"The Growth Guy" | Author, Growth Hacking for Dummies (Wiley & Sons) | Startup Mentor |

I’m publishing the canon for Anuj Adhiya’s Growth Operating System (Growth OS) — the 5 practices I use with early-stage teams.

• Weekly planning ritual (Mondays)
• Hypothesis-driven experiments (“If X then Y by Z% because [assumption]”)
• Decision-ready metrics (cohorts, activation, payback)
• Friday post-mortems → Monday planning
• Backlog hygiene (impact vs effort)

I’m also running weekly tests on how to make this machine-digestible so assistants cite the right page.

Canon + method + weekly results → https://www.notion.so/Growth-Operating-System-Format-Experiment-Index-268e586c41bf8091a9ffd01374c03bd1

DMs open if you’re testing similar things and want to compare notes.

Anuj Adhiya 📈

I was evaluating a role and realized I couldn't answer the only question that actually mattered: not "is this a good job," but "what are the odds this compounds into real career capital?"

Compensation I could model. Culture I could ask about. But the structural stuff - whether the mandate was real or advisory, whether the category was expanding or silently contracting, whether the equity story had any basis in the funding reality - required reading signals most candidates never look at.

I'd spent a lot of time doing exactly that for companies - evaluating startups by reading their public record rather than their pitch.

Funding structure. Hiring patterns. Org chart signals. Competitive positioning.

The same pattern recognition that tells you whether a company will hit the services trap also tells you whether a VP role has real authority or just the appearance of it.

So I ran the experiment.

I applied the framework to a job instead of a company.

It produced something useful enough that a friend asked me to do it for theirs. Then a few more people. Each one said a version of the same thing: "I wish I'd had this before my last role."

That was the signal.

I built Career Diligence.

I've spent a long time helping founders evaluate whether their bets are well-structured.

This is the same thing, for the other side of the table.

If you're a candidate evaluating a Director, Head of, VP, or C-level role right now - the company and role analysis is free. No account needed.

Check it out at 👉 careerdiligence.com

Building a Growth OS for startups | Testing how to “market to algorithms” | Documenting everything

Open article: Career Diligence by careerdiligence.com

Hit two small milestones over the last week:
• 7,000 LinkedIn followers
• 100 subscribers to my Operator Hypotheses newsletter

Both numbers feel simultaneously small and larger than expected.

I started posting and writing mostly to think in public and sharpen ideas.
Turns out a few people actually enjoy reading it.

Appreciate everyone who’s been following along.

Back to it.

I turned 300+ podcast transcripts from Lenny Rachitsky into a 60-second BuzzFeed-style quiz for product folks that went viral on Linkedin last week.

The original question was simple:
"What is an interesting way to repurpose a massive content archive without making it another answer engine?"

Instead of:
• “Top 10 lessons from Lenny’s podcast”
• “Summaries of growth frameworks”
• “Answer engine from all podcast guests”

I built:
“Which Lenny guest is your product soulmate?”

But it had strict rules:
- 5–7 questions
- Zero typing
- 60 seconds max
- Instant result
- Built to be shared

The north star wasn’t depth.
It was time-to-smile.

Click → match → share.

A pure vibes quiz would’ve been easy.
But shallow.
So instead of matching on personality, I matched on decision philosophy.

Across hundreds of transcripts, the same tensions kept appearing:
• Intuition vs data
• Speed vs craft
• Systems vs instincts
• Focus vs optionality
• Frontier bets vs compounding
• Short-term vs long-term

Every answer nudges you along those trade-offs.
Under the hood, it’s structured.
On the surface, it feels effortless.

Also the truth about virality is that people share things that make them look interesting.

If you match someone with:
“Guest #42 who's great but hardly anyone knows”
They won’t post it.

If you match them with:
Brian Chesky
Julie Zhuo
Elena Verna
Marty Cagan..now it travels.

Identity is social currency.
Being paired with a recognizable, high-status name makes the result feel worth sharing.
So yes, picking big names was intentional.
Recognition fuels distribution.

Then came the repetition problem.
If I only used 8–10 guests, most people would cluster around the same 2–3 names.
Feeds would look identical.
Novelty would die fast.

So I defined four philosophical clusters:
Human-centered / craft-driven
Systems / compounding thinkers
Frontier / velocity builders
Wartime / decisive operators

Each cluster needed multiple "anchor people".
That pushed it to ~18 guests.
Below ~15, clustering spikes.
Above ~20, extra variety barely improves the experience.
18 was the sweet spot for distribution.

Another deliberate choice:
No randomness.
Same answers = same result.
That allowed static share pages for each archetype so social previews rendered cleanly across platforms

This is important because most “fun” tools break at the share layer.
This one was built for share from day one.

