Thinking About AI and Your Product

Image by Andrew Kliatskyi on Unsplash

At WLCM, we’re full steam ahead on building AI apps for the OpenAI marketplace. This piece is Part 5 in a series helping entrepreneurs understand this new frontier. 

• Part 1 - Why we’re all-in on AI

• Part 2 - The major components of AI

• Part 3 - What you can build with AI right now

• Part 4 - How AI learns

Many founders and business leaders out there feel deeply that they should be using AI in their products and workflows… but they don’t know how or where. I suspect many of them are overcome by the pressures of the AI hype cycle, which spins the narrative that if you’re not using AI by now, you’ve fallen behind. 

As a result, they merely pepper AI into their decks and talking points and pray it yields something like success.

Implementing AI into your app or product or workflow is no different from any other technology. You have to understand what the technology can and can’t do, and you have to understand precisely where to put it where it can do the most good. Otherwise, it’ll just get in your users’ way. 

It takes time and due diligence to find the sweet spots. That said, I believe AI can make almost any software better.

Your product, your user base, and the product you solve for them are all unique. Rather than being prescriptive about what you should do with AI, I want to point you to a few places you may probe for opportunities. 

Re-imagine old-world functions

Many conventions of your average app – like forms, paywalls, etc. – have become obsolete. 

A friend of mine had a great example. Imagine you’re at the grocery store and you want to sample a new type of cheese from the deli. Instead of handing you a piece of cheese on a toothpick, they hand you a form to fill out. Then you give it back and the store mails a confirmation of your form fill to your house. Once you open that, then you can go back to the store and get your sample of the cheese.

All this before you even know whether you like the cheese or not. 

Many apps work exactly like this. Users must jump through hoops and fork over some of their data before they can tell whether they even like the app. 

Tech shouldn’t work this way anymore. It doesn’t have to, and AI plays a role in that.

What parts of your app feel like red tape? What parts of your app do users click/scroll through as fast as possible? 

AI can make these admin-type bits more of an experience. Typeform’s Formless AI product is a great example of how something as simple and dry as a form could be more engaging to a user.

Improve the user experience

The computers of the past have been more like calculators. You tell them explicitly what you want, and they do exactly as you say. If you make an error, the computer gets confused, because it can’t know you.

Google’s search engine was one of the first widespread examples of a program understanding us. By that I mean, you could pose questions imperfectly, informally, incompletely and it could surmise your intent. 

With AI, apps can think, rather than just calculate. Can you do some thinking or calculating on behalf of the user that gives them a better experience?

Apps that track something – calorie intake, moods, exercise, etc. - often require users to meticulously enter data manually. Generative AI can see, understand, and process imagery, sound, and text. Think of the way your phone can automatically find and group photos with a certain person’s face. 

AI could automate manual entry to a large extent. Parts that can’t be automated can be reworked into a more conversational, engaging experience that feels less like a chore. 

AI’s ability to match-make has big potential, too. Think how easy and effective Pinterest’s recommendation algorithm is. If your product requires match-making people, products, content, whatever, AI can perform that matchmaking much easier. 

Generative AI allows you to guide its decision-making through knowledge graphs built to serve your particular use case. Retrieval-augmented generation (RAG) lets you direct AI to use authoritative, current external sources to improve responses and stay current with its information. 

Improve backend decision-making

For products with some sort of appointment, reservation, or scheduling function, AI can help optimize availability and pricing to help drive revenue. 

Imagine you’re a restaurant, a therapist, a daycare, even a beautician. You make more money when your whole day is booked. If someone cancels at the last minute, it can be a pain to fill that slot. 

AI can observe traffic and learn when you will see the greatest demand, who and which appointments are most likely to cancel and when, and more. From that, it can generate dynamic availability and pricing and also project future demand.

What events cost you money that are somewhat unavoidable? 

AI can arbitrate the curve balls that come with scheduling in real time with minimal human interaction. This could work with scheduling employees too, in similar ways.

Perform data analytics

When it comes to data, our reach has exceeded our grasp. We can capture untold mountains of data, yet we have a hard time learning anything of substance from it. AI makes that possible. 

In fact, AI makes it possible to have a conversation with your data. Forget orchestrating intricate dashboards, graphs, formulas, and so forth. With AI, you can simply ask for what you want. 

What do you wish you could glean from your data that’s currently out of reach?

• Want to know what your most valuable customers have in common? Just ask.

• Want to know which behaviors indicate a customer is about to quit? Just ask.

• Want to know which types of jobs have the best margins? Just ask. 

You can certainly build and train an AI to deliver these insights and plenty of others that can massively inform the next step in your business. 

Exceed expectations of value

I think of AI as a means to close the delta between the value you purport to provide and the actual providing of that value.

In other words, it can deliver your value proposition better, more fully, more smoothly and help your app live up to its full potential. 

Remember, tech is a means to an end. It’s easy to get caught up in the “gee whiz” of it all and forget to ask if new features – be it AI or otherwise – are truly delivering an impact to your users. When it comes to UX, there are no neutral contours to the experience. Everything about a product either helps or hurts its usability and effectiveness.

So implement wisely and intentionally, my friends.