Game Theory 

New locations would take ~25 weeks to ramp up to decent utilisation. Built the playbook (pricing to impact preference & fill rate) to compress ramp up to 2 weeks.

Advisor, Growth

GrowthX®

Part of the early team, eventually grew to lead the Learning Experience vertical. Our charter was to scale the membership from an 8-week program into a high-value community. We started by driving an industry-leading NPS of 78, which was a core signal for referrals. Over time, we increased the value of the membership by co-building learning programs with industry leaders. This allowed us to scale pricing. We understood (fast) that India has a very low willingness to pay high subscription fees in education. The team solved that with a bunch of unbundling & GTM experiments. GrowthX taught me to break down the largest problems into tiny solvable chunks & I'm forever grateful for it.

Head, Learning Experience

Hoop Flo

Joined as the second team member at Hoop Flo, working alongside Eshna Kutty (one of India’s first Nike ambassadors & a leading flow artist). We combined content, community & commerce to make Hoop Flo the #1 destination for hooping in India. Over 12 months, we grew from a few tens to 1,000+ engaged members. In hindsight, what we built wasn’t just a brand, it was a movement. Members genuinely bragged about being part of Hoop Flo.

Founding Member

A few months after an ACL injury, I decided I wanted to do a Himalayan trek. On paper, it didn’t make sense. But, I was determined to make it happen. I got up every day, tracked every little thing on a sheet, adjusted the plan when my knee flared up & stacked the workouts day after day until I had the endurance I needed. At the end of 5th week, I had summitted a smol (but mighty) Himalayan peak.

I am that person. Once I commit to something, I’ll figure out what it takes, put in the work and get it done.

This has been a consistent theme with work too. True when I was building GrowthX (the most transformational & access-dense community for business leaders & founders in India), true when I helped build Hoop Flo from scratch (turning an experimental idea into India’s #1 place for all things hooping), true with starting Recess Club (a community that plays every Saturday at 7AM).

I’m told I’m a generalist. But, I really think of it as problem solving. I like solving problems where there’s no handbook, no structure, just fear (lol) and ambitious goals.

That’s where I do my best work.

We hosted the 10th recess break today. That's 40 of our friends (old and new) coming out to play, every Saturday at 7AM.

The idea has never been to expand this group massively. But to keep it limited to people who truly want to move, play and have fun for the love of it.

Honestly, this is the farthest away from a networking or a gym bro circle. There are no benchmarks. No oneupmanship. Just friendly competition and banter.

We play, just for the vibes of it.
On to many more recess breaks ✨

Ishita Agarwal

In my last role, I led two super important 0 → 1 projects. One failed, one succeeded. Here are some learnings:

The two projects: 

Project 1: Hey, can we take this existing unpaid feature & unbundle it to make revenue? Day 1. I’m excited! First, I need a Notion to make sure everyone is aligned. Day 10. Analysis-paralysis. Ok I think I have the structure. Let me just do a little more research. Day 20. Imposter syndrome. I have no idea what I’m doing. I'll get fired. Ugh.

Project 2. Hey, for our WAU metric, I think hosting great leaders in the community weekly will be great. Day 1. Action. I’m not sure what the format should be, or who these leaders are. How does one even go about this? Let me go ask some people & speak to users. Day 3. Calendars blocked, topic closed. Now we have to announce it so people show up. Day 7, shipped.

Now, what was different?

1/ Commit to a deadline before knowing the "how"

In project 1, I had too much freedom without constraints. In Project 2, we decided we wanted to host the first event in 7 days. Booking the calendars of members before closing the speaker was scary. But we had now made a user commitment, we had to figure it out.

When you’re doing new things, setting the deadline & then figuring it out always works. You could make the commitment in different ways: book the venue, book a meeting for a team review, just pick a launch date.

2/ Make small decisions, iterate as needed

We’re taught to plan → iterate plan → build. I tried that. But, hovering at 30,000 ft gave no feedback. In reality: build → iterate → plan works. In the second project, blocking the date forced us to figure leaders we want to get, a host, structure & the GTM in time. 