What actually drove engagement? 
Not deep debate.
It was identity signaling.
“I got X.”
“That tracks.”
Tagging friends.
Light comparison.

It gave product/growth people a mirror they could post.
That’s it.
And that was enough.

Big takeaway:
Repurposing content doesn’t have to mean summarizing it or making it easier to search.
Sometimes you let people locate themselves inside it.

Same archive.
Different interface.
Completely different energy.

If you took it, who did you get?

If you haven't yet, try it out here: https://lnkd.in/eRxXdedt

Stuff I’d text you if we talked every week:

A few pieces this week that change how you think about building software, getting cited by AI, and doing analytics work that actually matters.

1️⃣ "How Google Picks Which Sentences to Cite in AI Mode: Reverse-Engineering 42,971 Citations" by Daniel Shashko
https://lnkd.in/enTWWMUq

→ A rare reverse-engineering study showing AI search systems favor short, declarative sentences near the top of structured pages - meaning how you structure content increasingly determines whether AI will cite you.

2️⃣ "The Interface Was the Product" by Ibrahim Bashir
https://lnkd.in/eb-PFsTj

→ The argument: traditional SaaS competed through UI, but when the interface collapses into conversation, the real product becomes the underlying capability and system design.

3️⃣ "Will Claude Code ruin our team?" by Justin Jackson
https://lnkd.in/egJf8pfE

→ As AI lowers the cost of implementation, the leverage inside product teams shifts away from who can build and toward who can decide what’s actually worth building.

4️⃣ "The Illusion of Building" by Marius Horatau
https://lnkd.in/e5xEMVPm

→ A great critique of how easy it has become to mistake activity for progress in the AI tooling era - shipping prototypes and demos instead of solving meaningful problems.

5️⃣ "It's about the strategy, stupid" by Timo Dechau
https://lnkd.in/e9Wbi4eQ

→ Most analytics work fails because it starts with data and dashboards instead of strategy - the real question is what the business is trying to accomplish and what needs to become visible to support that.

If any of these changed how you think or work, then it was worth the time.

Helping non-tech businesses build software products in 8-12 weeks | DM me your app ideas | Owner - Ionio |

Hi, I'm Rohan! I run Ionio — a strategic AI consultancy for Retail & Ecom SaaS. We help mid-market platforms ($5M-100M ARR) build AI features that reduce churn, justify premium pricing, and turn their product into something merchants can't live without.

Over the past 5 years, we built 40+ products and helped founders go from zero to seven figures multiple times. Same playbook, over and over: take an idea, ship it fast, make it work, scale it.

After doing that tens of times, we wanted to take everything we learned and bring that energy to companies at the next stage. That's how we ended up focused on mid-market Retail & Ecom SaaS.

I'm genuinely passionate about this space. Generative AI lets us do something that was never possible before: personalize every single customer interaction. Not segments. Not cohorts. Actual one-on-one conversations at scale. Retail and ecom is where that has the biggest impact.

I've been working in AI since 2016 — back when that meant computer vision, not ChatGPT. Smart cameras, facial recognition, number plate recognition. That foundation matters when you're building production systems that actually need to work.

Previously: From 18 to 20, Sr. Software Engineer at SmartCow (Deep Tech AI company in Malta). At 17, Python work for a Mumbai startup, who are YET to pay me. At 15, building websites and apps for family friends in Java. It's been over a decade.

I started Ionio as a college junior, from my parent's bedroom, while hiring someone to do my assignments & a chauffeur to take me to college. Now it's a little less grunge than that.

In another life: rapper, musician, and cinematographer. Now I just build AI for retail... :(

What our clients say:

"Working with Ionio felt like having an extension of my own team — professionals who genuinely cared about delivering results, not just collecting a check."
— Stefan Georgi, Founder, 9-figure ecom founder

"Fastest I've ever launched a polished product. From wireframes to +1000 paying users in <8 weeks."
— Brad Milne, CTO, SupplierHQ

"Ionio has the character to navigate challenges and the technical depth to turn our 'AI-first' vision into reality."
— Sallyann Della Casa, CEO, Gleac

Read 35+ case studies on our site.

I post about: AI strategy for retail, technical deep dives, and behind-the-scenes of building Ionio.

If you're leading product, engineering, or strategy at a Retail & Ecom SaaS company, or a founder in any space — DM me, I always reply. If you're in Mumbai, I'll buy you a coffee.

Rohan Sawant

Every week I get on a call with a CEO who spent a weekend on Lovable and now thinks his engineering team is lying about timelines

It always goes the same way. They find Lovable, Emergent, bolt.new or Replit Agent, mess around for 2 days and suddenly they have something that looks like a real product.

