Growth is an ever-changing field, especially now while we are in the middle of this change.
With Lazyweb, I’ve tested a lot of distribution channels and growth tactics.
The results speak for themselves: 50k signups. $0 marketing. A couple of months.
Most of it didn’t work. Some did and a couple broke out.
Here’s everything we tried so far..
Sharing my story on X —
Great for customers, investors, and product feedback. The MCP launch post reached 365k+ views.
Ali Abouelatta@abouelatta_ali
I hated how bad agents are at design
I hated how Codex can’t access Mobbin
So I created Lazyweb
– 257k+ screens (apps/web)
– 6 opinionated design research skills
– 1 MCP (Claude/Codex)
100% Free…AI native…no rate limits…no subscriptions..
Enjoy (and tell a friend)
2:14 PM · May 2, 2026 · 366K Views
191 Replies · 376 Reposts · 5.49K Likes
Sharing my story on Substack —
Great for the soul (: I love writing this newsletter.
Automatic routing to the MCP from agents.md —
Guest accounts for agents —
We removed the email requirement so agents could download, connect, and start using the Lazyweb MCP.
One-line installation command —
Playground for humans before connecting to an agent —
A lot of growth work is just “being out there.” The lazy version of that is spam — Claude generating “9 best customer support AI agents 2026” for the 150th time.
It won’t work. It assumes the labs are stupid. They don’t make money from ads, and they definitely don’t make money from their agents being unreliable.
So what are OpenAI and Gemini optimizing their agents for? Higher quality, faster, cheaper. Not hard to guess.
If you wanna grow, make ChatGPT look good.
Took me a while to land there. I was wired on what used to work.
A free MCP server for our blog, so agents can search it faster
A
.mdversion of every blog post — neutral result.
Changing Hyperlinks → Citations
Both of these we never measured their impact, so I’m keeping the result as unknown.
We, as humans, have a weird bias. We assume people value their time way less than we value our own. They don’t.
What if they steal my idea? What if they steal my data? What if they steal my insights? What if what if what if what if.
Take inspiration from open source. No one wants to put in the effort to roll up their own services.
A free skill on GitHub, skills.sh, etc. —
Free access to data on the web app — — This worked really well for growth, but it became a time sink because of scrapers and DDoS attacks, so we turned it off.
Ali Abouelatta@abouelatta_ali
@edgarasben Since u already pay for all the tokens to do the research (I just provide the skills + data) my costs are rlly low (few hundred $$ / month)
I can afford it (:
6:22 PM · May 2, 2026 · 383 Views
1 Reply · 2 Likes
Ali Abouelatta@abouelatta_ali
I BUILT A GAME TO PROVE IT.
10 REAL EXPERIMENTS. REAL COMPANIES. ZERO HINTS.
50%+ = OG.
BELOW 50% = NEVER SAY “DATA-DRIVEN” AGAIN.
LIVE ON PRODUCT HUNT:
producthunt.com/products/grow-…
8:45 AM · Aug 13, 2026 · 4.35K Views
7 Replies · 11 Likes
/Explain-PR Skill — Fun to build. Didn’t really move any metric.
Ali Abouelatta@abouelatta_ali
YOU SHOULDN’T READ PRS
INSTEAD YOU SHOULD LET GILFOYLE AND RICHARD HENDRICKS EXPLAIN UR (agent’s) PRs TO YOU.
/explain-pr is skill I made to help me swift through the 100 PRs i am pushing with codex every day.
Here is an example for a PR migrating our APIs to bedrock
4:57 PM · Aug 18, 2026 · 1.58K Views
5 Replies · 1 Repost · 10 Likes
/code-map — I didn’t build this one, Francesco did. Same story: it didn’t really move any metric.
Ali Abouelatta@abouelatta_ali
Vibe coders….
Rejoice..u can now understand what the hell your agent built.
Track your product architecture:
– Every Feature
– Every API call
– Every network call
And how they all work together
100% Free. Tell a fellow vibe coder
Ali Abouelatta @abouelatta_ali
Instead of code maps for:
– PR by PR
– Only Pro and Max Subsribers
– Have to invoked everytime you want it
How about:
– Code map your entire app
– Track git-like changes to your architecture over time
– 100% Free
– Works on any coding agent
https://t.co/E6rflDcOik
10:06 PM · Jul 2, 2026 · 1.85K Views
2 Replies · 5 Reposts · 4 Likes
Million Dollar homepage revival , didn’t catch on. Funny enough someone tried something similar last week with outbid.lol but I had one too many battle-scars to give it another shot…that one took off (:
Facebook Groups , tried infiltrating vibe coding groups, got sniffed out
Reddit , got banned from a couple of reddit threads, but when the tweet went viral it started popping off there
X Radar for related convos , got a few hundred MCP from responding to a few threads related to Lazyweb. They got maybe <10k view but view → CTR was very high, in the 20% range
Its always nice doing these retros and reflecting back on what worked and what didn’t because its always surprising at the moment but obvious in hindsight.
Reporting back at the 100k milestone