By shipping tiny pieces, we wrote the strategy on the fly & could judge it against reality, not theory. Eventually, we were hosting great leaders every single week. We built the playbook by doing. Drafting the final launch copy, landing page,  hero image kills analysis paralysis & makes things tangible. Add a ‘beta’ label on anything new; the word alone lowers the stakes.

3/ Make your project more people’s problem

In Project 1, I acted like a lone-wolf. I had to do it by myself to prove my worth. This threw me into a spiral of self-doubt. In the 2nd project, I asked for help (loudly, clearly, often) & the project became the team’s project. That allowed me to get more input & support. In the end, when it was a success, it became a shared win. Way more fun, trust me.

Getting buy-in isn’t automatic though.
Here are three things that work:

1/ Get the founders to buy-in to your idea first. Their urgency becomes everyone’s urgency.

2/ Bring your manager along for the ride. The more they know about what you’re doing and your challenges, the more they will figure out how to support you.

3/ Use public slack channels, tag specific people for asks. Don't DM for favours, it tilts incentives.

Launching something new today
Canvas by Recess Club ✨

Every time you open a new tab, it’s the same thing: a blank page with your search history.

I personally don’t want to switch to a new tab & see my search history or analytics all the time. I wanted something light. Something I could play around with. Something I could scribble on. Like my softboard at home or the back of my notebook.

Which is why I built Canvas.

A playful new tab plugin. It turns the blank screen you open dozens of times a day into a small pocket of creativity.

You can doodle, add quick to-dos, quotes, images, checklists, whatever you want. And it’ll stay right there every time you open a new tab.

Personally, I’d been itching to build something & this started as a small tinker project. I’m not a technical person at all. But, I knew I could probably whip something up with Lovable. And, genuinely my mind was blown.

But, vibe coding isn't cake walk. As yet. It took me a little more than a week to ship this. About 60% of the time went into solving bugs. I have zero coding clue so I used GPT to debug, rewrite code & point me in the right direction until it finally worked.

For the first time, I don't feel as handicapped as I used to for not knowing coding. And the itch to create has only gotten stronger.

And yes, canvas isn't bug free. But it’s live.

Go ahead, give your browser some much needed play time. The link is in the comments ⬇️

AI launches just in the last 2 days:

- Chatgpt + Commerce
- Sora 2
- Claude Sonnet 4.5
- Perplexity Search API
- Lovable Cloud + AI
- Bolt v2
- Granola Recipes

And this is just from my feed.
We're not even in techtober as yet 🤯🤯

Who knew lagori could be so intense.
12 of us played lagori on the 11th recess break.

My biggest takeaways from Nikita Bier on how to consistently go viral:

1/ Chasing latent demand. Nikita asks to look for places where people are trying to obtain value, but are going through a messed up process to obtain it. If you can compress their motivation into a product: adoption tends to be explosive.

2/ Building where invites are abundant. Interestingly, the older you get, the lesser you invite people to use an app. Invites drop ~20% each year per user from age 13→18 and collapse after they turn ~22. Teens are the only ones referring in abundance.

3/ Sequencing validation like a checklist. Make one thing 100% true at a time. Example: – Do users love the core loop? – Will it saturate a school/peer group? – Will it hop to new groups? – Will people pay?

Nikita explains to think of it as: If X is true, what else must also be true? Then stack the next layer. The more layers you add, the higher the risk, so keep it to 4 core truths max.

4/ Engineering the spark, not the wildfire. Nikita’s playbook is to saturate one school with ~100 users (IG follows, hyperlocal ads, targeted outreach). If it spreads, the product carries the wildfire. If it doesn’t, no amount of seeding will save it.

5/ Delivering a 3-second aha. Nikita says: in consumer apps, you only get a few seconds before users bounce. Your product must show value immediately. If users don’t get value in 3 secs, they won’t stay for the rest.

6/ Treating branding as part of growth. For example: his app, “Crush” (pink) was not shared by males. When rebranded to Gas (black flame), send rates lifted exponentially.

