Ravish Agrawal

Growth @GammaApp | Built & scaled Consumer App 0 to 3 Mn+ Users | @ycombinator Founder | @ForbesUnder30

[Ravish Agrawal @ravishagrawal]

We’ve somehow convinced an entire generation of founders that the only way to start a startup is to apply to YC, SPC, or Antler.

To be honest, when someone at seed asks me how to fundraise, that’s my default advice too,“go apply somewhere.”

YC and others truly solved a problem and accelerated early founders.

But it still feels a bit awkward. A lot of this used to be permissionless or at least, there were many other ways to figure it out. I’m not sure when that changed.

[Ravish Agrawal @ravishagrawal]

@itsajinkyaraj 💯

[Ravish Agrawal @ravishagrawal]

@SimplyColeB Haha, we all have been there😅

[Ravish Agrawal @ravishagrawal]

@suraclaudio yes, pre-pmf is also a great learning playground!

[Ravish Agrawal @ravishagrawal]

The demand for good marketers has never felt higher, but the shelf life of what makes someone a good marketer has never felt shorter.

a lot of the demand for cracked marketers is fundamentally misunderstood.

A surprising amount of it is coming from pre-PMF companies.

Somewhere along the way, we've started believing that hiring an exceptional growth person can change the trajectory of any company.

I don't think that's 100% true.

Good marketing looks easy when the product is great. But even a cracked marketer can't manufacture PMF. And pre-PMF growth is a very different problem from scaling something that's already working.

I think we'll eventually see a correction in how we think about this. But I don't see it happening anytime soon, especially as technology becomes easier to build.

as someone who's been doing this for a long, long time, this shift feels quite personal.

[GREG ISENBERG @gregisenberg]

i think we've reached the point where the cracked marketer is more in demand than the cracked engineer

[Ravish Agrawal @ravishagrawal]

@aidan_nguyentrn 🙏

[Ravish Agrawal @ravishagrawal]

@NakatomiHarry There's no silver bullet 💯

[Ravish Agrawal @ravishagrawal]

@djforge 🙏

[Ravish Agrawal @ravishagrawal]

I also feel that the world wouldn't look the same 3 or 5 years later. My belief is that neither the traditional engineer nor the traditional marketer is where this settles.

We're going to see an entirely new kind of role emerge. Someone who sits at the intersection of product, engineering, and marketing.

Someone who can build, distribute, understand users, and move between bottom-up product decisions and top-down growth strategy.

Not just someone who knows how to acquire users, but someone who can figure out what needs to be built, why people should care, and how to get it into their hands.

Today, that's largely the founder's job. I suspect we'll see more people who can do this without being founders.

We don't really have a title for this role yet. Maybe "builder" comes closest, but I don't think it quite captures it.

And maybe we don't need a title just yet. I think we'll first start seeing more people doing this, and the titles will follow.

[Ravish Agrawal @ravishagrawal]

10 years ago, when I was graduating from IIT, I had zero interest in coding. And I remember feeling like a second-class citizen because of it.

This was the era of Android and iOS apps. Full-stack developers were in crazy demand. Some of the smartest people I knew became engineers.

I stumbled into growth at Udacity, learned a lot building Able Jobs from zero to 3 million users, and now at Gamma, I've been lucky enough to see growth at a completely different scale.

And for the first time in my 10-year career, I genuinely feel marketers are in crazy demand.

At the same time, marketing is changing faster than I've ever seen.

We've borrowed ideas from other companies' playbooks. Other companies have borrowed from ours. Something that works brilliantly today can become completely commoditized in months.

Everyone is trying everything in the hope that something sticks.

It's a very different world from the one I started my career in. And personally, I'm quite happy to be doing what I'm doing right now.

[Ravish Agrawal @ravishagrawal]

@rabi_guha 🙏🚀

[Ravish Agrawal @ravishagrawal]

One trend I've noticed with recent model updates: people are increasingly asking AI to show them things, not just tell them things.

I see this with friends who are heavy AI users, and I've noticed it in my own behavior too.

Between reading an explanation and interacting with something visual, I'd almost always prefer the latter.

And this is where AI products are heading.

