Luke Productwalker

Walking the product side of the Force. Building digital products in public. Product, AI, startups & growth.

[Luke Productwalker @product_luke]

@foxtomb232 I had problem — kept building things nobody wanted.
Now I run every idea through https://t.co/bcPABpGWdG first (100 synthetic personas, honest score). It rated last idea 40/100, so at least it’s not a yes-man

[Luke Productwalker @product_luke]

I built an app that evaluates startup ideas.

Then I asked it to evaluate itself.
Score: 40/100. Out of 100 synthetic personas, most said “meh.”

So either I built an honest product or a stupid one. Both are rare in AI.
Try roasting your own idea: https://t.co/bcPABpGWdG

[Luke Productwalker @product_luke]

I built an app that evaluates startup ideas.
Then I asked it to evaluate itself.
Score: 40/100. Out of 100 synthetic personas, most said “meh.”
So either I built an honest product or a stupid one. Both are rare in AI.
Try roasting your own idea: https://t.co/bcPABpGWdG

[Luke Productwalker @product_luke]

Free roast, no signup.

Reply with your landing page URL. I'll run it through https://t.co/aJPSfosS16 -- 50 AI-simulated customers from YOUR target audience -- and post the verdict here:

→ score /100
→ top objections
→ the single most brutal verbatim quote

First 10 replies. Fair warning: 15% of the panel are hard rejectors. They don't do mercy.

[Luke Productwalker @product_luke]

Free roast, no signup.

Reply with your landing page URL. I'll run it through https://t.co/aJPSfosS16 -- 50 AI-simulated customers from YOUR target audience -- and post the verdict here:

→ score /100
→ top objections
→ the single most brutal verbatim quote

First 10 replies. Fair warning: 15% of the panel are hard rejectors. They don't do mercy.

[Luke Productwalker @product_luke]

Free roast, no signup.

Reply with your landing page URL. I'll run it through https://t.co/aJPSfosS16 -- 50 AI-simulated customers from YOUR target audience -- and post the verdict here:

→ score /100
→ top objections
→ the single most brutal verbatim quote

First 10 replies. Fair warning: 15% of the panel are hard rejectors. They don't do mercy.

[Luke Productwalker @product_luke]

Free roast, no signup.

Reply with your landing page URL. I'll run it through https://t.co/aJPSfosS16 -- 50 AI-simulated customers from YOUR target audience -- and post the verdict here:

→ score /100
→ top objections
→ the single most brutal verbatim quote

First 10 replies. Fair warning: 15% of the panel are hard rejectors. They don't do mercy.

[Luke Productwalker @product_luke]

Free roast, no signup.

Reply with your landing page URL. I'll run it through https://t.co/aJPSfosS16 -- 50 AI-simulated customers from YOUR target audience -- and post the verdict here:

→ score /100
→ top objections
→ the single most brutal verbatim quote

First 10 replies. Fair warning: 15% of the panel are hard rejectors. They don't do mercy.

[Luke Productwalker @product_luke]

Day 5 of building Prolpy: money time.
✅ Stripe live -- $29/mo (20 tests) or $79/mo (100 tests, bigger panels)
✅ New test type: feature validation -- 50 synthetic users tell you if anyone would actually use that feature you're about to spend a month building

Honest math: A real focus group costs $3,000+. Tomorrow: I run my own ideas through it. Publicly. Including the scores I won't like.

[Luke Productwalker @product_luke]

The persona engine works.
First real test: 50 synthetic customers evaluated an idea. Time: ~2 minutes.

Best part -- the haters are real. One persona, a CFO type, rejected it with: "AI pretending to be a CFO... is not research, it's a hallucination with a dashboard." My own product roasting itself. That's exactly the skepticism I wanted.

Day 4: building the report UI.

[Luke Productwalker @product_luke]

Day 3 of building Prolpy: shipped!

✅ Auth (Google login)
✅ Credits system
✅ Dashboard live in production

Tomorrow: the persona engine. 50 AI customers with enforced skepticism -- how many of them will hate your idea no matter what? We will see... That's the point. Real focus groups have haters. So should synthetic ones.

