Despite the growing adoption of Android...
A brief introduction to AI chatbotsA modern chatbot (aka chatterbot, talkbot or simply bot) is a lightweight Artificial Intelligence app that runs inside a messenger like Facebook or WeChat and answers users’ questions asked in natural language the way a customer support person would. Its natural language processing (NLP) capabilities are driven by machine learning algorithms, so bots normally get smarter over time. First introduced by Alan Turing, the chatbot concept was further developed by Joseph Weizenbaum who created ELIZA – a semi-smart computer program that processed user questions and matched patterns. Also, there was A.L.I.C.E., a minimalistic app aimed at promoting the Artificial Intelligence Markup Language. The growing adoption of smartphones – as well as multiple technology advancements in smartphone technology – gave a boost to mobile app development. Creative vendors were trying to fit website functionality on to a smaller screen and designed mobile software for every service possible. “There’s an app for that” – that’s what we ended up with. And things were going well until the 2015 study by Comcore revealed that 78% of all smartphone holders were using free apps only. Although users now spend 90% of mobile time in apps, only 3 to 5 applications (that is, social media and occasional gaming titles like Angry Birds and Clash of Clans) see heavy use. Unless you come up with a unique engagement strategy and provide real value to users, there’s no way they’ll download your brand application.
Bots are the new appsAmid the chaos WeChat, a popular Chinese messenger with over 800 million active users, launched the first chatbot development platform, thus enabling companies to reach out to their target audience through a channel they do use and love.
Here are some astonishing facts on WeChat bots:
- If you want to start business in China, you’d better create a WeChat bot before you build a website. Some local VC-funded companies that generate millions of dollars in revenue don’t have websites at all;
- The messenger supports two types of chatbots. These are subscription accounts (which allow content providers to send new articles to subscribers) and service accounts (assist users to book a hotel room or complete an online transaction);
- The majority of Chinese bots are triggered by “keywords” and do not engage in long conversations. When facing a question they do not understand, bots will connect you to a human assistant;
- Microsoft’s Xiaoice, an experimental entertainment bot that mines the Chinese Internet picking up conversation patterns, is used by 40 million smartphone owners. 25% of those have told Xiaoice they loved it;
- WeChat bots have access to users’ personal data (including WeChat Wallet number);
- There are 10 million official chatbots on WeChat.
Top 3 chatbot advantages
- Wider reach. Although the total number of app downloads will surpass 268 billion this year (up 17% from 2016), the growth is now primarily driven by emerging countries. The US mobile app market, on the contrary, saw a 20% decline last year. It is estimated that social media and messaging apps account for 60% of all mobile app downloads. Can your company compete with Facebook, WhatsApp and YouTube? Probably not. What’s more, Business Insider claims messaging apps will soon have more active users than social networks. A database of 2 billion customers – isn’t it something a sales manager would kill for?
- Increased user engagement. Judging by Kleiner, Perkins, Caufield and Byers 2016 Internet Trends study, users spend 80% of their mobile app time in just 3 applications. Over 65% of smartphone owners do not download new apps for months on end. Meanwhile, conversational interfaces have become one of the biggest UX trends. 63.9% of consumers believe brands should be available through messaging apps. As a result, bot-delivered messages have a 85% open rate (compared to 20% generated through emails). Also, chatbots do not consume smartphone storage, load faster than mobile apps and websites and help you optimize marketing efforts;
- Lower dev costs. One of the main reasons why bot vs mobile app question has come into prominence is the growing mobile app development costs. How much does it cost to build a mobile app? The answer depends on several factors including the type of an application, its feature set, the target platform (now it’s just iOS, Android or both), the size of a mobile app dev company and its country of residence. iOS app development costs, for example, range from $ 150 (US) to $ 35 (Eastern Europe) per man-hour. According to Pavel Shylenok, CTO at R-Style Lab, a simple mobile app will cost your company at least $ 10 thousand (for each platform). Building a high-quality Facebook bot will cost you around $ 5-10 thousand, too, but the program will run on any smartphone regardless of its OS.
Why do brands still invest in mobile apps?
- Bots are immature. Companies that do bot commerce on WeChat quickly realized they needed a human assistant somewhere in the background in case something goes wrong. Microsoft can tell you firsthand what “going wrong” means. Its Tay (which ironically stands for Thinking About You) went from a human-loving bot to a racist freak in less than 24 hours. Also, bot conversational abilities are most often limited to one topic;
- Bots can replace promotional, retail and content-focused apps only. Although there are 2.4 million apps on the App Store, 75% of its revenue is driven by games. Some bots do rely on gamification to engage younger audience (Miss Piggy, Unilever’s Signal Pepsodent, Oyoty); however, they are not endless runners and arcade games users go crazy about. The success of Niantic’s Pokémon Go (7.2 million downloads during its first week), Prisma (the most downloaded app in 10 countries) and MSQRD (which made it to Apple’s top 5 iOS apps of 2016) proves mobile apps is still a business of its own – and that’s where chatbots fall short.
Mobile app success stories from last year uncovered a rapid change in users’ expectations. If you want to dominate the App Store or Google Play through 2017 and beyond, your application should either provide real value to end users (like niche business and productivity software) or use cutting-edge technologies (neural networks, sensor data or natural language processing). It’s the same with chatbots. If it takes customers 80 messages to order pizza through an AI assistant, you’d better hire a human one.There’s always room for improvement, right? And that’s why you should invest in market research, monitor mobile app development trends and partner with a reliable vendor.