29. In summary,
• What are chatbots? Future potential
• Different types of businesses (B2B, B2C, etc.) that can use chatbots
• How business processes can be personalized with chatbots?
• Benefits of using them for the organization
• Deployment strategies (based on organization size)
• Challenges (people, process and technology)
31. References
• Cameron , G., Cameron, D., Megaw, G., Bond, R., Mulvenna, M., O’Neill, S., . . . McTear, M. (2017). Towards a chatbot for digital counselling . Retrieved
October 12, 2017, from http://hci2017.bcs.org/wp-content/uploads/BHCI_2017_paper_110.pdf
•
Cearley, D., Burke, B., Searle, S., & Walker, M. (2017, October 03). Top 10 Strategic Technology Trends for 2018. Retrieved October 12, 2017, from
https://www.gartner.com/document/3811368?ref=solrResearch&refval=192157643&qid=75554fd63dea0ed47dd208ff63a829df
•
Fugu Chatbots : Revolutionizing Brand – Customer Interaction. (n.d.). Retrieved October 12, 2017, from http://click-labs.com/launching-fugu-user-
insights-marketing-automation-bot-apps-2/
•
6 Clever Ways to Drive Website Engagement With Chatbots. (2016, September 13). Retrieved October 12, 2017, from https://blog.proto.io/6-clever-
ways-to-drive-website-engagement-with-chatbots/
•
Merritt, Arte. “The Language of Chatbots – Chatbots Magazine.” Chatbots Magazine, Chatbots Magazine, 19 July 2017, chatbotsmagazine.com/the-
language-of-chatbots-99a38be54b58.
• Kostick, C. (2017, June 13). Five Situations Where You Should Use a Chatbot – Chatbots Magazine. Retrieved October 12, 2017, from
https://chatbotsmagazine.com/five-situations-where-you-should-use-a-chatbot-5a3f97fd56da
•
Kim, L. (n.d.). 11 Interesting Examples of How to Use Chatbots. Retrieved October 12, 2017, from https://www.inc.com/larry-kim/11-interesting-
examples-of-how-to-use-chatbots.html
•
(n.d.). Retrieved October 12, 2017, from http://www.icmi.com/Resources/Customer-Experience/2017/03/What-Do-Customers-Really-Think-About-
Long-Wait-Times
Editor's Notes
In the summer of 2014, I was in Europe. I wanted to travel from Dresden, Germany to Prague, Czech Republic. So, I googled around for trains and arrived on the national German railways website (Deustche Bahn). Although today the website is much friendlier, it is pretty bad then. Firstly, the website was in German and I just couldn’t find the German to English translate option. However, a railways booking website is somewhat predictable, so I booked my train with some guesswork and Google Translate help. Now that’s a happy ending, right? Nope, it isn’t. I booked a First Class ticket, but I had no seat on the train. It turns out buying a ticket just ensures that you are allowed to board the train. If you want a seat, you have to pay an additional reservation fee (5 euros, I think).
Was there a way to improve this situation? Well maybe not in 2014. But today with advances in technology, chatbots and other virtual assistants can come to our rescue.
So, what are chatbots? We are all familiar with a version of chatbot that comes pre-installed with our phones, e.g. Siri for Apple, Google Assistant for Android, etc. A chatbot is a neat piece of software that can help you do a variety of things – set a reminder/alarm, show photos, play music and answer basic questions. Sometimes, they try to emulate human features like humor or in this case, sass.
However, applications of chatbots extend beyond personal use. Today, they are commonly used by companies to automate their customer service functions – resolving basic issues and/or forward the call to a subject matter expert.
However, that is just the tip of the iceberg. Chatbots are capable of so much more. With advances in Natural Language Processing, Machine Learning and Robotic Process Automation, there are limitless opportunities.
I would argue that chatbots will fundamentally change the way we interact with devices in the near future.
Going back to my Deutsche Bahn example - Imagine when I went to the website, I was greeted by a somewhat humane bot. They could probably detect my location and speak to me in English. It could access my Calendar and understand that I probably have meetings in two different cities. It could ask me if I want to connect my social media and payment app (Facebook, PayPal, etc.) information. Furthermore, it could pull some information and arrive at an accurate estimation of my intended trip. So, instead of opening a browser, looking for English, filtering trains according to budget, etc., the bot would make a reservation with very little input on my trip, just like a friend or family member would.
There are 3 things that make chatbots unique and powerful – context, input variety and multiple devices. Chatbots can parse context in terms of location, relationships, etc. to make intelligent recommendations. They are capable of taking inputs in different forms like text, sound, image, video, etc. and can be embedded into various ‘smart’ devices like phones, laptops, computers, speakers,.. even home automation devices.
One of the biggest areas where chatbots can make a huge difference is the overall customer experience.
Customer experience is a cover-all term. In fact, a customer could have different meaning in different organization types an therefore different applications for chatbots.
Let us start by exploring the generic CRM (Customer Relationship Management) process. These functions stay the same across all types of organizations. In a survey conducted by an Inbound Marketing Agency – Incisive Edge, they found what consumers think would be the most useful applications of chatbots. Some of the major ones are quick emergency answers, forwarding to humans, buying basic items, complaint resolution, etc. You can see 46% think that quick emergency answers would be the best application. Only 4% thought that app consolidation was an useful application. However, as the technology matures, the results are bound to change and consumers will appreciate the benefits of consolidation.
