Even though it's the same technology, ChatGPT is taking off way faster than GPT-3.

A few lessons we can learn from its growth 👇


GPT-3 has been publicly available for more than a year. You can go to OpenAI's website and access it via a UI.

ChatGPT launched last week and grew to 1 million users in 6 days. (Link below 🔻)

So what gives?


Takeaway 1: Constraints are good

By putting some constraints on the UI and forcing the user to input into a chatbox, ChatGPT makes it clearer what the user can do with it.

SaaS companies that keep their product simple and don't provide *too* much freedom for their users tend to do better too.


Takeaway 2: Viral loops are powerful

ChatGPT allows you to 'talk' to the AI. This generates interesting interactions which users then share on social media. This in turn acquires more users for the product.


Takeaway 3: Ship fast instead of pursue perfection

ChatGPT still makes many errors. And it sometimes produces results that are offensive, even with the built-in safeguards.

This didn't stop OpenAI from releasing it. The immense volume of user interaction helps them refine the product much faster than if they chipped away at it for a few more months.


Have you tried it out? What are some of your favorite interactions?

#growth #ai

А какой у вас MRR? 😈

🔗 https://www.linkedin.com/posts/kyle-poyar_ai-saas-finance-activity-7255220443127726080-bqMw

AI could kill ARR as we know it. And, in doing so, it could kill SaaS metrics as we know them ⤵

Annual recurring revenue (ARR) is the building block of SaaS metrics. And it's the basis of SaaS valuation multiples.

It's (usually) high margin, predictable and growing. Which means SaaS companies are (usually) on track to become highly profitable at scale.

AI throws this off.

What's not classic ARR:
- Charging per successful AI resolution (Intercom, Zendesk)
- Charging per credit used (Clay)
- Charging per task completed (11x)
- Charging per photo edited (Imagen)
- Charging per demand package generated by AI (EvenUp)
- Charging per conversation (Salesforce)

We're moving away from charging for *access* to software and to a model of charging for the *work delivered* by AI & software.

This might mean greater volatility. Variable margin profiles. Seasonal revenue. Project-based, non-recurring use cases 🤯

Welcome to the new "ARR": annual *re-occurring* revenue.

Some implications of this shift:

1️⃣ Spending much more time unpacking the components of revenue.

2️⃣ Moving away from ARR multiple valuations to looking at last 12 month revenue (or, even better, last 12 month margin $).

3️⃣ Looking much more closely at revenue concentration -- I suspect there will be a far wider variance between the smallest & largest accounts.

4️⃣ Measuring newer things like "time to ramp" and "share of wallet" as predictors of future success.