“Bots are the new apps” according to Microsoft CEO Satya Nadella in a keynote given at a Microsoft developer conference in March 2016.
On the basis of the amount of start-ups working on bot related activities at Web Summit 2016, it would seem as if bots may be about to become mainstream after all.
I counted 23 bot related companies exhibiting at the early stage (Alpha) area at Web Summit last week.
These included bots for different kinds of services as well as bot building platforms. A year ago, this area might have been taken up by start-ups working on mobile apps.
Is this a clear sign that bots are about to move beyond the nascent stage? Perhaps.
What is a bot?
We’re actually a lot more familiar with bots than we might realise. Think of Apple’s Siri or Microsoft Cortana. They’ve been around for a while but until recently, haven’t really gained traction for various reasons.
Search engines may be considered as a type of bot. A user types in a command or request in the form of a search query and the search engine returns a number of results based on that query.
Or let’s go even further, remember Microsoft’s paper clip virtual assistant? That was discontinued years ago but bots have been taking off again recently as advances in machine learning and artificial intelligence make them more accessible and versatile than before.
In terms of the more recent bot experiences, brands are starting to use them for more personal, proactive and streamlined interactions with people. In this sense, a bot is just a new type of user interface.
According to Ted Livingston, CEO and founder of messaging app Kik, speaking on the digital marketing stage at Web Summit, “people think bots are about chatting. They’re just a better way to deliver software. It’s just a user interface.”
Chatbots are just automated computer programs that can simulate conversation with people to perform tasks or answer questions.
How sophisticated are bots?
There is a scale of complexity when it comes to bots. At the most sophisticated level, a bot is an artificially intelligent creation capable of understanding complex interactions.
At the lower end of the spectrum, a bot is just a simple interface that can respond to a limited number of pre-programmed commands.
For an idea of how basic bots can be, there is a plethora of basic bot-building platforms online. I created Seanbot at robots.me. It’s not going to do my work for me anytime soon.
As basic as some bots may seem, Kik’s Livingston says “calling bots basic today is a bit like calling websites basic 20 years ago”. We’re going to see bot sophistication increase far more quickly than we did for website functionality.
Why are bots getting popular so quickly?
According to Ted Livingston, one answer might be the growth and use of messaging platforms which is providing some of the infrastructure for delivering bot interfaces.
Back in April 2016, Facebook opened up its Messenger platform to developers at its F8 developer conference. That means that brands that want to reach people on mobile can build bots to share weather updates, order pizza, confirm flight reservations or send receipts after a purchase.
For example, Mark Zuckerberg presented the use case of ordering flowers by chatting with the 1-800-Flowers.com bot, ironically bypassing dialling the telephone. As of July 2016, there were more than 11,000 bots on the Facebook Messenger platform.
Also in April, Kik launched its own bot store, which according to Livingston, has already attracted more than 20,000 bots.
He told users at Web Summit that in China there are more bots launched on TenCent’s WeChat every day than websites added to the Internet. In other words, according to Livingston, “WeChat is the Internet” in China.
Kik bot store
Google revealed its chatbot strategy in May. Unlike M, the virtual assistant in Facebook’s Messenger, the Google Assistant will respond to voice queries and not just text input.
Microsoft also has its own open source bot builder. Unlike Messenger and Kik bots, these bots can be deployed on other platforms.
Amazon has also opened up its bot building platform, Echo, to developers since 2015.
What do bots mean for the future of apps?
Chatbots can be delivered via website interfaces for managing basic customer service queries.
There could be a time though when instead of visiting an ecommerce site, we simply message the relevant stores bot on Facebook Messenger which would ask us what we are looking for and we simply tell it.
If we think about mobile apps, there are a number of reasons why bots may in fact be the new app.
Why might bots succeed apps?
1. Difficulty getting cut through on mobile app stores
If we examine mobile, according to Livingston, three quarters of American smartphone users download zero apps per month. Also, research suggests that users only use 3–4 apps on a regular basis.
