From audio apps to food delivery, picking winners in era of change

Trend 3: Online food delivery will stay, but the survivors will become super apps  

When people say the COVID-19 crisis is continuing to devastate the economy, that doesn’t apply to all sectors, nor the startup scene. Like many shifts in consumer behaviors, demand for online food delivery is unlikely to drop in the post-pandemic era. But here’s the problem: except for consumers, it’s a sector in which no one wins.  

Delivery platforms such as Deliveroo or Uber Eats typically charge a 15% to 30% commission on each order, but restaurants are subsisting on razor-thin margins (often less than 5%). Take the example of two Californian pizzerias that had once generated profits between $50,000 and $100,000; they started to lose as much as $40,000 annually once their customers shifted to delivery, the New York Times reported. The owner of a Chicago pizzeria pocketed only $376.50 from Grubhub after more than $1,000 in delivery orders, according to the Los Angeles Times. 

There are many reasons for this. The total food order for each delivery are often cheaper. Most people don’t order fancy cocktails or champagne together with takeout; they buy booze at stores. Since the customers don’t see the servers, tips are less generous. Then there are the additional expenses for drivers and bikers.There’s no single cause for the bad economics. It’s death by a thousand cuts.  

Meanwhile, delivery platforms are seen by the public as interchangeable. You can’t tell the real difference when ordering from different platforms. The one differentiation is the number of restaurant listings, but then most restaurants would sign up for multiple platforms anyway.  

That leads to ruinous competition. There are too many players in a crowded space. Delivery platforms become big spenders on paid ads on Google and Facebook. They call it market investment, but they are merely buying temporary market shares.  

What we’re beginning to see is online food delivery companies becoming super apps. They branch out beyond the single function of bringing hot meals to houses. We have seen that with Meituan, operator of China’s largest food delivery and local services platform. Singapore’s Grab is offering everything from food and parcel delivery to hotel and airline bookings and access to financial and health services. This is also the logic of Uber. A ride-hailing service is easy to copy, but stitching Uber Eat and Uber Ride together can create higher customer stickiness. Becoming a super app is the new game in town.   

Trend 4: Audio apps will be big, but Clubhouse’s format is not yet ready 

Before Clubhouse, there were podcasts. Before podcasts, there were radio talk shows. When screens are proliferating and screen time is lengthening, voice-only media fits well with our ever-growing multi-tasking lifestyle.  

Like many of our readers, I have an iPad, an Apple Watch, a Kindle, a flat-screen TV, two laptops, and iPhone. That is, by any measure, an excessive number of screens for one person. And all of my devices and apps are screaming for my attention. They demand my total focus, even if that’s only for a few minutes or seconds: the instant messaging, the text-based social-media feeds, the video- and photo-heavy apps.  

Audio apps are distinct. They can simply run in the background. I can listen to an audio book while grocery shopping or a podcast on Spotify while cycling. Not only are these apps less intrusive, they actually make mundane tasks more enjoyable. Waiting for the next train can feel productive when I’m listening to Think Again, by Adam Grant.  

That’s the allure of Clubhouse. When every app these days seems to promise an “immersive experience”, the “drop-in audio” social network is clever enough to focus on voice-only chat rooms. Community guidelines prohibit recording and transcription. Conversations on Clubhouse are therefore only available in real time.  

Being ephemeral adds further appeal, especially at a time when user privacy and data protection are coming to the forefront. But if Clubhouse is to become another billion-dollar business, or for any voice-only app to do so, it must resolve one missing link: organizing information.  

Google’s search engine helps us locate the information we need. Amazon is powerful not because of its extensive selection, but because you can locate what you want quickly. Spotify is cool because it recommends new music that you may love. Netflix is (almost) addictive because it shows what you can’t stop watching. What we see here is the power of matchmaking when powered by artificial intelligence.  

What that means is for a voice-only app (and that includes Spotify’s podcasts), the biggest problem has not yet been solved. We need better AI that can digest human speech and then classify and score it autonomously. Whoever can go beyond granular recommendations based on genre, and offer them based on content and stylistic analysis, will becoming the Google of voice apps. Is the technology there yet? Judging by the current performance of Alexa, Google Assistant, and Siri, I think there may be still some way to go.  

Now what about Clubhouse? Can it last, with its ephemeral proposition? It can, in theory. But it will involve quite a bit more work. And I am not sure if Clubhouse has deep enough pockets to fund such fundamental research and development. Still, one can imagine an AI machine that understands the real-time discussion inside the chatrooms. It’ll then label a speaker’s performance across multiple dimensions. Those will include subtleties like regional accents, political leaning, sentiment, pitch and tone, in addition to the speaker’s content. The more dimensions it scores, the more precisely it will be able to match a speaker’s quality, style, and topics with the listeners’ preferences. Of course, Clubhouse will never be able to give perfect recommendations, since nothing is pre-recorded. But that surprise element, while tightly controlled, will be its strongest appeal. Humans like surprises, just not too much.