Triple Whale's pricing page starts at $149/month, bumps to $429/month at $1M GMV, and crosses $899/month for $5M+ brands.

Northbeam publishes one number — a $1,500/month Startup plan, capped at brands doing under $1.5M/year in media spend. Even that requires a demo. Everything above it is a custom quote.

Hyros runs two pricing tracks. The Shopify track lists tiers from $69/month (under $5K/month tracked revenue) and $129/month (at $10K), scaling up from there. The Business track — for info-product and direct-response brands — starts at $230/month annual. Every Hyros tier, on either track, still requires a setup call before you can buy.

Three attribution tools. Different in how they publish. One thing in common: the price you pay scales with your revenue, independent of what it costs them to run your account. And nearly every meaningful tier puts a demo call between you and the buy.

So I went looking for the other number. What does it actually cost to run attribution and analytics for a single Shopify brand?

Ingesting Shopify + Meta + Google data for one mid-sized brand: a few cents a day. Storage (ClickHouse handles it at ~1GB per client, compressed): negligible. Dashboard rendering on a shared instance: fractional. Orchestration, query layer, alerting — all open-source, running on commodity cloud.

Total cost of infrastructure for one attribution client, running 24/7: somewhere between $2 and $4 a month. I've built this stack. I know what the invoices look like.

Everything above that number is the business model.

The business model is GMV-scaled pricing. The tool charges you based on how much revenue it "tracks" for you — and the vendor's cost to run your account doesn't change when your revenue grows. Same script fires. Same rows get stored. Same queries run. The bill goes up anyway.

The mechanics underneath:

When your success becomes your vendor's upsell, you stop being a customer. You're a revenue curve on someone's quarterly slide.

There's a second layer propping up the apparatus: your data stays inside their platform.

You can see dashboards through their UI. You can filter and group. What you can't do is pull the underlying dataset into your own warehouse, connect it to your own Claude or ChatGPT, or run queries they haven't pre-built.

Same thing happens at the intelligence layer. Want to ask your AI assistant a question about your ad spend? Can't — the vendor ships a proprietary chatbot instead, and you pay for their tokens on top of their subscription. Their AI is the only one allowed to see your data.

If you could query your own dataset and point your own AI at it, you wouldn't need most of what you're paying for. So the platform is built to make sure you can't.

Every piece of this is a choice. The VC raise. The GMV pricing. The walled data. The proprietary chatbot. The category got locked into this model because the incentive structure pulled every company in the same direction. None of it is a law of physics.

We know this because we ran into it from the other side. Our clients at Inceptly — seven-figure to nine-figure D2C brands — almost always come in with an attribution tool already installed. For eight years, our job has been running it for them. Optimizing it. Troubleshooting it. Working around its limits.

We've seen every major tool in this category up close, across dozens of different business shapes. Somewhere around year five, it stopped being "hope the next one is better" and became "what would we build if we were the ones building it?"

Bratrax Lite goes live May 12. Attribution and analytics for Shopify D2C — $79 flat for the first 100 founding members, locked for life. $99 for everyone after.

Same price at $500K revenue. Same price at $5M. No tiers. No demo. No contract. Cancel whenever.

Your data is yours. When you want to leave, you leave with everything. We don't hold you hostage.

Your AI is yours too. Every Bratrax account ships with an MCP endpoint. Point Claude, ChatGPT, Gemini, or whatever comes next at your own data using your own API key. No proprietary chatbot. No token markup. We provide the data layer. You bring the intelligence.

The obvious question, once you know infrastructure is $4. Why not charge $5 and call it a day?

Because running a good software product costs more than infrastructure. You need a small, excellent team that builds it, supports it, keeps it running, and improves it every week. That team is paid fairly. That's the actual cost of running a software business the right way.

The gap between $99 and $1,500 sits somewhere else entirely. Sales teams, SDR orgs, CSMs whose job is to upsell you to the next tier, investor floors demanding 10x returns. That's the apparatus, not the product.

Jason Fried and David Heinemeier Hansson have been arguing for twenty years that companies can be profitable, sustainable, and fair without taking outside money or racing to the bottom. Their company, 37signals (Basecamp, HEY), has run profitably for over twenty years — no sales team, no SDRs, no traditional VC raise. They charge what their products are worth to build and support them well. Nothing more.

$99 a month is what it costs to build and support this tool honestly, with a small team doing real work, without a sales apparatus layered on top. The first 100 founding members lock in $79 for life — early adopters subsidized by future customers, not by an investor floor. No exit timeline to chase.

First 100 founding members lock in $79/month for life. 30-day money-back once we're live, no questions.

If you're tired of paying for someone else's Series B, you know what to do.

— Brat