There’s two hills and one path. You have to crest the first one before considering climbing the second. Most teams skip straight to the second hill and wonder why nobody showed up.

AI tools made writing code cheap. This makes “clone the competitor” feel rational. When you look at the math, it looks pretty good: 10 products × $100K = $1M. This many products, something something, this much revenue. Profit!

Distribution doesn’t fork. The products you cloned had distribution baked in. Customers, brand recognition, SEO, word-of-mouth, integrations. When you rewrite their code, you don’t get any of that.

Clones are orphans from birth. They worked in the original context because distribution was inherited. You’ve written perfect code. It just never learned how to find customers.

The Trap

Rdio launched first with a superior product and founders with Skype pedigree. Spotify launched in the US nearly a year behind. Rdio treated free tier as conversion tactic. Spotify treated it as distribution strategy: get everyone in first. Today, Spotify has 600 million users. Rdio is a Wikipedia page.

I’ve had front row seats to the same movie—AI era, different cast. Early-stage company, serially cloning products to sell cheaper. Build costs were plummeting. That part was right. Products were built. Customers didn’t show up.

“It only took a week to build, let’s give it another month” — such a tempting trap.

Experiments running too long become commitments. Killing them feels wasteful, so you don’t.

Twin Peaks, Not One

Basecamp’s Hill Chart is a simple model: every project climbs uphill while you’re figuring it out, crests when you have clarity, then coasts downhill as you execute.

Most teams put Product on the hill. They grind uphill scoping what to build, crest when risks are resolved, coast downhill shipping it. Meanwhile, Distribution is hanging out in the parking lot. It never made it onto the hill. They shipped a product nobody knows how to find.

Distribution has its own hill. It should crest before you start climbing the product hill.

DISTRIBUTION "Can I get someone to care?" PRODUCT "Can I build it?" uncertainty clarity unknowns execution FIGURING IT OUT EXECUTING FIGURING IT OUT EXECUTING 1 2
The Twin Peaks Framework: Distribution has its own hill that should crest before you start climbing Product.

“Can I get someone to care?” comes before “Can I build it?”

How Your Team Runs This

Time-box the Distribution Hill:

  • 2 weeks, not 2 months
  • Single owner, not a team
  • No Slack channel until crested
  • No name/logo/landing page until you have signal
  • Goal: keep abandonment cost lower than continuation cost

The kill trigger:

  • If it hasn’t moved in 2 weeks, it’s not an experiment anymore
  • It’s a zombie. Kill it before it gets a Slack channel
  • Dead experiments don’t rot. You can always resurrect them later.

Role clarity under compression:

  • If everyone’s coding, who owns distribution?
  • PM’s job: Acquire, Activate, Monetize, Retain—not glorified feature QA
  • Weekly check: “What did we learn about distribution this week?”

What counts as “Someone said yes”:

  • A deposit or pre-order
  • A commitment to pilot when it’s ready

What doesn’t count:

  • “We have design partners” (but the value prop is diluted across 7 products)
  • “We raised $10M to figure out our ICP”

Can you describe the acquisition channel specifically? Not “posting stuff on LinkedIn”, more like “CTOs in fintech who post about compliance pain, DMed individually, a few converted to calls.”

If you can’t describe it, you haven’t crested. If the weekly answer to “What did we learn about distribution?” is nothing, you’re building another Rdio.