← ALL CASE STUDIES
FüreverWell (Oömph!)

FüreverWell — A Two-Person Company Hired the Best Working Dog in E-Commerce

A ROCKET NOW BUILD · SHIPPED & IN PRODUCTION

Ask any farmer why they keep a working dog and you'll get the same answer: it does the jobs you can't be everywhere for. It watches the gate at 2 a.m. It herds what wanders. It never forgets a face, never misses a shift, never asks for a weekend.

Now meet FüreverWell: a founder-led canine wellness brand — Oömph!, a PhD-formulated daily gut-and-vitality supplement developed by animal nutritionist Dr. Leroy J. Hanson — with a real product, a real brand, a fulfillment partner… and a two-person team facing the math every supplement company faces: the business only works on repeat purchase, and a subscription business generates operational surface area a two-person team cannot absorb by hand. Dosing questions at midnight. "Where's my order?" emails. Support living in a personal inbox. Lifecycle marketing that didn't exist. The ceiling on growth wasn't demand — it was operational capacity.

So Rocket Now built them a working dog. The whole kennel, actually.

The Build: One Animal, Every Job

The delivery is a single full-stack application — ~19,000 lines across 133 files, 14 RLS-secured tables, 80 typed server functions, 18 managed email templates, four integrated systems — that is simultaneously the storefront, the subscription engine, the account portal, the support desk, and the founder's entire back office. One animal, every job on the farm:

It answers the door — at 2 a.m. Rex, the on-site AI agent, is a first-class part of the shipped product, not a scripted widget. He answers product, dosing, and shipping questions in a streaming chat — and when a customer wants a human, he does something about it: he calls a real create_support_ticket tool against the production database, collects an email from guests first, and hands back a deep link into the correct thread, tagged so the admin queue knows Rex made the handoff. Guest conversations persist and are claimed onto the account at sign-up, so support inherits the full pre-signup context. When the front-line question arrives at midnight, then it gets answered at midnight — and if it needs a human, it arrives as a ticket with its history attached, never a cold start.

It herds the sale. The single highest-leverage conversion feature is disarmingly simple: a weight-based dosing calculator that turns "how big is your dog?" into an exact jar quantity and monthly supply, pre-filled straight into the cart — dissolving the number-one pre-purchase question at the exact moment it blocks the purchase. Behind it, live Stripe commerce across three SKUs (one-time jar, monthly subscription, 90-day challenge) with a signature-verified webhook driving order and subscription state — and the retention masterstroke: immediately post-purchase, one-time buyers are offered $10 back, instantly refunded, to convert to a cancel-anytime subscription. An upgrade offer at the point of highest intent, issued idempotently against the original payment so a replay can never double-refund.

It never forgets a name. Eighteen managed lifecycle templates — welcome, order confirmation, shipped with live tracking, invoice paid and failed, subscription started and cancelled, win-back, broadcasts, a daily founder digest — with delivery, open, click, and bounce events flowing back through a signed webhook into an in-app send log. A 20-day re-engagement campaign runs on schedule with a promo code named exactly what it should be: GIVEREXATREAT. And consent is enforced at the query level — opted-out users are never even selected into a batch, so a template bug cannot leak promo mail — with HMAC-signed one-click unsubscribe on every send.

It guards the yard. The security spine this series always insists on, at DTC scale: RLS across all 14 tables, roles in a dedicated table behind a security-definer check, signature verification on every webhook (Stripe, Resend, XPS Ship), server-side role verification before any privileged operation, and money paths built idempotent. The test-to-live Stripe cutover was run as an explicit migration with each buy path individually re-verified — which is precisely how a stale default-price lookup got caught before a customer ever did.

And it reports to one master. The admin console consolidates what used to be three separate logins — Stripe, the fulfillment portal, email — into one surface: a 360° customer view (orders, subscriptions, tickets, notification preferences, even Rex transcripts), an order page with charge and refund history, invoice PDFs, full or partial refunds, retry and resend actions, plus subscriptions, the support queue, and the email engine. The founder runs the entire business from one screen.

The Results

  • From no owned checkout to live Stripe payments across three SKUs on buyoomph.com — with a subscription upgrade offer sitting at the point of highest intent
  • From dosing questions handled one-by-one to a self-serve calculator that resolves the top pre-purchase blocker and pre-fills the cart
  • From support in a personal inbox to a full ticketing desk — with an AI agent answering front-line questions 24/7 and escalating with real tickets and full context
  • From zero lifecycle marketing to 18 automated templates, engagement tracking, a daily digest, and a scheduled win-back campaign
  • From three logins to one console — the operational ceiling that capped growth, lifted
  • Delivered as one continuous build in reviewable batches — commerce, email, support, notifications, admin depth, brand polish, live cutover — with the client redirecting priorities between batches

"[Quote pending FüreverWell sign-off — suggested angle: hours per week returned, and what an agency would have quoted for this scope.]" — Dr. Leroy J. Hanson, PhD, FüreverWell

Why It Worked: The Right Step Was One Animal, Not a Kennel of Tools

The Rocket Now principle, house-trained: a single step in the right direction is worth exponentially more than many steps in the wrong one. The wrong steps were the DTC defaults — stitch together a storefront app, a subscription plugin, a helpdesk, an email platform, and a dashboard, then spend the founders' scarce hours being the integration layer between them. The one right step was building the entire customer lifecycle as one system with one data model: the calculator feeds the cart, the webhook feeds the orders, the orders feed the emails, Rex feeds the tickets, and everything surfaces on one console. When the platform is one animal, then it learns the whole farm — and a two-person team gets to grow like a twelve-person one, because the ceiling was never demand.

Is your team's capacity the ceiling on your growth? Get a working dog. We breed them. → RocketNow.com


Editor's note — before publishing:

  • Confirm client approval and replace the placeholder quote; confirm Dr. Hanson as the attribution
  • Conversion, attach-rate, LTV, deflection, and hours-saved figures are deliberately unquantified — the source docs mark them for backfill from live analytics; add only measured values
  • The review wall is currently seeded per the roadmap — avoid citing customer reviews as social proof until verified-purchase reviews replace it
  • Delivery-scale figures (lines, tables, functions, templates) are codebase-verified and safe as stated
Your build next

Start with a Blueprint. Get the quote as a document.

Start a Blueprint →See what it costs