Like many emerging technologies, AI is being thrown against several walls to see if it sticks. The result is some fitting applications and some that are ‘solutions in search of problems.’ This exploratory process historically results in serendipitous – and sometimes accidental – discoveries. Killer apps take time.
There’s no real way to predict which applications will fly and which ones will die… just as no one envisioned Uber and Spotify when the iPhone first launched in 2027. But often, there is at least a set of principles that drive and define killer apps, such as a common trait of solving tangible user pain points.
Applying that basic framework to AI, we’ve seen a few integrations that make sense, such as Web.Com’s ChatGPT-based writing assistant and domain generator. These address real SMB pain points. Similarly, NextDoor’s AI writing assistant offers automated help for individuals who aren’t natural writers.
Where else are we seeing these principles play out, particularly within the Localogyverse? To further demonstrate, here are a few examples from the past week’s news cycle. – including one fitting AI integration and one that may see challenges. This will continue to be a moving target.
Brain Transplant
Starting with a positive example, Intuit has integrated AI across its suite of financial software, such as Quickbooks. The thought is that these apps are filled with rote tasks that people don’t particularly enjoy – the most painful example being tax preparation. The goal is to automate it as much as possible.
This has always been the case with tools like QuickBooks and Turbotax, including features to automatically categorize expenses in one’s bookkeeping. But like many recent AI integrations across the tech sector, automation and intelligence functions are now getting a GPT-based brain transplant.
The result is Intuit Assist. As we examined last week, it can generate personalized recommendations from your tax profile (e.g., “consider filing as an S-Corp”). It also serves as a reactive chatbot that fields questions with greater reliability than Intuit’s rarely-helpful FAQs and help-center message boards.
Intuit is also applying an intelligence layer across its suite. That opens the door for crossover integrations like generating email marketing content in MailChimp based on product inventory in Quickbooks. Altogether, it’s a logical AI integration that alleviates real pain points – again, the name of the game.
Fries with That?
As for an AI integration that may be more challenged, Doordash has launched a voice-based AI agent to field incoming phone calls to restaurants. As we wrote last week, this was driven by the fact that most customers prefer to call restaurants, but most of those calls go unanswered – a legitimate concern.
The result is a chatbot that will answer calls, handle orders, and even suggest complementary items (read: upsells) that traditionally go with the items ordered. Those recommendations are based on AI training sets from past Doordash orders, indicating what foot items likely complement each other.
So what’s the challenge? This pushes against an embedded cultural sentiment: people hate automated voice response (AVR) systems. Though the performance may improve with new AI infusions, the connotation still exists. It will take a while for these deep-rooted and widespread attitudes to change.
Another challenge is that this AI integration is expressly built to serve one particular constituent: restaurants. Its primary goal is to boost order sizes through upsells. There’s nothing wrong with that, but it should balance customer needs and UX, which goes back to the previous paragraph.
Market Reaction
So there you have it… just a few examples from the past week. We’ll see many more examples on both sides of the equation. And the AI playbook will evolve beyond the success factors we’ve raised above. We’ll also see how the market reacts to the above moves, which is more telling than our speculative take.
Header image credit: Iker Urteaga on Unsplash
