Revamp Wants to Kill the Merge Tag in Your Inbox

Open your promotions tab and scroll. Somewhere in that pile is an email that opens with "Hi {first_name}" - because the software that sent it never actually rendered your name. That small failure is the whole reason Revamp exists. The San Francisco company, part of Y Combinator's Summer 2023 batch, has a blunt view of most "personalized" marketing: it isn't personalized at all. It's a template with a slot.

Revamp's answer is to hand the writing itself to AI. The platform connects to a brand's Klaviyo and Shopify accounts, reads how each customer actually behaves, and then generates a distinct email for each of them - not a name swap, but different copy, different product picks, different reasoning. The company describes what it builds now as "omnichannel AI agents for customer experience." Two years ago it described itself more modestly, as smarter customer segmentation. The gap between those two sentences is the story.

01 — THE ITCHA problem someone had to live inside

Revamp started the way a lot of useful software does: someone was tired of doing a job by hand. Co-founder Amin Akhtar was the Chief Revenue Officer of a bootstrapped direct-to-consumer brand, and his customer data was scattered across platforms that didn't talk to each other. Building a decent email segment meant exports, spreadsheets, and copy-paste. When he finally wired the data together into a prototype, revenue from email campaigns jumped almost immediately.

That number is the seed of the company. Akhtar teamed up with Stephen Campbell, who had worked at Airtable and studied at USC, and Xinchi Qi. They took the pain Akhtar had felt personally and pointed it at everyone else running an online store.

The insight underneath Revamp is unglamorous and correct: in eCommerce, the cheapest revenue you can get is from a customer you already have. Retention beats acquisition on cost every time. Yet retention marketing is where the least craft usually goes, because doing it well means segmenting thousands of people and writing to each group differently - and nobody has the hours.

02 — WHAT IT ACTUALLY DOESFrom segments to sentences

The first product was about understanding. Revamp stitches Klaviyo and Shopify data together, automatically builds behavioral segments with calculated metrics, and syncs them back into Klaviyo with one click. Work that took a marketer weeks collapses into seconds. That alone was enough to launch on.

The second product is about writing. Revamp's AI agents compose real-time, 1:1 emails based on a shopper's browsing history, past purchases, and preferences. And in April 2026 the company shipped the piece that ties it together: Revamp Campaigns.

Campaigns runs a four-step loop - pick your audience, configure the content blocks and which data sources feed them, preview on sample profiles, then generate and sync. Crucially, it doesn't spit out a finished, frozen email. It generates personalized profile properties that get written back into Klaviyo, so the same generation can power a campaign, a flow, or an A/B test. There's undo/redo, per-block control over data sources, and the ability to cancel or retry mid-run. That plumbing is what separates a slick demo from something a team trusts on a Tuesday.

03 — WHO'S USING ITNames you already know

You can usually tell whether a marketing tool works by whose emails you already recognize. Revamp's early user list reads like a tour of the D2C aisle: Dollar Shave Club, Fellow, Outdoor Voices, and Neuro. These aren't obscure test accounts - they're brands with real lists and real revenue on the line. The target buyer is a direct-to-consumer brand or a marketing agency, and the tooling leans toward companies large enough to feel the pain, roughly the $10M-plus revenue tier.

One of the sharper endorsements is almost a side note. Talking about scaling operations, Campbell described getting back roughly 20 hours a week and reaching "a whole new range of clients we couldn't have targeted before." The through-line of everything Revamp builds is the same: give the team back the hours that hand-personalization eats.

04 — THE BUSINESSPriced like it means it

Revamp's pricing tells you exactly who it's for. Campaigns is usage-based: 1,000 free profile generations a month with no credit card, then two cents per profile after that. No seat fees. No minimums. That's not enterprise pricing with a procurement cycle - it's built for scrappy D2C teams and agencies who want to try it on a real list this week and pay for what they use.

The model also reflects where Revamp sits in the stack. It doesn't try to replace Klaviyo or Shopify. It layers on top of them, reading their data and writing back into them, which lowers the switching cost to roughly zero. You keep your email platform; Revamp just makes it smarter.

05 — THE COMPETITIONEveryone owns the tools; few use them well

Ask who Revamp competes with and the honest answer is "the status quo." Klaviyo already ships segmentation and AI features. Retention and personalization tools like Black Crow AI and Rebuy crowd the same shelf. And plenty of brands still lean on human email agencies and copywriters. Revamp's wedge isn't a feature nobody else has - it's the claim that everyone owns the software and almost nobody has the hours to use it properly.

Directional illustration of personalization depth, not a benchmark.

06 — WHERE IT FITSThe AI layer on the retention stack

Zoom out and Revamp is a small, specific bet on a big shift. AI has made it cheap to generate text that's actually tailored, and the first industries to feel that are the ones drowning in repetitive, data-driven writing. Retention email is a near-perfect fit: high volume, rich behavioral data, and a direct line to revenue. Revamp is trying to be the AI layer that sits between a brand's data and its customers' inboxes - starting with email, expanding into SMS and onsite experiences.

The team is small - around nine people out of San Francisco, YC S23 alumni still hiring founding engineers for AI infrastructure and data pipelines. Whether they become the default retention layer or get absorbed into the platforms they sit on is the open question. But the framing is worth stealing for any "boring" industry: the bottleneck usually isn't the software. It's the human hours. Hand those to an agent, and a tired job turns into a product.