7/ Assuming the growth stack is temporary. Nikita's last app's SMS invites became illegal by the time Gas launched. iOS contact permissions now make sharing harder. Expect regulations to change the rules & keep reinventing how you spread.

8/ Nikita's viral consumer playbook: (1) spot latent demand in distorted user behaviour, (2) design the pixels, copy & flows yourself, (3) ship many small apps fast (15+ before TBH hit), (4) seed into tight, dense groups, (5) use live support for instant feedback, (6) stack learnings with a repeatable testing process, (7) ruthlessly cut features. 

According to him, this exact process is what turns experiments into breakout consumer hits.

Andrej Karpathy on how old technology vs AI is getting adopted: 

Most breakthrough technologies followed a path:
- First, used by governments (military, space)
- Then, adopted by corporations (industry, enterprise)
- Finally, consumer (computing, internet, GPS).

But, LLMs are different.

Instead of trickling down top to bottom, they’ve emerged in the hands of everyday people first. Interesting & unprecedented inversion.

Adding the link to the talk by Andrej at YC in the comments:

AI adoption is happening bottoms-up.
Individuals try → teams adopt → companies purchase.

Bottoms-up adoption starts with individual contributors who love and spread your product. Top-down starts with execs and budgets.

Communities are what make bottoms-up possible: early adoption, fast feedback, organic growth, brand love.

We’ve seen this pattern before:
- ⁠Figma → Friends of Figma
- ⁠Notion → Notion Pros
- ⁠Airbnb → Airbnb Community
- ⁠Canva → Canva Creators
- ⁠GitHub → GitHub Community

AI tools are leaning into the same playbook that Figma did. We're already seeing this with products like Lovable, where their community just hit 100,000 on discord.

It'll be interesting to see which AI communities come through in the next few years.

P.S. If you want to read about bottoms up adoption & GTM, Claire Butler has broken it down beautifully. Linking it in the comments

My biggest takeaways from Andreessen Horowitz's podcast on how AI will change online shopping:

1/ Google isn’t screwed (yet). Interestingly, they say that Google could lose 95% of search volume & still grow as long as they keep monetizable queries (eg: best dress for date night). AI is eating the useless search queries first (“what's the weather in Bengaluru”) + Google doesn't make money on those.

2/ Not all commerce is equal. A16Z breaks down shopping into 5 parts:

– Impulse buys (chocolate near checkout, Insta tees)
– Routine essentials (detergent, diapers)
– Lifestyle buys (bags, skincare, home decor)
– Functional buys (laptops, bikes, furniture)
– Life buys (house, wedding, college)

They think AI is going to eat the middle three. Impulse is too in-the-moment. Life purchases are too high-stakes.

3/ Agents will stop watching and start acting. CamelCamelCamel today: “We’ll alert you when the price drops.” Tomorrow: “We’ll just buy it for you at that price.”

4/ Teens are already hacking the system. They’re pasting Taylor Swift concert pics into ChatGPT & asking “what sweater is this?” → buying the cheaper dupe. It's messy today, but clearly the future of trend-led shopping.

5/ A16Z says aggregators will still win. A single brand can’t ride every trend (think New Balance one year, On Running the next). But, Amazon and Shopify can ride all of them. With agents directing purchases, expect even more power to shift toward aggregators over single-SKU brands.

The thing is, it's starting already. Recently, ChatGPT launched checkout inside chat. Exciting times, curious to see where this goes over the next few months. Adding the link to the podcast in the comments:

Taplio

Ishita Agarwal's Linkedin Analytics

's Best Posts (last 30 days)

Want to drive more opportunities from LinkedIn?

Taplio

.css-1jgreyo{background-image:linear-gradient(284deg, rgba(39,43,72,1) 0%, rgba(33,47,183,1) 100%);color:transparent;-webkit-background-clip:text;background-clip:text;}Ishita Agarwal's Linkedin Analytics

's Best Posts .css-qqfgvy{font-size:var(--chakra-fontSizes-md);}(last 30 days)

Want to drive more opportunities from LinkedIn?

Ishita Agarwal's Linkedin Analytics

's Best Posts (last 30 days)