UGC creators sourced and produced.
Our angle-testing system installed and run, so every message is tracked and ranked by profit.
A buying structure that tests cheap, then pours budget into the winners as they prove out.
Pages matched to every winning angle, plus the CRO to convert the traffic you send them.
Weekly performance reports and a direct line to me. No juniors, ever.
Three brands, anonymised. Click any one to walk through what we changed and see the receipts.
A mix of our founder's own clients and brands where he led creative strategy. Every brand's situation is different. Figures are taken straight from the clients' own Stripe and Shopify dashboards.
Cold Angles is founder-led by design. The person who audits your account, interviews your customers, writes the angles and reads the data is the same person on your weekly call. A direct line, not a ticket queue, and not a junior account manager learning on your budget.
Tatsuki ThomasFounder, Cold AnglesEmail players hat tip to Ronaldo Del Luca and Tatsuki Thomas for our new BerserkerMail ads.
…the numbers speak for themselves in terms of the number of people on the waitlist, the number of conversions that we did and the number of people who signed up. It's much bigger than we've ever done in the past.
One of the only people I trust when it comes to Meta ads
Shut up and feed me.
Where would I be right now without Tatsuki? I'd be lighting money on fire and sucking my thumb in the corner. If you're on the fence about working with Tatsuki, don't wait. Everyone else is going to give you big fluffy advice. Tatsuki works with you in the trenches, he knows the big picture strategy, he knows cold traffic.
Your copy is just insane honestly
When you need anything regarding copywriting, you need to go with Tatsuki Thomas.
This guy has helped me navigate a lot of copywriting things that I've been working on for my sales pages and just been absolutely excellent at calling out the tiny details that make a big difference in conversion rate.
Tatsuki is the mind behind some of Money X's biggest launches
Man, you're killing it.
What do you actually do for work?
Hands down the best Meta ads help I've ever recieved.
You've got a gift with words (and sales).
Tat's newsletter is one of the few daily emails I read - and I don't read many. There's always a golden nugget I can apply to my brand and business.
Tell me about your brand. If it's a fit, you'll book a call at the end.
Most brands have a winning ad live within the first 30 days.
Subscribers were coming in at a trickle. Money was going out, but no one knew which message was actually pulling buyers in, so spending more just meant losing more.
We opened with customer calls, not ads, and let the interviews hand us the exact angles that turned cold traffic into subscribers. Then we tested them one message at a time and put the budget behind the two that paid.
Paying subscribers went from 286 to 2,199 in two months, and weekly revenue climbed with them. Here is what that looked like in their own dashboard.
A subscription product stuck on the same handful of tired ads. Recurring revenue had flatlined because the account had stopped finding new messages that converted cold buyers.
We rebuilt the testing system so every angle was tracked on its own, found the messages that actually moved cold buyers, and scaled them while we kept searching with the rest.
Recurring revenue compounded fast once the search was running properly: MRR up 84.65% and paying subscribers up 57.5% in the window.
An ecom brand that had never really searched. They were spending on the formats they could buy, not the messages their buyers responded to, and monthly revenue had stalled.
We mapped the angles straight from their own buyers, tested formats cheaply, then scaled the winners across channels as each one proved out.
Monthly revenue went 8×, and in one four-month stretch we added $126,152 a month in new-customer revenue alone.
Three quick steps. If it's a fit, you'll book a call at the end.
Step 1 of 3
Step 2 of 3
Step 3 of 3. We read every application and reply within 48 hours.