Conversation
Highlights

How Remark Is Rebuilding E-Commerce From the Ground Up — One Shopper at a Time

Walk into any well-run retail store and something invisible happens. Someone asks a few questions, steers you toward the right product, handles your objections, and makes the whole thing feel effortless. You leave having bought exactly what you needed.

Now open a browser and visit any brand’s website. Generic homepage. Nav tree. Grid of products. A chat widget that promises help and delivers frustration.

That gap — between what retail figured out a century ago and what e-commerce has never solved — is the entire premise of Remark. In a recent episode of BUILDERS by FrontLines.io, Theo Satloff, Co-founder and CEO of Remark, walked through why that gap still exists and what it actually takes to close it.

The Problem That’s Been Hiding in Plain Sight

Theo’s observation is blunt: “If you really think fundamentally about e-comm, it hasn’t really changed since 1996. You still have kind of a generic homepage that you land on when you visit a brand’s website. You still have a nav tree that you can kind of follow.”

Pages load faster. Design is more polished. But the fundamental dynamic is unchanged — a static experience designed for the average visitor, optimized for no one in particular. Every session starts from zero. Nobody is learning anything about the shopper in front of them.

Theo’s window into this problem came from his time at Amazon, where he was building early LLM prototypes for Alexa — specifically testing whether small language models were suitable for voice interactions. By 2020 and 2021, it was clear the technology had crossed a threshold. A genuinely personalized, consultative shopping experience was no longer a research problem. It was an engineering one.

Remark is what he and his co-founders built.

Not a Chatbot — The Distinction That Matters

The first instinct most people have is: chatbot. Theo anticipates it.

“Chatbots are cheap. The moment you start chatting with it, you know it’s the most annoying thing ever. And the immediate response that we all have is, get me to a human.”

Remark takes over a larger surface on a brand’s website — styled images curated for the specific user, interactive cards that let shoppers change an order, reset a password, or track a shipment without leaving the interface. Beyond discovery, it handles inbox automation, live on-site content creation, and proactive outreach on behalf of brands.

But the deeper differentiator is what Theo calls the interview. “We basically interview every shopper — making sure that every shopper is given a proper voice and that they’re not being taken through these click trees of generic autoresponders.” Knowing a shopper is buying a gift for their girlfriend’s birthday or heading to Colorado to ski for the first time changes every recommendation that follows. That individual context is the actual product.

The ICP Assumption That Was Wrong

Early on, Remark built for high-consideration categories — bikes, skis, technically complex gear where guidance feels necessary. That assumption didn’t survive contact with the customer base.

“What we found is while those product types certainly do need guidance, they’re not the only category that needs guidance,” Theo says. Darn Tough Socks became a clarifying customer. “They sell socks. Yes, they happen to be expensive socks, but it’s socks — they still have an enormous amount of friction before a user checks out.”

Today Remark works across luxury, home furnishings, beauty, skincare, and medical devices. The signal that unifies them: “Any sector that evokes some sort of emotional response during the purchase journey is a great fit for this type of technology.”

That’s a more durable ICP filter than category or price point. Emotional friction — not SKU complexity — is the variable that predicts whether a shopper needs guidance. Theo arrived at it through customer evidence, not hypothesis.

A Sales Motion Built on Proof, Not Persuasion

Getting brands to change their web experience is a high-friction sale. These companies have invested heavily in how their site looks and functions. Theo doesn’t try to win on vision alone.

“We can take them through a reduced cost proof of concept period and then run an A/B test that can prove to them that against their existing solution, Remark’s way of thinking about commerce can drive 10, 12, 15% more revenue for them in these really true and fair controlled experiments.”

What makes this motion work is what it doesn’t ask of the buyer: belief. The experiment design removes the need for a leap of faith. The brand keeps its existing setup as the control; Remark runs against it. The data closes the deal.

There’s a parallel insight in how Remark gets positioned in the budget conversation. Some buyers fund it out of customer service software spend — a line item next to Zendesk, Intercom, or Gorgias. Others frame it as a headcount replacement: “Instead of hiring temps during peak season, they’ll just allocate that budget to us.” Theo is explicit that knowing which bucket you’re drawing from before you walk into a sales conversation changes your champion, your framing, and your competitive set entirely. It’s a variable most sales teams discover mid-cycle. Remark maps it in advance.

Old School Selling in an AI Outbound World

Ads underdelivered. Mass cold email underdelivered. What replaced them is what Theo calls “old school selling” — handwritten notes, deep research on every prospect before first contact, and gifts chosen with enough specificity that they stop reading as outreach.

“We will research in depth about each contact before we even send them an email. The response rate that we get from doing cold outreach is a lot higher than you’d see as an industry norm, because it actually isn’t that cold. It’s a lot warmer.”

The chef’s knife campaign made the mechanism visible. Theo’s team sent custom Japanese chef’s knives to top prospects. “When we sent the knives, everyone on LinkedIn was posting about the knives. And even today, we still hear from folks who received those.” The knives worked not because they were expensive but because they signaled genuine attention in an environment saturated with automated personalization. In a world where AI can generate a sequence that looks personal in seconds, actually being personal is the differentiator.

Their strongest inbound channel requires no outreach at all. Brands discover Remark while browsing a competitor’s site, go through the experience firsthand, and reach out. “We know that they found us in the field and they’ve already experienced what makes us special,” Theo says. It’s pipeline the product generates by being good in public.

The Contrarian Positioning Bet

The most deliberate GTM choice Remark is making heading into 2026 is one of scope — and it runs against the dominant trend in AI marketing.

“We’re getting kind of bored and overstimulated by the same repeated AI taglines that every brand has right now. AI is going to change your workflow. AI is going to automate your email. It’s like, okay, great.”

The standard response to that noise is to go broader — to position as a horizontal platform that works for every use case. Remark is going the opposite direction. “We want to be really good at making the customer journey for e-comm exceptional. And we don’t want to do everything. We don’t want to be the thing that makes you a doctor’s appointment.”

The strategic logic is straightforward: when every AI company sounds identical, specific and credible beats broad and forgettable. Owning a lane — and making it exceptional — is a more defensible position than competing for the widest possible surface area. In a market where buyers are overwhelmed with generic AI claims, Theo is betting that clarity compounds.

Listen to the full conversation with Theo Satloff on BUILDERS by FrontLines.io.