What I'm Learning Building an AI Company for Restaurants

Managing your brand and customers online is increasingly important and complex. It requires skills and time that many small business owners struggle with. I had a front-row view of this struggle as I created a movement to mobilize community members on social media to support local restaurants through the pandemic. This became Takeout 25 and was eventually established as a not-for-profit to pursue that mission beyond the pandemic.

Takeout 25 brought community and businesses together to support and strengthen the local economy. It attracted new business to the area, brought more customers to existing businesses, and created opportunities for the community to support local causes and nonprofits through its events and initiatives. Much of this impact was made possible through social media and other online channels. It gave us a firsthand appreciation for both their immense potential and the significant challenges they pose for small businesses.

So, as Takeout 25 turned five, we decided to invest in finding solutions to that problem. It was important that whatever we created be easy to use and affordable for an independent business. That set us on the path to creating 20fy.ai.

That path had many unknowns. We had to conceive and design the actual solution, select technologies and technology partners, and build and test the product with simplicity and affordability as key design parameters. Most importantly, we had to put it in the hands of actual businesses to determine whether it solved a problem they cared enough about to use—and eventually pay to solve.

That last part turned out to be the most important.

What I have learned is that building the technology is only part of building a company. AI has dramatically lowered the barrier to creating sophisticated technology. Things that would have required large development teams and significant capital only a few years ago can now be built surprisingly quickly. But AI hasn't eliminated the difficult parts of building a business.

You still have to understand the customer, identify a problem important enough to solve, fit the solution into someone's already busy day, earn their trust, and create enough value that they are willing to pay for it. Those questions cannot be answered by AI. They are answered by customers.

Our experience with restaurants reinforced this. A restaurant receiving a large volume of reviews has an obvious review management challenge. Another receiving relatively few may reasonably ask why it needs a tool to manage them at all. The technology can work perfectly in both situations and still create very different levels of value. That distinction between what technology can do and what a customer actually needs has been one of my biggest lessons.

It also taught me to listen carefully when customers say no. When you have invested time, money, and emotion in building something, the instinct is to explain why someone should want it. I am learning instead to ask why they don't. Is the problem too small? Are we solving the wrong part of it? Is the real value somewhere else?

Those questions are changing how I think about 20fy.ai. Perhaps the opportunity isn't simply helping restaurants manage reviews, but helping them extract more value from whatever customer feedback they receive: What do customers love? What should the restaurant fix? What should it promote? What can it learn that it didn't know before?

I've also learned that small businesses generally don't need more technology. They need less work. The best AI may be technology that largely disappears behind useful answers and actions. I have found simplicity to be surprisingly hard. Adding features is easy. Removing complexity is much harder.

There is a broader lesson too. I am not a traditional technology founder, yet AI has allowed me to participate deeply in conceiving, designing, testing, marketing, and continuously improving a software product. It hasn't eliminated the need for experts, but it has dramatically expanded what small teams and individual entrepreneurs can attempt.

Perhaps that is the biggest lesson from this journey so far. AI makes it increasingly easy to generate answers and build solutions. Building a business still requires us to ask the right questions, listen carefully to the answers, and be willing to change.

And the learning continues…