Andrew Melián

Helium 10 · 2024–25

AI Advertising

Helium 10’s first AI product: fully automated Amazon advertising for small sellers. I led the design team from discovery and problem definition through MVP scope.

Role
Product Design Manager
Team
My design team, Product, R&D, and GTM
SMB adoption
36%
Ad spend / month
+$1.8M
Faster setup
7×
AI AdvertisingProduct Goal DetailsLearnAll TimeX-Large Saucer Chair, Oversized Folding ChairOverviewStart dateJan 26, 2024StatusEnabledACOS goal30.00%Daily budget$150.00GoalACOS goal30.00%ACOS27.45%BudgetDaily budget$150.00Utilization95%ROISpend$150.00Sales$1,250.32OrdersSponsored200Organic100

The problem

Ads was core to Helium 10’s growth plan, but revenue had stalled and small sellers were churning. The product was rules-based automation built for agencies and enterprise advertisers.

Sellers told us they spent 1.5+ hours a day learning PPC, learning our tool, and managing campaigns. They wanted automation that took little effort and worked on small budgets.

How I led it

  • Framed the problemI worked with leadership early to define the problem, agree on priorities, and surface our riskiest assumptions before committing to a solution.
  • Kept it groundedI led discovery through user interviews, competitor benchmarking, and usage data, keeping the strategy grounded in user needs on an aggressive timeline.
  • Made the AI callI compared AI with the alternatives instead of using it for its own sake. It won buy-in because it adapted to each seller’s goals and built on years of our campaign data. No competitor offered AI with rules-based automation.
  • Scoped and shippedI set direction and ran critique for the design team. I scoped the MVP with Product and R&D, then supported GTM through launch.

Key decisions

AI with a choice

Sellers could hand campaigns to the AI or keep rules-based automation. Giving them the choice built trust instead of forcing it.

Ask only for what matters

Setup asks for just a product, an ACOS goal, and a daily budget. We traded control for guidance and guardrails.

Progress against the goal

Performance shows trends against the seller’s own targets, not walls of numbers, and settings can be changed right where results are reviewed. Less detail meant setting clear expectations.

Screens

Outcome

One of the fastest-monetizing launches in Helium 10’s history: 36% SMB adoption and $1.8M a month in incremental billable ad spend.

Setup became 7× faster, and sellers save hours by letting the AI manage their ads.

Links

Next case study

Get in touch