Why I built Ripple

I've built a lot of brand decks. 12 to 30 pages each. Mission, vision, positioning, buyer profiles, a few pages on fonts and colors.

They sat on a shelf or in a document repository, and no one looked at them. And AI can't use them well, because they aren't enough context. So I built Ripple.

What happens to a brand deck after you hand it over?

You know how this goes. The team reads it once. Then it gets filed.

Most of the time, someone reopens it for the fonts and colors. But nobody goes back for the rest, because the rest keeps moving. The core positioning stays. The messaging, the market, the competitors, and the campaigns don't. Brands grow and evolve, and a deck can't keep up.

And a new hire needs onboarding, and there are 5 versions of the ideal customer profile. No one knows which one to use. Sales needs a flyer and can't tell which version is right. Product releases new features, but sales, marketing, and customer support are each using their own messaging. I built those decks, so I know this problem. I wanted to solve it.

What does drift look like?

Somebody's slide always has the wrong logo or the old headline.

I once read a set of RFP responses with the same phone number written three different ways. The company name showed up four ways. Nobody decided that. Everyone grabbed whatever they had.

We've had slop forever. AI just made it faster.

My hunch is that's one reason so much marketing sounds alike now.

Why did I decide to build this?

I led marketing at two startups that sold, as marketing director and CMO. Somewhere after those were done, I decided I wanted to build something of my own.

Then I put in the time to learn AI. Hands-on, with the tools, until I understood where they work and where they fall apart. That's when I realized AI could solve a problem I'd been fighting most of my career.

The problem was this one. A brand that lives in a deck few people open. Or in a pile of folders where nobody knows which version is right.

What happened when I gave AI my brand?

I started with what I had. A voice and tone document, with some design direction. The results were bad.

The AI didn't know my ideal customer. It couldn't understand the buyer. It had no picture of the market. It didn't have my voice fingerprint either. That's how long my sentences run, the words I use, how I open and close. I'd handed it a document about how to sound, and I expected it to know who it was talking to.

Then it happened again. While testing Ripple, I forgot to connect it to ChatGPT. It decided on its own to write a proposal. No context at all. The proposal was terrible.

Nobody had told it who we sell to, what we can prove, or what had changed. So it did what AI does. It filled the gaps.

Here's my take: AI is only as good as the input. And a person has to create that input.

What is Ripple?

Ripple is a brand context engine. In plain words, it's what ChatGPT was missing when it wrote my terrible proposal.

It's where your positioning lives. Your buyers, in their own words. Your proof, with the source next to every fact. Your voice and your approved content. Every AI tool and every teammate works from the same place. When your message changes, it changes there. Once.

We build it from your calls, your decks, and your site. We interview your staff and your customers. Where there's still a gap, we fill it with research and label it as research.

Then ask it for a proposal. It already knows who you sell to and what you can prove. If it can't find a fact, it says so instead of guessing. And it tells you how confident it is in what it gives you.

Every draft gets checked before it goes out. Against your sources and your rules. By a second AI. And by buyer models built from your real customer conversations, who react the way your buyers might.

Where in the loop does a person step in?

AI's value depends on humans in the loop. Judgment without a system doesn't scale. A system without judgment produces the average.

A person builds the foundation with you. That's where the strategy gets right. Then you choose how much a person reviews before anything goes out.

Right now, Ripple is a service. I build it with you, and I'm the human in the loop on every project. My vision is to have a self-serve option in the future. Someone has to look at what the AI made and say this is right, or this is wrong. That's why I sell to marketers.

Why can't a brand sit still?

Brands aren't static. They never have been, but now they really can't be. Your messaging changes. Buyers say something new on a call. A result comes in that surprises you. And the AI tools you set up in March are still working from March.

I want every foundation to get a regular look at what's working. What buyers say and what the results show should make it in, with a person making the update.

What can you try today?

Open the AI tool you use most. Ask it to write a proposal for your company. Give it nothing else.

Read what comes back. Mark every line that's generic or wrong. That list is what your AI doesn't know about you yet.

Then ask whoever owns your brand: is all of that written down in one place?

Want to see it work? Ask Ripple five questions about Pebble →

(If you try the test, message me and share what came back!)