[Luke Productwalker @product_luke]

I've built 5 products. Abandoned all of them... 3 days ago I decided #6 will be different - built in public, from day one.
Prolpy: test your idea on AI-simulated focus groups before you build it. 100 synthetic customers, results in minutes.
Day 1: landing page live → https://t.co/jyMbp84coV Next: building the persona engine and publishing accuracy tests (synthetic vs real users) as I go. Follow along.
Especially if I fail.

[Luke Productwalker @product_luke]

I've built 5 products. Abandoned all of them... 3 days ago I decided #6 will be different - built in public, from day one.
Prolpy: test your idea on AI-simulated focus groups before you build it. 100 synthetic customers, results in minutes.
Day 1: landing page live → https://t.co/jyMbp84coV Next: building the persona engine and publishing accuracy tests (synthetic vs real users) as I go. Follow along.
Especially if I fail.

[Luke Productwalker @product_luke]

I've built 5 products. Abandoned all of them... 3 days ago I decided #6 will be different - built in public, from day one.
Prolpy: test your idea on AI-simulated focus groups before you build it. 100 synthetic customers, results in minutes.
Day 1: landing page live → https://t.co/jyMbp84coV Next: building the persona engine and publishing accuracy tests (synthetic vs real users) as I go. Follow along.
Especially if I fail.

[Luke Productwalker @product_luke]

I've built 5 products. Abandoned all of them... 3 days ago I decided #6 will be different - built in public, from day one.
Prolpy: test your idea on AI-simulated focus groups before you build it. 100 synthetic customers, results in minutes.
Day 1: landing page live → https://t.co/jyMbp84coV Next: building the persona engine and publishing accuracy tests (synthetic vs real users) as I go. Follow along.
Especially if I fail.

[Luke Productwalker @product_luke]

I've built 5 products. Abandoned all of them... 3 days ago I decided #6 will be different - built in public, from day one.
Prolpy: test your idea on AI-simulated focus groups before you build it. 100 synthetic customers, results in minutes.
Day 1: landing page live → https://t.co/jyMbp84coV Next: building the persona engine and publishing accuracy tests (synthetic vs real users) as I go. Follow along.
Especially if I fail.

[Luke Productwalker @product_luke]

I've built 5 products. Abandoned all of them... 3 days ago I decided #6 will be different - built in public, from day one.
Prolpy: test your idea on AI-simulated focus groups before you build it. 100 synthetic customers, results in minutes.
Day 1: landing page live → https://t.co/jyMbp84coV Next: building the persona engine and publishing accuracy tests (synthetic vs real users) as I go. Follow along.
Especially if I fail.

[Luke Productwalker @product_luke]

I've built 5 products. Abandoned all of them... 3 days ago I decided #6 will be different - built in public, from day one.
Prolpy: test your idea on AI-simulated focus groups before you build it. 100 synthetic customers, results in minutes.
Day 1: landing page live → https://t.co/jyMbp84coV Next: building the persona engine and publishing accuracy tests (synthetic vs real users) as I go. Follow along.
Especially if I fail.

[Luke Productwalker @product_luke]

I've built 5 products. Abandoned all of them... 3 days ago I decided #6 will be different - built in public, from day one.
Prolpy: test your idea on AI-simulated focus groups before you build it. 100 synthetic customers, results in minutes.
Day 1: landing page live → https://t.co/jyMbp84coV Next: building the persona engine and publishing accuracy tests (synthetic vs real users) as I go. Follow along.
Especially if I fail.

[Luke Productwalker @product_luke]

I've built 5 products. Abandoned all of them... 3 days ago I decided #6 will be different - built in public, from day one.
Prolpy: test your idea on AI-simulated focus groups before you build it. 100 synthetic customers, results in minutes.
Day 1: landing page live → https://t.co/jyMbp84coV Next: building the persona engine and publishing accuracy tests (synthetic vs real users) as I go. Follow along.
Especially if I fail.