You can’t move far in the tech ecosystem without the subject of conversational interfaces coming up. In a March 2016 keynote presentation, Microsoft CEO Satya Nadella went so far as to say “Bots are the new apps”. So are they living up to the hype? Let’s zoom into what’s happening in the enterprise and workplace, what’s stopping us, and what’s next.
Pieces of the puzzle
Let’s break this down to basics. When we talk about “bots” there are a few inter-dependent elements:
- Bots Little bits of software, often interacting with other software via APIs that contain some logic, often dedicated to a specific task.
- Natural Language Interfaces The ability to interact with a system using natural languages. This requires things like Natural Language Understanding (Google Natural Language API), and Natural Language Classifiers (IBM Watson’s API). It could be via voice (Bing Speech API) or via typed words.
- Chatbots The combination of bots and natural language interfaces to solve a problem, usually hosted on a messaging platform.
- Messaging Platforms & Apps Mobile & desktop apps providing discovery identity, APIs, contacts and the messaging interface.
- Voice Platforms & Devices Physical device, often an app and the platform, providing discovery (or not in some cases), identity, APIs.
Note: With voice, today the physical device is mostly ubiquitous with the platform — some providers will stick with this (like Apple), but others (like Amazon) are allowing third parties to include their voice platform in third-party hardware. History repeats.
Bots, bots and more bots
Single purpose bots
Most bots you’ll find today in workplace chat apps like in Slack or Microsoft Teams are notification bots. They hook up to your existing apps and tools, replace email notifications and often add a great deal more notification types. They inform you when something ‘interesting’ happens — say, when someone subscribes to your mailing list, a shared document changes or a lead is updated in Salesforce.
You can add these to group channels to put real-time information in the place where the conversation is happening. They are also super useful for creating personal news or activity feeds.
A significant number of bots — including most created by start-ups in this past two years — are helper bots. They are designed to help make you a better you, or your team a better team. They might also do things like help you close sales or ask you for a weekly report at 5pm on Friday to share with the team.
It won’t be long until these bots leap into being ‘intelligent bots’ but today most are simple rules-based workflows, and suffer from low engagement. For a few examples take a look at GrowthBot, Meekan and Troops.
Ah, finally we’re getting closer to our sci-fi future — our own virtual assistant that will one day (and not too far away) be augmented into our brains, helping us be our best selves.
Facebook M, Alexa, Siri, Cortana. These type of bots are still in their infancy and difficult to build, but the exponential curve of improvement tells us one thing: this is the future of human computer interfaces.
These bots will get to know our likes and dislikes, they’ll be able to understand our intent, summarise what’s important to us and take action on our behalf.
If this is the nirvana, why aren’t we there yet? Virtual assistants are fast entering our personal lives, but not so much at work. The primary reason for this is that access to data at work is much more difficult.
In our personal lives the data we need is available on the web or stored on our phones. At work this data is spread across multiple systems, with complex access mechanisms, and with varying security and compliance requirements.
The machine learning algorithms behind AI need both access to this hard-to-reach data, and access to ‘see’ our daily behaviours in order to function. Solving this will be the key to everyone having a personal assistant at work.
One of the most fruitful areas of bot development right now is where people are writing their own bots to solve real problems. These bots are often pretty simple, but also hugely effective. One of the first widely shared examples of this in Slack was at Al Jazeera with their Newsroom bot. An awesome example of bots solving real problems and making a team more efficient.
Bots replacing apps
So, will bots replace apps? Occasionally yes, but it’s probably the wrong question. We’re at the start of a shift that will see the interface separated from the applications and services we use. The interface we’ll use will depend on context and what’s most convenient; voice, chatbot, notification screen, widget/integration with another app. I’d predict 80 per cent of the most common interactions won’t involve opening the original service that provides the data.
The apps and services that win in this next phase will be those with the most flexible APIs (often with no interface at all), that allow virtual assistants and other interfaces to leverage their capabilities in the way that is most convenient for the user.
Bots will also enable new types of solutions to age-old problems. The first wave of SaaS companies (Salesforce, ServiceNow, Workday) largely moved forms (systems of record) into the cloud, but the user experience didn’t fundamentally change. Conversational interfaces will remove the form, and thus new companies and new winners will emerge.
For users, the shift to conversational interfaces will have 10x impact than the shift from on-premise/desktop apps to cloud/browser based software, which mainly benefitted the IT department.
What’s holding back progress?
At work, this shift has a key dependency: an enabling platform must be adopted enterprise-wide. A platform that delivers enterprise-grade security and compliance, with the team messaging and bots that enable all the new benefits for users.
These platforms will finally allow communication, real-time data and workflow to come together in the same place. For the first time, an end to context switching in order to get stuff done!
Q: So what’s stopping us all having a personal virtual assistant at work?
A: Access to data — or lack of it.
As mentioned above, one of the primary restrictions on virtual assistants at work is the ability to access data. It’s therefore no coincidence that messaging platforms like Microsoft Teams, Slack and ours, Blink, are starting out with the request for permission to connect to your workplace applications. In our personal lives our phones already contain all this data — but at work we’re missing a unifying layer and the plumbing behind it. Data is the oxygen virtual assistants need.
Messaging platforms — the new operating system
Our vision at Blink is to build a universal interface to work: a single intelligent app that lets you get stuff done, quickly and easily.
The new generation of messaging apps at work are coming along at a time when several trends are converging to create a wave of change that is now unstoppable:
Explosion in the number of apps — BYOS (Bring Your Own Software) is resulting in an explosion in the number of apps we all use everyday and the context switching this causes is killing productivity. Did you send me that file via Dropbox or Google Drive? Did you email me the link or WhatsApp it to me?
Frustration with traditional software Why am I so productive on my personal phone, but getting into the office is like stepping back into the 1990s? Our use of personal software like messaging apps and travel search engines is changing our perception of what good software can do. People are fed up of the old crappy software they’re forced to use at work.
Accelerated teamwork and innovation Traditional tools, like email, encourage hierarchical working and the creation of fiefdoms. Team messaging enables teams across the organisation to form quickly, deliver/solve a problem, and for everyone to then move on to the next thing.
Team Messaging allows large companies to make scale a competitive advantage again, enabling them to move at start-up speed.
Increasing need for compliance and security Often seen in conflict with speed and innovation, but with messaging apps acting as the cloud operating system, shining a light on shadow data, compliance and security actually get easier.
You can think of messaging platforms as the connectivity and productivity layer for all of your cloud and API-based applications, an operating system for a modern workplace that can predict what’s most important to you right now, and that can search everywhere from one place. Apps like Blink might look like ‘messaging’ apps today, but together with bots, virtual assistants and unified search they will change the face of enterprise software.
Blink makes possible a digital workplace that lets people work they way they want — with instant messaging, bots and team-working — while providing enterprise-level security and compliance. We’re a new type of enterprise software company and we’d love to talk to you.
If you're interested in improving your workplace digital experience, we'd love to talk. Drop us a line at firstname.lastname@example.org now.