Let us move on to the B2C (business to consumer) segment. Multiple functions could utilize chatbots –
1. Customer Service – This is the most familiar usage of chatbots. With the addition of more context and more humane features, they can be almost indistinguishable from human representatives.
2. Sales & Marketing – Using messaging, organizations can actively identify leads, prospect them and personalize their products and services according to need. They can reach consumers - sharing deals, promotions, offers, in a more personal way.
Bill Payments – As customers and organizations interact through a data-rich platform, it can quicken the billing process with one command.
Feedback – Customer engagement is a two-way street. Gathering feedback from clients is made easier with chatbots.
For B2B (business-to-business) organizations, the customer experience is targeted at maintaining good vendor/ supplier relationships. Complicated processes with many steps (like procure-to-pay) can be simplified with chatbots. They form the perfect user-interface for processes that involve multiple systems. In the procure-to-pay example, the sub-processes of supplier selection, billing, invoicing can be taken over by the chatbot.
Finally, internal customers (or employees) can benefit from simplification of processes. For example, a HR (Human Resources) professional has to create duplicate entries on at least 3 to 4 systems when onboarding a new employee. She has to create a entry in the payroll system, benefits system, IT system, etc. A chatbot command of ‘induct employee XYZ’ can reduce that load and help her to focus on more productive tasks.
There are many advantages to implementing chatbots/ virtual assistants, namely cost savings, 24/7 availability, quality improvement, scalability.
The chart is the potential annual US salary savings created by chatbots used in insurance companies, financial services, sales and customer service.
From the chart, the total annual salary expenditures for Insurance Sales Representatives, Securities, Commodities, and Financial Services Representatives, Sales Representatives, Customer Service representatives are 20, 32, 43, 79 billion USD, respectively. The most cost is for Customer Service. But Chatbots can potentially save about 30% of the total expenditure on annual salary.
In the long run, chatbots will create savings on salary expenditure, not by replacing labor employment, but by saving staff hours to work more productive. This graph is against the potential replacement of jobs by using chatbots.
For BOA, they used to have 24/7 customer service for credit cards. But they started to have automated 24/7 customer service for debit cards.
AirBnb, Evernote, and Spotify also started to use chatbots on TWITTER to provide 24/7 customer service.
Therefore, chatbots have been adopted to provide better customer service experience by responding to customer requests anytime needed.
Quality can be improved in 4 ways.
1. Attitude: The attitude of customer call representatives can influence the quality. Chatbots can be programmed with “humor” feature, which brings better customer service experience with positive attitude. Don’t know the humor of chatbots? Try to say “Do you know beatbox?”to your Siri.
2. Language: Chatbots has been used commonly in 10 languages: English, French, Spanish, Portuguese, German, Italian, Arabic, Russian, Chinese, Dutch, and more in various languages.
3. Culture: Chatbots can be programmed to distinguish different cultures and locate where the calls are from, so they can customize what and how they will respond to customers based on culture to avoid some unnecessary offenses, which may lead to bad customer experiences.
4. Holding time: In the 2017 Arise Survey, it showed, Almost two-third said they would wait two minutes or less and 13% said that “no wait time is acceptable.” But chatbot can serve multiple customers at the same time to shorten the holding time.
With any businesses growing fast, there could be some challenges to scale the resources.
For example, as the business grows, there will be more customers needing more customer call representatives to provide service. But we can just use chatbots to solve this challenge by creating more bots in the system, which it’s easier and faster.
Deploying bots for your business can be tricky. There are 3 strategies to help you deploy bots –
Converting your website/ mobile app to a bot
Platform integration
Hybrid
Converting your website/ mobile app to a bot – This is especially suitable for large enterprises, with enormous amount of customers and web traffic.
How e-commerce functions today without chatbots.
How e-commerce could be improved with platform integration. Small-to-medium enterprises benefit from this type of deployment.
This diagram from Gartner reflects the architecture of a chatbot. It helps us appreciate how a chatbot consolidates input-output channels with third-party services.
Hybrid – This deployment strategy should be considered for custom requirements.
In order to make chatbot better functional, it should know you better than yourself
So you have to give access of your private information
Ongoing issue right now and has to working on it.
*Fail to understanding- lead to wrong info or discontinue the communication
-Today, the focus for AI is on "narrow AI". Narrow AI consists of highly scoped machine-learning solutions that target a specific task
-Artificial general intelligence refers to the use of machine learning to handle a broad range of use cases.
A lack of the relevant data sciences will probably hamper AI adoption in the short term.
*Counterfeit chatbot – mislead customer to fake/unauthorized website
- SSL Certificate
-Blockchain(Using a public blockchain can remove the need for trusted central authorities in record transactions and dispute arbitrations. This is because trust is built into the model through immutable records on a distributed ledger.)
*Serving Malicious URLs
- intelligence software to detect
Also, there is a concern of privacy
In order to make chatbot to be more friendly use, it should know you as well as your best friends or better than yourself
So you have to give access of your private information
Ongoing issue for every technology facing right now and has to keep working on it
Change Management is hard. Although the process will become simpler, we need to deal with organizational inertia. In addition, people may be fearful of the bot taking their jobs, which could lower employee morale.
Chatbots have the risk of being used inappropriately. In 2016 Microsoft launched a chatbot on twitter, known as Tay. This chatbot was programmed to learn and form replies based on its interactions on twitter. This caused people to interact with the bot inappropriately, using offensive language and content which ultimately caused Microsoft to take the chatbot down after only 16 hours