AcmeCorpBot: Hi John. Ready to receive yesterday’s store highlights?
John: Yes, let’s get started.
AcmeCorpBot: Visits to your ecommerce store were within normal variance most of the day but there was a spike after 6pm. Despite the increase in traffic conversion to sales dropped by 10 per cent. Do you want to know more about the spike?
AcmeCorpBot: It was due to a post by Facebook page FavouriteAcmeProducts, linking to our new AmazingWidget. Sentiment analysis on the post indicates that it was positive and enthusiastic. The traffic was mainly US-based.
John: @AcmeCorpBot - set a task for the ecommerce team “Please provide an analysis of why conversions were down despite increased traffic and positive a message to the digital team“. Attach today’s store highlights. Due as soon as possible.
AcmeCorpBot: The e-commerce team has been contacted with the task and Simon has committed to doing further analysis and presenting initial findings by 2pm.
We’ve all seen the films. From Minority Report (2002) to Her (2013) and, more recently, Ready Player One (2018), we are constantly dreaming up futures where we can fluidly converse with computer programs and access and manipulate data effortlessly. What is changing now is that some of these movie scenarios are about to go the way of the Star Trek tricorder. They are becoming reality and in a few years the movies will look quaint as real tech surprasses them in unexpected ways.
Case in point, the scenario laid out above is entirely feasible today. All the technological pieces are in place and the workplace provides the perfect environment for these technologies to flourish and enable a whole new way of working. Compared to consumer-facing chatbots and automation the workplaces offers a much more stable and limited domain meaning we can employ technology much more effectively. In order to make this scenario real an organisation needs to master three things. Data, automation and conversational interfaces.
Let’s look at the three aspects that come into play and how they can be employed to your advantage now.
A decade ago the biggest challenge with data was working out how to store it and access it reliably and economically. Since then classes of storage such as NAND memory have increased in capability faster than Moore’s law and the cost has dropped to such low levels it no longer features as a concern for most organisations.
At the same time our ability to process data with powerful algorithms through easy to use tools has exploded as well. Cloud computing and advanced software make it easy to crunch through huge amounts of data. Tools like Google Analytics with its almost ubiquitous usage actually act as a standardising force.
The end result is that an entirely new set of services is being built that takes advantage of standardised access to data (such as data within SQL or Google Analytics), cheap computing and storage costs, and advanced AI techniques such as Natural Language Generation. These services can provide easy to understand narratives about data in a way that is immediately actionable by organisations.
Access to more relevant information through data analysis is crucial but will remain inefficient if it is not accompanied with increased capability to act on it. This is where automation can provide support. There are two ways of thinking of automation.
On the one hand you can automate reaction to new information. For example, if through data analysis we are seeing an increased request for a specific item on our store we can automate the ordering process to have more units delivered to distribution centres. That is the kind of decision-making Amazon relies on to have the most efficient distribution network possible. However this requires a high degree of sophistication in terms of data analysis and automation.
On the other hand you can look to automate more mundane tasks, including data analysis tasks, so your team has more time to focus on the higher-level more complex and more high value decision-making. These are the low-hanging fruits that most organisations can grab today. Looks at all your processes and workflows and you are sure to see steps that can be better connected via programs that take advantage of API-level (application programming interface) access to the functionality. Connecting services via APIs creates this smart pipe system between your organisation and reduces the mundane tasks. These types of tasks typically fall under the heading Robotic Process Automation and can involve anything from automating change of address processes (affects multiple systems), benefits administration, statement reconciliation, various compliance activities, manual order processing, or even training (as we will see in the next section).
The advantages of combining data and automation are intuitively clear. Conversational interfaces, however, may not be such an obvious complement. What do we mean by conversational interfaces and why do we think they are going to play a crucial part in changing the workplace of the future?
A conversational interface is one where the main means of interaction is through a conversation. The way we all learn to interact from our very first years. A chatbot is simply a software program that we interact with by conversing with it rather than manipulating web pages or windows. The benefits come from two sides. Firstly, we are most likely already within a conversational interface interacting with colleagues. Whether it is Slack, Skype, Microsoft Teams or Facebook for Work we are increasingly using the same tools we use for interacting with friends and families to get work done. Messaging platforms such as Facebook Messenger, WhatsApp, Telegram, etc.
As the Slack tagline loudly says, conversational interfaces are “Where work happens”. A chatbot is a way of providing access to data and automation through the same interface that we use daily and within the context of work discussions already taking place. It is far more natural, effective and more likely to react and respond to information provided via Slack than email or having to pro-actively think of logging into a specific service in order to access a small piece of data. A chatbot can proactively come to your with relevant information and right within that context provide you the tools to take action - just like our example at the start of the article where we find out about something relevant and create a task for the right team to action further.
GDPR is a perfect example of how chatbots can prove extremely useful in providing widespread training. Instead of having to task someone with the job of following up with people to make sure they’ve read the relevant material on GDPR, a bot (such as the one we have recently developed and which can be installed on any Slack team) can do it automatically. For example, our GDPR Training bot will proactively contact all team members and walk them through the relevant GDPR training material. It also offers a quiz at the end of it with scores that can be compared throughout the organisation adding a fun element that will make it more likely and more entertaining to go through the material.
Just like the point and click interface paradigm opened up computing capabilities in the 1980s, conversational interfaces will be our window to the incredible combination of data and automation.
Agile digital transformation through data, automation and conversational interfaces
Digital technology is already changing shaping how we work in a myriad of ways. While a lot is said and written about digital transformation, once you cut through the marketing, what remains is the ability of an organisation to change how it works through the introduction of digital technologies. Conversational interfaces offer a natural way of introducing changes without having to go through lengthy processes and costly IT projects before anyone perceives benefits. Introducing a training bot or a marketing data analytics bot into the daily habits of people shows the benefits immediately. It allows people to start the most important transformation, which is a cultural one, by seeing tangible change and positive impact with minimum intrusion.
Ronald Ashri, Technical strategy, Deeson
Image Credit: PHOTOCREO Michal Bednarek / Shutterstock