Why I Started Helping SaaS Teams Fix Their Marketing Data
Here’s something I didn’t expect when I started building Arcana and Sigil: the hardest parts of trading data engineering are the exact same hard parts that SaaS founders deal with when they try to figure out where their money is going.
I spent months building systems that pull millions of trades from exchanges, clean them up, enrich them, and turn them into something you can actually make decisions with. The whole time, I was solving problems like: how do you make sure you don’t miss data when a connection drops? How do you combine information from different sources that don’t agree with each other? How do you run through hundreds of scenarios fast enough that you can actually learn something?
Turns out, those are the same problems that show up when a SaaS founder asks “which of my ad channels is actually bringing in customers?”
It’s the same work, just different data
Every data pipeline has the same basic shape. You pull data in, you clean it up and combine it with other data, you process it, and you get answers out the other side.
In trading, that looks like: connect to Coinbase, stream trades, build analytical bars, run strategies against them.
In marketing, that looks like: connect to Meta and Google Ads, pull campaign data, combine it with Stripe or Shopify revenue data, and figure out which channels are actually driving paying customers.
The data sources are different. The APIs are different. But the engineering challenges are almost identical. Unreliable connections. Data that shows up late or out of order. Multiple sources that tell conflicting stories. The need to reprocess without duplicating anything.
If you can build a system that handles millions of trades without losing data, you can build a system that tells a founder exactly what their CAC is by channel. It’s the same muscle.
Where this gets interesting for businesses
Most marketing teams are stuck in a frustrating spot. Google Ads says one thing. Meta says another. Stripe says a third thing. Nobody agrees, and there’s no easy way to reconcile the numbers. So decisions about where to spend the next dollar end up being based on gut feel or whoever shouts loudest in the meeting.
That’s not a strategy problem. That’s a plumbing problem. And plumbing is what I do.
The same kind of enrichment work that turns raw trades into meaningful trading signals can turn raw campaign data into honest attribution. The same optimization engine that tests hundreds of trading strategies can test hundreds of audience and creative combinations to find the ones that actually minimize your cost per customer.
How Position5 approaches this
When we take on a consulting engagement, we bring the same process that built Arcana and Sigil:
- Understand the data. What sources do you have? What’s the quality like? Where are the gaps?
- Build the pipeline. Pull everything into one place, clean it up, make it trustworthy.
- Make it production-grade. Not a spreadsheet that works once. A system that runs reliably and handles the weird stuff gracefully.
- Optimize for speed. The value isn’t in one report. It’s in being able to ask new questions quickly and trust the answers.
The first consulting engagement is already underway. Same engineering, different data, same goal: giving people numbers they can actually act on.
If your team is sitting on marketing data that isn’t telling you what you need to know, let’s talk. This is the work I’m built for.