Why we’re building Kascade

Every B2B company wants inbound. There's a real difference between customers finding you through Google, ChatGPT, Reddit, or something they happened to read, and having to go chase every single one of them through outbound. Everyone knows this.

But building that channel is still weirdly hard.

You have to work out what your buyers are actually searching for, what questions they're asking, which topics are even worth fighting over. Then you research, write, edit, publish, distribute, update, internally link, track rankings, watch conversions, and at the end of all that try to figure out whether any of it turned into revenue. Most companies deal with this by building an internal content team, hiring an agency, duct-taping together a bunch of SEO and writing tools, or just not doing it at all.

And even when you do all of it properly, the feedback loop is thin. You know which pages get traffic. You might know which pages rank. But knowing which content actually created customers, and then using that to decide what to do next? That part is much harder. That's the problem we're building Kascade around.

Search is changing too

For a long time "organic" basically meant Google. That's not really true anymore. People still google things, obviously, but they're also asking ChatGPT, Claude and Perplexity. They're searching Reddit. They're reading discussions and recommendations without ever clicking through ten blue links.

So the interface people use to find information is shifting. The underlying behaviour hasn't gone anywhere though. People still have problems, they still ask questions, they still do their research before buying. There are just more places now where those questions get answered.

Which makes the old workflow of "find keywords, produce articles" feel incomplete. We think the opportunity is bigger than SEO. It's about understanding what your market is asking, wherever they happen to be asking it, and becoming part of the answer.

Content shouldn't start with keywords

Most content systems start with the question "what should we write about?" We think the better question is "what does our market want to know?" They sound close to the same thing but they lead to pretty different systems.

A keyword is one signal. A question that keeps showing up on Reddit is another. So is something prospects keep raising on sales calls. So is a competitor suddenly getting attention around a specific problem. So is a question surfacing across AI search. So is a page that's already producing qualified signups.

All of these are demand signals. What we want Kascade to do, eventually, is collect those signals continuously, turn them into opportunities, and then help you actually execute on them.

The missing piece is the learning loop

Honestly, generating another 100 articles doesn't interest us much. AI already made content cheap to produce, and that created its own problem: now everyone can produce more, so producing more stops being an advantage. The interesting problem is figuring out what deserves to exist in the first place.

We want Kascade to close the whole loop. Discover, decide, create, publish, measure, learn, repeat.

If something gets impressions but no real engagement, that's a signal. If a piece starts showing up in AI answers, that's a signal. If five articles bring traffic but one of them consistently produces qualified signups, that's a much stronger signal than either of the other two. The next content decision should be shaped by those outcomes. Over time the system should understand more about your market and more about what actually works for your business specifically. That's the bet.

Why now

AI made creating content cheap. It didn't make building an organic acquisition engine easy. In some ways it made things worse. When anyone can generate content, production isn't the bottleneck anymore. The bottleneck becomes knowing what to create, why it matters, how it connects to everything else you've published, where it should show up, and whether it produces any business at all.

At the same time, discovery is fragmenting across traditional search, AI answers, communities and whatever comes next. We think that leaves room for a new kind of system. Not another AI writer. Not another SEO dashboard. Something that understands demand, executes against it, measures what happens, and gets better from the results.

But first we have to prove it

There's an obvious problem with building a product that promises to help companies grow through organic acquisition: we'd better be able to use it ourselves.

So Kascade is going to be our first customer. We're starting with almost no meaningful organic distribution. Over the next 90 days we'll use the exact same process we're building for customers to grow Kascade itself, and we'll write up what we try, what works, what doesn't, what ranks, what gets picked up in AI search, what drives signups and, in the end, what turns into paying customers.

The goal isn't 100,000 visitors. It's simpler than that. Can Kascade acquire customers for Kascade? If we can do that repeatedly, we'll know we've built something useful.

This is where we're starting.