top of page
Search

How will our apps change in the future?


We used to have this saying that software eats the world.


And today, apps do run the world.


Hungry? There’s an app for food.


Need a ride? There’s an app for transport.


Bored? Come into a beautiful app that connects you with your friends.


Or the other beautiful app that allows you to chain smoke short videos.


Today’s tech ecosystem is built by and dominated by apps, but as AI agents come online, will we see more or less of them? Or will most things run in the background?



With recent announcements, it seems that as AI eats away at the app layer, their first stop is Ecommerce.


We’re going to see an interesting shift in the ecommerce space in the next few years. Amazon used to be the big cheese. That might be changing.


A few days ago, Visa publicly announced a product called Visa Intelligence Commerce, in which the Visa protocol will run within LLM’s. Yes, it means having something like a ChatGPT also become your personal shopper. Next time you need new supplements, car parts, diapers? Just asked ChatGPT. Next time you’re hungry and want lunch? Ask it again. And if physical goods is the first thing that LLM’s want to eat away, what is next?


The other use cases in their demo video is ticketing, getting Ubers, hotels. That means a possible disrupting to all sorts of major websites we have: everything from Uber, Expedia, all types of airlines and the list goes on.

 

Website Type

Examples

Impact Description

Ride-Hailing & Transportation

Uber, Lyft, local taxi apps

Users book rides directly via AI, reducing app visits and simplifying comparisons

Travel Booking

Expedia, Kayak, Airlines

AI handles flight, hotel, rental car bookings, bundling services seamlessly

Food Delivery & Reservations

DoorDash, Uber Eats, OpenTable, Yelp

Ordering food or booking tables via AI, personalized recommendations, less app switching

Event Ticketing

Ticketmaster, Eventbrite, Meetup

Users book event tickets and manage attendance with AI’s instant search and purchase

Healthcare & Pharmacy

CVS, Walgreens, Telehealth platforms

Scheduling appointments, ordering meds, and telehealth visits done via conversational AI

 

So will we even have all these apps downloaded into our phone in the future?

My guess is as these companies start to broker similar relationships as Shopify has done with OpenAI and these LLM companies, these front end apps will eventually become back end databases. TicketMaster will still run a web app and mobile app for more traditional customers but its app will likely become less and less updated. It makes me think back to when I worked at Tmall and used to make sure that certain campaigns would appear properly on the web app and mobile app side. During moments of major time crunch and little resources, we would forgo the web app, because there weren’t that many users visiting us there anyway and focus most of our resources on the mobile side (ecommerce in China is 99% mobile). In the age where LLMs take on transactions, the same thing might happen, where increasing the findability and attractiveness of a SKU is much more important.


LLM’s are going to increasingly take on the role of marketplaces


Marketplace type of applications, whether for healthcare, food delivery or physical goods solved a findability problem. As AI becomes more mature and capable of scraping all of the internet, we have less use for dedicated marketplaces, because the entirety of an all-knowing app like ChatGPT is a marketplace itself.


I think companies will respond in one of two ways:


  1. Double down in investments of the experience side for customers so that they don’t leave altogether

  2. Investment in becoming more findable by LLM’s


#2 is usually part of most keynotes and workshops that I do where I teach brands and website owners how to show up in LLM’s responses.


A few words of advice for companies if they want to go to route #2:

  1. Focus on your core product. Bring more value to your customer.

  2. Make sure you are well reviewed and regarded. The results that show up are put together by the LLM scraping hundreds of data points including reviews, articles, and opinion from trusted sources.

  3. If you’re selling a physical product, beef up your product detail page. Think about long tail keywords. Your users will be querying with a specific situation or context, make sure you are capturing that.


As we get more and more comfortable asking our agents to help us do things on the internet, visit stores for us, order groceries and book tickets, we will become more comfortable with handing over our credit card information.


The only thing is current surveys say something opposite:

  • Only about one-third of U.S. shoppers are currently comfortable handing full purchase control—and therefore card details—to an LLM-powered assistant.

  • Trust gap remains the main brake: eight in ten people list stolen-card risk as their top online fear; data-misuse worries sit above 50% in most polls.

  • Net-net: interest is real but cautious. Expect uptake to hover in the 25-40% range until providers pair tokenization and visible fraud-controls with rock-solid opt-outs that reassure the other 60-75 % of shoppers.


Shortly after the Visa announcement, Perplexity followed with their announcement of choosing Venmo and Paypal to be their trusted payment partners. By the summer, we’ll be able to also use Perplexity as a personal shopper and use Venmo or Paypal to pay for the item.


Change is happening, just not at the speed we think.

 

 

 
 
 

Comentários

Avaliado com 0 de 5 estrelas.
Ainda sem avaliações

Adicione uma avaliação
bottom of page