Getting cut through in the app store and then retaining users is also incredibly difficult. Bots on the other hand are available via some of the most popular messaging apps and so can provide a new way to manage frictionless interactions with consumers.
Livingston says that if you can create a great bot experience, that experience can spread virally.
“The problem with the app store is that they (apps) can’t go viral. The top 50 apps take up the majority of downloads. Bots make it easy for experiences to go viral via mentions which allow bots to be put into conversations”.
He gave the example of a fun chatbot Kik built called Roll that went from 0 to half a million users in 30 days.
95% of the user base came by being shared peer to peer. It now has 1 million users.
2. Ubiquity of messaging apps
Considering the incredible growth in the number of bots available via Messenger, Kik, WeChat and other platforms, then according to Livingston “if you are a developer at this point and you are still building an app, you are crazy”.
- WhatsApp, owned by Facebook, is the most popular messenger app in the world with 1bn monthly active users (Source: Statista).
- Facebook Messenger takes second place with over 900m monthly active users (MAUs). This number has been consistently growing since 2014 when it had 200m MAUs.
- Chinese company Tencent’s instant messenger QQ isn’t far behind with about 880 MAUs (ChinaInternetWatch).
- Tencent’s WeChat is also on the list with 806 million monthly active users as of October 2016 (Statista).
Further, the number of mobile messaging app users is forecast to nearly double between 2014 and 2019 (Statista). That means a projected total of 2.19bn people using mobile messaging apps by 2019.
3. There is less friction using a bot than installing an app
To make user of a bot, a user just needs to search for the bot within their preferred messaging app and start “chatting”. Because you are interacting via your installed messaging app, the bot will have access to your identity.
This contrasts with searching for an app, installing and creating an account. This may be problematic if you are outside a wifi zone and don’t want to use data to download the app.
I tried using 1800 Flowers this morning. While I didn’t actually order anything, connecting with the bot and starting the interaction was quite easy.
4. Bots may be better options for businesses that don’t have an immediate business case for an app
Users can add a bot to their contact list rather than downloading an app.
What this means is that small businesses or companies that don’t have a clear use case for downloading and keeping an app on your phone may benefit from a bot instead.
Think of hotels or your local hairdresser / barber. To spend the time, effort and money to build an app for these services wouldn’t be worth it but there may be a long tail in having them in your contact list to make bookings etc.
5. Bots don’t use up valuable memory on users smartphones.
Use case – how KLM is starting to use bots for customer service
At Econsultancy’s Festival of Marketing 2016, KLM’s Social Media Manager Karlijn Vogel-Meijer discussed the airline’s approach to customer service via social media.
KLM has been a poster child for using social media as a customer service tool. Karlijn told FoM attendees that KLM has 235 agents dealing with social mentions around the world, 24/7.
To put it into context, the KLM team responds to 15,000 social mentions per week in 12 different languages.
They used to have a 60 minute promised response time but in reality, most customer don’t want to wait that long. These interactions are managed manually but the volume of interactions is increasing year on year.
For this reason, KLM is exploring AI and bots to help reduce the strain, whilst working to maintain a human feel.
KLM added in a Facebook Messenger chatbot in March. Since its launch, it is receiving 5 questions per hour via Messenger. This number can increase to 13 in a peak hour.
The brand is using Messenger for automated updates around checking in, potential delays and sending boarding passes. However Karlijn was clear that if a customer has a more complex question, a KLM agent will still get involved.
Bots right now may not be in a position to interact in a personal way she says. In that sense, KLM’s social customer service strategy hasn’t changed. They’ve just added a layer of technology to support common requests.
KLM is exploring adding more functions to Facebook Messenger and expanding its chatbot to other platforms like WhatsApp and WeChat. These are expected to roll out within the next year.
Learn more about bots
Econsultancy has published a number of posts and reports about bots and artificial intelligence in the last year. In particular, readers may find the following helpful: