Loop: a brand for a product that never shipped

An AI tool built to show its work, in a market full of black boxes.

The problem: Brand a new AI product that didn't exist yet, in a market where every AI tool looked and sounded identical.

My role: Built the whole brand from the story up. Positioning, voice, a content system, a custom typeface, a mascot, and a mascot system that stretched across other products.

The move: Skip the magic-black-box playbook everyone else was running, and build a brand with its work showing.

Every AI tool at the time looked the same. You typed something into a chat window, it handed you a finished thing, and it never told you how it got there. Surprise, here's your magic asset.

Loop was the chance to build the opposite of that.

The idea was big. Loop was going to be the Leadpages people actually wanted, after years of the original sitting untouched. Analytics first, an easier builder, real testing baked in, and AI running through it instead of bolted on top. The plan was to eventually move everyone over to the thing that fit them better. When I saw the early dashboards I liked it for real, because it opened on data instead of a chatbot, and metrics are the thing marketers live on.

Branding a promise

The product didn't exist yet, so I was branding a promise. That meant the story had to come before the logo, and my CEO was already asking for logos.

I started with what Loop actually was underneath the feature list. People were A/B testing everything, messaging and images and layouts, over and over, chasing better numbers. Reviews across every landing page tool said the same thing, that the analytics were a pain and nobody could tell where the numbers came from or what they meant for a campaign. Loop was built to fix exactly that.

So the brand became a lab. Somewhere you run experiments, keep notes, and get smarter about what works. The Living Lab. \

The problem was AI itself

The harder part was AI. Everyone was using it, us included, and that was the trouble. I didn't want Loop to vanish into the pile of identical chatboxes.

So I gave it a personality with a point of view. The reference in my head was Bill Nye. Curious, a little delighted by the data, and happy to be wrong out loud, because being wrong is how a test teaches you anything. No guru statements, no standing on a stage telling marketers what to do. I wanted the tool to show its work, admit when a test bombed, and treat the failure like something worth looking at.

I wanted the tool to show its work, admit when a test bombed, and treat the failure like something worth looking at.

That personality drove a whole content system, built to run in public. We'd document experiments like a lab and publish the notes. It broke into a few repeatable formats:

• Lab Notes: numbered experiments people could follow to learn something in their own field

• Receipts: the actual data, shown plainly

• Hot Takes: our own opinions, formed from what we saw

• Pattern Drops: the long-form data digs

  
  

The plan was an experiments section on the site that doubled as SEO content and as genuine resources. Everything did more than one job.

And I kept the AI in the background on purpose. It ran all through the product, and it stayed out of the headline, because the headline was the part everyone else was fighting over and losing. The lab led. The AI was one of the instruments in it.

The voice needed a face

I went looking for a typeface that felt like the personality. Something that moved like a data chart, academic but excited, whimsical without being a toy. Serifs read too serious. The usual sans-serifs read too obviously tech, the same handful every software company reaches for. Nothing I could license had it, so I drew it by hand. Loop SemiSans. It carried the personality nothing off the shelf could, and it made the logo.

Then Bloop, the mascot. Putting a face on a B2B AI tool sounds like a bad idea until you remember the entire point was making the thing feel human. Bloop was that. Clippy, if Clippy were cute and actually helpful. The waves on him oscillated like a chart with shifting data, so even the mascot was made out of the product's core idea. He was simple enough to drop anywhere, and I planned him as the face of the AI chat agent, so using Loop felt like being walked through an experiment by someone who had your back.

And Bloop was built to scale. He was the first of a mascot system meant to tie the whole product family together through their chat agents. Spark for Leadpages and the smaller companies. Div for HTMLPub. Each one different enough to have its own personality, and close enough that knowing one meant you knew the family. Cohesion you could feel without anyone having to explain it to you.

None of them ever shipped. The whole system was built on a logo that never launched, so Bloop and Spark and Div all stayed on the shelf.

Why it never launched

Loop never launched.

The CEO loved the story and loved the design. The product direction is what changed. He didn't want to bet everything on Loop before seeing whether he could fix Leadpages first, and he had a point. Leadpages had a real following, years of brand presence, and a reputation for being easy to use even while it was hard to control. It just hadn't been touched in a long time. Betting the company on a brand-new product before repairing the one that already has the customers is a reasonable thing to not want to do.

What actually shipped

The product didn't. The parts of it that mattered did.

The biggest one is easy to miss. Loop's analytics and the whole iteration idea, the actual point of the thing, became the newest version of Leadpages. What I built the brand around didn't launch on its own, it got folded into the product that already had the customers, which is the exact bet the CEO was making. Some of that same foundation seeded HTMLPub too.

The mascots stayed on the shelf, but their parts didn't all go to waste. Div's colors carried into HTMLPub, and the little pixels that float off his head are part of what it looks like now.

The bigger thing that carried over was building in public. I pitched it for Loop, and the company started running Office Hours on LinkedIn, still going every week, where the product team openly shows what they're building. That's the clearest proof any of this worked, because you can go watch it. The honesty also worked its way into the email messaging, quieter, but real.

The whole brand was built on the idea of showing your work, and the piece that outlived everything else was the company deciding to actually do that, out loud, every week. Want to check out the full brand overview? Sneak a peak below.

Loop Brand Summary.pdf

Download PDF • 5.14MB