AI Startup Innovation Process

A quick mention to Wispr Flow, Mac keyboard app. I’ve only made time to make this post because I can (mostly) dictate it.

How the heck to come up with an idea!?

Read this first:

https://www.nfx.com/post/10-places-to-find-product-market-fit

Chad GPT-based idea generators

Here are some pre-trained LLMs below, but Claude, Perplexity, ChatGPT, etc. also work if you tell the LLM that it’s a business expert or a coach for founders, and then serve your prompt. If you have a certain type of business that you want to explore like those with network effects or community platforms, include that in the prompt. Spend some time iterating and save your best ideas.

https://maze.do/saas-ideas

https://www.hyperwriteai.com/aitools/startup-idea-generator

https://chatgpt.com/g/g-kzS669NF0-venture-drinks

Customer Discovery on Reddit

https://gummysearch.com/

Greg Isenberg has an extensive video on how to do this, so check that out:

Search volume-based market research

I would also highly recommend: https://explodingtopics.com

If you don't already have SEMRush or another search volume analysis tool, this one is a great free Google Chrome plugin: https://keywordseverywhere.com/ And there are lots of other options.

Look for growing search volume trends ideally with high volume and low cost per click (competition)

Now things are getting fun! We have a basic idea, so let's visualize the business.

The "Business Model Generation" book from Strategyzer is a great primer for understanding value propositions, customer segments, revenue streams, and the rest of the business model canvas.

This is where we figure out pains and gains, jobs to be done, the avatar or ideal customer profile you serve, and to have a starting point for your MVP (minimum viable product).

Using the canvas above will help think through how the business will actually work, and Strategyzer offers an AI business prototyping tool to make the process faster: https://www.rapidvisual.ai/

Here are a few screenshots to see what you can build with it:

The fastest road to failure (yes please)

We have a business idea, we think we know what the market might look like, and what we need to do now is find the fastest road to failure - which means we want to disprove this idea as quickly as possible. Because if it's going to fail, we want to know so we can move on to the next idea before spending a lot of money and time on it.

Here are another couple of fantastic GPTs from Strategyzer.

Put your elevator pitch idea here: https://chatgpt.com/g/g-Ng8gDifwI-assumptionsgpt

Take that output and drop it into their Experiments GPT: https://chatgpt.com/g/g-ob1QQB5gD-experimentsgpt

Now that we have some known points of failure, we can go out and test our hypothesis. You can go stright to the market, or save potentially a ton of time and use synthetic data to start.

https://www.syntheticusers.com/

This entire process for one idea could take less than a day, but I also make space for new ideas that come up throughout the process. Keep after it until you find the fast moving water!