# Experimentation and the product team

Turned one product's testing discipline into a portfolio standard, and grew from owning products to leading four PMs.

- Author: Robert Botto (https://robertbotto.com/)
- URL: https://robertbotto.com/work/portfolio-experimentation
- Published: 2026-10-04
- Updated: 2026-10-04
- Context: Adly portfolio, Apr 2026 to now. I lead the product team.

## Key figures

- **8**: Products on one shared pricing-test standard
- **4**: PMs led across the portfolio
- **10**: Marketplace tests run to a decision
- **145**: Point, three-stage PM hiring rubric

## Context

Each brand in Adly's portfolio has its own PM, team and numbers. Once the marketplace's cohort framework was producing decisions, the next step was making that discipline the portfolio's default.

## The problem

A portfolio only compounds what it can compare. Without a shared standard, a pricing win on one brand teaches nothing to the next, and each PM carries the risk of every test alone.

## What I built

1. A hypothesis and an assigned cohort
2. A fixed test window
3. A decision-grade read
4. Promote to core, or kill

A weekly price testing sync shares what each brand learned.

### A shared standard

A five-step testing framework, presented at the Apr 2026 all-hands. A weekly price testing sync since Jul 2026. A shared pricing tool, built by engineering, adopted across **8 products** by Sep 2026.

### Discipline by example

On the marketplace, **10 tests reached a decision and 4 were killed** inside their windows. Promoting winners and killing losers at the same speed is the standard I ask of every PM.

### Leading the PM team

Since mid-2026, four PMs on the portfolio's other brands work with me weekly, plus a monthly brainstorm on each PM's big bets and monthly one-on-ones. I co-proposed the maintenance-mode tier framework leadership adopted in Apr 2026.

### Hiring for AI fluency

An APM programme with a three-week ramp, and a **145-point, three-stage hiring rubric** that weights AI building and prototyping as essential.

### Research that reframes the question

For one social brand, stated cancellation reasons pointed more to delivery (**42.5%**) than to price (28.6%).

## Results

| Measure | Value |
| --- | --- |
| Products on the shared pricing tool | 8 |
| PMs led | 4 |
| Marketplace tests run to a decision | 10 |
| Tests killed inside their window | 4 |
| PM hiring rubric | 145 points, 3 stages |

### Notes

- Apr 2026 to Sep 2026.

## About the author

Robert Botto is an AI-native product leader and forward deployed engineer. He leads a product team at Adly, a portfolio of internet businesses, builds his own AI products, and advises early-stage and AI-native startups on product, pricing and growth.