It doesn't matter where your agent lives. WhatsApp, SMS, or inside your own product. At some point, agents will need to become more visual and interactive.

We're visual creatures. It's only natural that the way we interact with AI will evolve in that direction.

That's why I'm excited about what @rabi_guha is building with OpenUI. He's one of the best people I know to take a crack at this.

Go Rabi! 🚀

[Rabi Shanker Guha @rabi_guha]

Today, we're shipping OpenUI Dashboards.

Your users are taking their data to Claude to build dashboards. Now your agent can build them - right within your product. Your own components and APIs, any model.

Claude's best feature, now for every agent.

[Ravish Agrawal @ravishagrawal]

@ChrisUniverse @comet By getting there, I meant in terms of positioning and not in terms of product. The idea of this tweet was: the adoption for agents in companies. Positioning will play an important role for that adoption, and they are first ones to move towards that positioning.

[Ravish Agrawal @ravishagrawal]

@MovesInShadows Great internet marketing company was enough to get them where they are, and they will figure out the rest 🤞

[Ravish Agrawal @ravishagrawal]

Canva is in a uniquely difficult spot here. ~$4bn in ARR to protect, while ChatGPT and Claude keep getting better at things people pay Canva for.

They’ve rolled out AI over the last year rather unsuccessfully. But I think we underestimate the inference cost of getting a consumer base that large to actually use it.

Even Google and Meta, which own much more of the stack, are taking on more debt to fund their AI buildout. Owning the infrastructure doesn’t make those costs disappear. They show up in infrastructure spending and borrowing too.

Canva has its own models, but it doesn’t have the same advantages as a big tech. So you’re potentially hurting margins to roll out AI, while the models themselves come after your existing revenue.

I don’t think this is just an execution story. They may have to cannibalize the old business before the new one can make up for it.

Maybe the best time to do that was a year ago. Maybe it’s a year from now, when some of this settles. But today feels like a particularly bad time to have to make that call.

And still, six months’ salary, removing the vesting cliff, and $10k grants for departing employees who want to start something of their own. A pretty incredible package.

Reminds me of what Airbnb did in 2020. Truly a people-centric company.

[Mark Di Stefano @MarkDiStef]

Faarking hell: Canva announces deep lay offs hitting 22 per cent of its employees.

[Ravish Agrawal @ravishagrawal]

@therigger33 nice

[Ravish Agrawal @ravishagrawal]

You have to give Perplexity credit. They’ve generally been ahead of the curve. They were early in AI search, and then again with the browser.

With Teammates, I don’t think it’s as much about being ahead or behind. We’ve seen Muse,Instinct, Dot, Duet, and others coming up. People have been wowed by how proactive these agents can be, and I think they’re now starting to expect that in their work too. This feels like the first wave of that expectation translating into something practical.

I think everybody will move towards some version of teammates. Perplexity is one of the companies trying to get there.

I’ve been a big proponent of verticalized AI. Today, AI is still very horizontal. What you get out of it depends a lot on your imagination, and people still struggle with “What should I actually do with this?”

Package it as “here’s your legal teammate” or “here’s your store manager,” and people already understand the role and what they would hand over.

That alone doesn’t make it verticalized AI. But I think it could drive a lot more adoption, even if the idea isn’t new.

I think we’ll see more and more companies position themselves in this direction, and I’m curious to see how it turns out for Perplexity.

[Aravind Srinivas @AravSrinivas]

Today, we're launching Perplexity Teammates.
They are personal agents for work. And change the paradigm from writing prompts in a session to giving high-level goals. Rolling out to Max users on the web today. Native apps (mobile, desktop) coming soon.

[Ravish Agrawal @ravishagrawal]

@prabhathpani I have used it when the first version came out, but eventually bundling took me away, and now I use Claude and the ChatGPT computer use.

[Ravish Agrawal @ravishagrawal]

@MrBinspire 🙏

[Ravish Agrawal @ravishagrawal]

@scspeier Ohh wow, excited 🤞🤞

[Ravish Agrawal @ravishagrawal]

@victorgu0 GOAT 🚀

[Ravish Agrawal @ravishagrawal]

@charlieholtz Congratulations Charlie!
@aaron_epstein is the best 🚀