I scraped 119,412 conversations for this

The absolute worst thing you can do for your business is guess what your customers actually want.

You end up finding out months and thousands of dollars later that you were completely wrong.

I call this the guessing tax.

But here's the truth... the exact answers you need are already sitting right there. You just can't read 4,000 reviews and a thousand TikToks fast enough to catch market trends in time.

So I got tired of paying that tax, and I built a solution.

It's a completely free, open-source tool called Frontrun (scroll down to clone it).

Frontrun reads your entire category and breaks down two things:
1) where is my category moving, and 2) where are my competitors exposed

And if you're lucky, it can even spot a lottery ticket, a trend that's about to pop before it becomes a bidding war.

Let me show you exactly how it works.

Yup. I scraped 119,411 user conversations in 30 days.

This is for a big project where every convo, every creator, every hook, is tagged, broken down by source, angle, geographic, demographic, format, you name it. The moves we’ve been getting from it are nuts. And while that builds running at serious scale, I pulled one slice out for you.

That’s Frontrun.

And the first thing Frontrun did was prove me wrong.

Let me tell what you happened.

This ingredient kept blowing up my feed. Magnesium L-threonate. I don’t usually get supplement content but I kept seeing creators stacking it.

And the thought crossed my mind… “This thing's about to pop. I could grab a few friends already in the space, spin up my GTM agents, and pull some cash out before the window closes”

So I threw Frontrun at it.

Search demand for threonate was already at 74% of its all-time high. Everybody knew about this thing except me. Ol’ Matt is late to the party.

I wanted a hockey stick. What I got was a flat line that peaked months ago.

And here's the part that matters for you. That same check runs on your own brand, every time you point it at your category. It lays the whole board in front of you. What's still climbing. What your customers are already done with. Where your competitor's exposed. Then you make the call. Five minutes, receipts in hand.

That board is the product. Let me show you one.

(Or skip the setup, reply to this email and I’ll run it for you. Free for the first few

I pointed it at magnesium. The whole category, not one idea.

It read the market and handed back the board. Ten angles staged:

Seven angles are rising. Ride those now, while there's still room. Three are mature. People still want them, but the category's crowded, so you win on language, not on novelty/angle. Zero are pre-wave this run. Nothing's early here.

Now I have a map of exactly where every angle stands and what to do about each.

This is what works for your category, whatever you sell. Magnesium, mattresses, B2B payroll software. Point it at the URL, get the board.

Reply to this email with your category. We’ll run it and send you the read (free for the first few)

Half the map is YOUR angles. The other half is your competition.

Take oxide, one of the three mature magnesium angles. It still pulls heavy search, near the top of the category. So a competitor sees that volume and leads their whole pitch with "high-potency magnesium oxide."

The tool's read on that lane: the category already settled here.

That's the second thing every run gets you: where a competitor is spending cash on angles their customers don’t want.

Frontrun pulls that straight from the reviews and comments, (in the customer's own words) and shows you the opening.

Every brand has a competitor who's exposed somewhere. And only a select few categories find a pre-wave trend.

A map is a snapshot. But the big value is mapping the data over time.

Run Frontrun on the same category next week, and next month, and the angles move. A rising one tips into mature. A mature one starts to die and frees up the budget you were wasting on it. The thing creators were whispering about finally hits search.

One run tells you where you stand today. A standing weekly run tells you where the category is heading before your competitors feel it. That's the difference between reading the market once and watching it move.

Point it at your category. Then keep pointing it.

A map is worthless if it makes things up. So this one can't.

A $5k/month Meltwater sub (or Brandwatch Sprout, etc) just counts. It pulls 4,000 comments and tells you "pricing" was mentioned 47 times. So you slap "address pricing" on a deck, and nothing gets done.

Frontrun ranks by usefulness instead. Same engine on a SaaS pool here, the math doesn't care what you sell:

The 3-mention line beats the 47-mention cluster, 26.0 to 3.2. Mentioned 15 times less, worth 8 times more. The loud, useless theme gets barred.

The specific, commercial one gets promoted.

So that's the floor. Here's the ceiling.

Every so often, buried in the noise, Frontrun finds the one shape it actually hunts for: creators loud, search still empty. The thing nobody's clocked yet. To see how rare that really is, I ran some tests.

These are six categories printing in ecomm right now:

295 candidates, everything hot across all six:

Take skincare. It threw out every ingredient you've already heard of and showed me where each one really sits:

That number is how hard people are searching each one right now, out of 100. The word next to it is the verdict.

Brutal news if you’re looking for novel arbitrage. ALmost everything blowing up your feed is already late. You’d be buying the top.

But one lane out of 295 stayed open.

So what made it through all 295? One ingredient.

Spilanthol. Acmella oleracea. The one creators are calling "nature's Botox", a tingly flower extract they're pitching as Botox without the needle.

Here's why it lived. Search sits at 14.5 out of 100, near the floor. And it's never had a real wave. Most low terms are low because they already peaked and crashed. But this one's low because it hasn't happened yet.

Meanwhile creators are catching on:

There’s even a brand is already shipping a "Fountain of Youth" face oil built on it.

The language is jumping into Spanish. There’s a creator's who’s hook is that a billion-dollar cosmetics giant patented it and buried it.

Creators loud. Search empty. In 295 tries, it found it once.

On September 1 I’m going to run the numbers again (and I’ll post the results). If search has jumped past 20 and the creators are still at it, the call hit. If it's flat and they've gone quiet, it was noise, or we’re still early.

If I'm right, whoever moves on it now owns the lane before it turns into a bidding war.

The old way to learn all this: an $8k research agency, a $5k positioning strategist, three weeks of ad spend confirming what your reviews already said. Call it $35k a quarter, and at least a month gone.

Frontrun gives you the same map in five minutes.

And the real power isn't the one run. It's running it every week and watching the board move, rising terms cooling off, dead ones you can finally cut, a new lane cracking open before anyone else sees it.

Or skip the setup. Reply with your brand URL and we'll run the whole map for you, free for the first few who ask.

Point Frontrun at your category if:

→ you want a map of where every part of your market stands, rising, mature, or dead, with the move for each
→ you're sitting on reviews, comments, and competitor ad libraries you've never actually read
→ you want to know where your rivals are exposed before your next campaign, not after
→ you'd rather hear "this isn't early" than get flattered into a bad bet

Not for you if:

→ you've got a research-ops team tracking your category's demand stages in real time. you're the 1%.
→ you want a button that spits out ten confident trend calls from no input. that's the slop machine this exists to beat. mine killed 294 out of 295.

The full engine is open source. Point it at your category and read the map yourself. (You wire up your own scraper API keys. The README walks you through it.)

You get:

No upsell. No catch. MIT license. Fork it.

Stop paying the guessing tax. Start reading the map.

If you have been building flows and custom builds.

Go big,

Matt

P.S. Frontrun is the layer that sits in front of the rest of the stack. Slideshow Kit ships the angles to your feeds. Outcome Kit tells you which ones produced buyers. Frontrun maps the move, Slideshow posts it, Outcome scores it.

P.P.S. Frontrun is opensource slice. If you want the full power of the build it came from, reply to this email with your brand or project. That's the whole machine: hundreds of thousands of user and creator voices, sliced by demographic, geography, and segment, delivered as a daily Intelligence Briefs, alerts, and a dashboard that shows you exactly where your market is moving and who's moving it first.

P.P.P.S. Star the repo if you grab it. It tells me to keep building. And in six weeks, hold me to the spilanthol call.