Are we entering a post-app world? “The app is dead!”, “We are approaching the post-app era”, “It’s the end of the app boom, what now?” I look at the headlines and I see a lot of confrontation. Some say we are entering a new age in software and the way it is built and delivered, but it is still unclear where we are exactly going. In this post, written in 2016, I reflect on conversational interfaces like chatbots and voice assistants, the end of apps, and the future of technology.
Chatbots are the new cool kids on the block. Some say the conversational approach is revolutionary and will change interfaces as we know them (look at all the noise around Alexa), some say it’s meant to fail and that apps are still the way to go (ask the VCs).
The App is Dead
A study by Gartner found that a large number of users from the US, UK, and China used just six to ten apps in a given month in 2016. Coming from 26 apps per month in 2013 the experts are calling this a crisis for the app world. That’s what Gartner says. My take:
The app is not REALLY dead: The pieces of software we download from a store and assign a place in our smartphones’ home screens ARE ANOTHER STORY
The app stores are dominated by a handful of powerful companies and a few categories get it all when it comes to users’ time: Social Networking, Music and Games represent nearly half of the time spent on mobile devices. We are beasts of habits: We spend half of our mobile time on our favorite app, and almost 75% their top three apps. It’s hard for the rest of us to show up on the consumers’ screens.
We don’t download as many apps anymore and we tend to put apps in folders, where they die in silence because we are less likely to allow push notifications
And because we have so many apps we are engaging less with them: Less than 25 percent of people return to an app the day after they install it. App retention is at around 11 percent within a week of install. After 45 days, that number is less than 5 percent, before hitting 4.1 percent after 90 days. Looking at these numbers and the review figures one wonders where all the engagement really comes from. Has the world of reviews been botified before apps?
Experts say so: The process can’t be stopped — users are getting used to the quick interactions provided by the web, and look to reduce the friction of going from one app to the other, opening and closing icons once and again.
After the app, what’s next? Messenger apps’ ubiquity, their low use of data and their little need for local storage make chatbots an interesting platform to explore as the first step in this change of paradigma.
WhatsApp is used for everything in Brazil, the number of messenger app developers is counted at the tens of thousands and a lot of companies from huge players like Facebook to startups like my team at Reply.ai are building tools for chatbot developers. Speaking about WhatsApp; this will all just grow bigger as soon as they open their long-awaited API for bot developers.
Long Live The Bot!
Or not. Let’s not be killed by the hype: bots are not the cure for everything and they have a couple of downfalls worth considering before diving in:
- Not all apps are botifiable. Actually most of the apps that generate the most revenue are not. Think of games, totaling 75% of the app store revenue, or productivity apps like word-processors, social networking apps, photo editing apps… With messenger chatbots, the look in which some apps are presented to users will change, but a lot of the processes will still rely on an app that was once downloaded from an app store.
- There’s a belief that bots are convenient, cheap and easy to build, which is partly true at least when talking about prototypes. But don’t be naive: chatbots are just software and the fact that they don’t have buttons only means you don’t need a button designer anymore. Software costs are about a lot more than that an anyone that promises you a chatbot for 500 dollars is pulling your leg.
- Bots are still dumb: The technologies that are necessary for really smart bots are still on the works: NLP, artificial intelligence, cognitive services… The expectations are high thanks to years of science fiction, PR and Mark Zuckerberg’s personal butler, but don’t get overexcited (yet).
- In addition to the technologic moment, society is not yet ready for this change of paradigm. New Yorkers may be buying into the craze and ordering their lattes on messenger, but most of your regular Joes anywhere else only know about stickers for messenger apps. Ontop of general awareness, dynamics and social uses need to be created around customer service bots (is it illegal for someone to sexually harass a customer service bot?) and voice interfaces like Alexa (is it polite to talk to someone else’s voice assistant?). Interesting times ahead.
Live together, die alone
As we noted above a lot of the processes one may want to have as a chatbot would still require some kind of application, previously downloaded. A chatbot is an interface but the processes still need to happen somewhere.
A good conversational interface is not only connected to a number of services but also will typically be linked to a) an external server where data is kept and processes are run or b) an app in the device that communicates with an external server.
If you think of stuff like calendar bots, bots for yoga routines, or banking bots this becomes quite clear. You may be talking to a bot, but the events, the asana videos and the movements need to come from somewhere.
Even if chatbots end up eating apps for breakfast, why would we get rid of a perfectly working calendar app where I can see how busy I am at a glance, in favor of a bot? If you have tried scheduling assistants like Amy or Meekan you know what I mean.
Apps and Bots can be great companions. Look at one of the most popular chatbots out there, Swelly. They’re famous for a chatbot where you can ask an audience to A/B like something for you (when you want to know which pair of shoes to buy or what holiday to book). HOWEVER they have native mobile apps for both iOS and Android (with 10k-50k downloads), AND a web app for businesses. There, there.
This is not Beta vs VHS. Apps, messenger bots, voice bots… This is no one-course meal. It’s a buffet. And you can have it all.
I encourage you to give chatbots a go and see how they can integrate with your existing technologies. Go to bot building informed – to keep your expectations under control – and with clear, measurable goals that are specific to your conversational strategy: a chatbot is not an app, so don’t measure its success against that of one.
You do not need to kill your apps, instead think of what features could be enhanced by a conversation: maybe it’s the boring notifications and emails you send your users every once in a while or your expensive and annoying e-commerce site. Look at your value chain and think of what steps don’t belong on a form anymore, but keep in mind that any actions that require a certain complexity or have different data levels or layers (a map, a chart…) still belongs in your apps.
It’s Apps & Bots. For now.
My company, The Neon Project, can help your organization get started with a conversational strategy and develop and release your first chatbot or voice app. Through the years we have developed a methodology for conversational building and a series of tools that can help us speed up the development and reduce project costs. With proven benefits in different parts of the value chain (including a conversion rate up to four times higher than traditional contact forms), chatbots are a great way to establish relationships with your customers.
Let’s work together! Get in touch through The Neon Project website or using the email below.
Leave a Reply