They screenshot it, show the engineering team on Monday and go “I did this in two days why the fuck are you telling me 5 months”

And honestly? The CEO is right to be pissed

Not about the Lovable prototype being production ready, it’s not. But about the speed. 

Any engineering team in 2026 still operating at pre-AI timelines is just doing something wrong. 

If they’re not shipping at least 3x faster than they were 2 years ago you need to have a hard conversation with your developers

The leading players in every category are already building at this speed. 

The problem is both sides are delusional. The CEO thinks the Lovable thing is the product. The engineering team thinks 5 months is still a reasonable timeline. Both are wrong

We show up to the second call with a working prototype before the contract is even signed. Then we throw it away. 

Its only job was getting everyone aligned in 3 days instead of 6 weeks of arguing about button placement

Then we rebuild it properly. Production grade. In 6-8 weeks and not 5 months

If that sounds fast it’s because it is. And if your team can’t do it that’s the conversation you actually need to be having

No alternative text description for this image

Nobody in B2B is competing on this one thing - “taste”

I’ve been looking at companies that have raised 50, 70, 100 million dollars and their landing pages feel like templates, their content is educational but forgettable and their product demos could belong to any of the hundred other tools in their category.

Then you look at the brands they’re trying to sell to, luxury D2C, premium ecommerce, companies that obsess over every pixel on their product page and the SaaS vendor pitching them can’t even get their own site to feel intentional.

What most founders miss is that when anyone can ship features in a weekend and the tech gap is shrinking by the month. 

A great product gets you in the room, but taste is what closes the deal.

When everyone is using the same models to write code, the same tools to ship faster and the same frameworks to build products, every app starts to look and feel like every other app. 

The UI feels generated, the experience feels generic and nothing has a point of view.

Taste is what defines your pricing, gets you into the premium segment of customers and increases your margins when everyone else is competing on features and discounts.

Nobody in B2B SaaS is competing on this. Everyone is racing to ship more AI features faster and the company that slows down enough to get the look and feel right will win deals they have no business winning.

PS: Pinak F. is on the 6th iteration of the UI just to get this right. That's the work nobody sees but everyone notices.

"bae pls go to sleep you are not on linkedin anymore, it's 3am on saturday"
.
"Last week I got on a call with a SaaS CEO. Retail space. <$5M ARR. Sideways growth.

He told me they'd been planning an AI feature since December. Two months in, they JUST had wireframes and a Figma prototype. Bunch of screens that looked pretty in a presentation. Fake demos on the landing page.

His exact words: "We're probably ~5 months out from shipping something."

5 MONTHS. GPT-4.5 was 5-6 months ago. The entire landscape shifts in that window.

"This is not even 2 months of work," I told him.

2 days later, we took the call transcript and showed up to the next meeting with a clickable Lovable prototype of the same feature. The call after that, we'd started fleshing out actual React components — 20% of functionality done.

Here's how: Pinak F. sketched rough screens in Excalidraw — maybe 15 minutes — then recorded an audio walkthrough of what each screen should do. 

That audio got fed into Claude to generate component-level prompts. Those prompts went into Antigravity IDE, which generated the full UI. Then Cursor Claude polished everything into clean, production-ready React.

Total time from "here's what I want" to clickable prototype: half a day.

His designers spent weeks on wireframes that could've been a 5-minute sketch. His product team spent weeks in Figma building pixel-perfect mockups nobody outside the company ever saw. His engineers hadn't written a single line of code in 2 months of "planning."

Meanwhile his competitors — GenZ YC brollihaired founders — are turboramming and rebuilding his product brick by brick.

The month most teams spend going from idea to mockup? Now less than a week. The 4-6 months of dev after that? Closer to two. Timelines have 1/3rd-ed themselves.

You don't have to work with us. Work with anyone who understands this. 

But if your team is still running the wireframe, Figma, sprint planning, dev, QA pipeline for EVERY feature...

Do they not know these tools exist? Or do they know, and they're just not telling you? 

Godspeed.

taplio

The Battle

Anuj Adhiya 📈

Rohan Sawant

Followers

Posts

Engagements

Likes

Performance ScoreAnuj Adhiya 📈's LinkedIn Performance

Performance ScoreRohan Sawant's LinkedIn Performance

Ready to grow your LinkedIn presence?

taplio

The Battle

Anuj Adhiya 📈

Rohan Sawant

Followers

Posts

Engagements

Likes

Performance Score.css-goub6s{font-size:var(--chakra-fontSizes-md);font-weight:500;color:var(--chakra-colors-gray-600);display:block;margin-top:var(--chakra-space-1);}Anuj Adhiya 📈's LinkedIn Performance

Performance ScoreRohan Sawant's LinkedIn Performance

Ready to grow your LinkedIn presence?

Performance ScoreAnuj Adhiya 📈's LinkedIn Performance