The Many Functions of the Chatbots

A chatbot is a hot topic. How to utilize a chatbot to boost your business?

The chatbot is one of the critical features of modern digital sales and customer service. Chatbots help customers 24/7. What kind of bot would be the best fit for your company? Book now time for a free chatbot consultation!

Chatbot

What is a chatbot?

Chatbot on tietokoneohjelma, joka automaattisesti reagoi saamiinsa viesteihin. Chatbotin tarkoituksena on simuloida ihmisten välistä keskustelua, vaikkakin chatbotin kautta keskustelua käyvät ihminen ja kone.

Chatboteja on olemassa toimintaperiaatteiltaan erilaisia:

Multiple choice bot

The simplest way to build a chatbot is to set it to offer a predefined set of questions and answers. The chatbot can launches when the website visitor's behaviour matches chosen rules.

For example:

A visitor browses the website of a mobile phone network operator. When the visitor has been browsing the site for a while, a chatbot appears. The chatbot asks:

"How can I help you? Are you looking for..."

  • a new mobile phone
  • internet connection package
  • customer service contact details?

Chatbot can offer different topics for the website visitor so can find more easily the information they are looking for.

Do you want a more detailed analysis of how a chatbot could boost your business? Book now time for a free demo!

The chatbot should always solve a real business problem

Building a chatbot should be the result of a company strategy implementation. The chatbot provider should customize the chatbot to fit the client's needs.

One use of a chatbot is to fix the website's mistakes. For example, if the website user experience is challenging to navigate, bot fixes the problem but not the root cause.

Is the aim of the chatbot to increase sales or cut costs? If the company automatizes customer service tasks, there might be a consequence that there are fewer calls to the customer service centre.

The history of a chatbot

Even though the term chatbot has surfaced during the past five years, it's not a new phenomenon. The Turing test is one of the starting points of the chatbot and machine development. It was made famous by British mathematician Alan Turing. Alan Turing researched artificial intelligence in the 1950s.

Bots are part of our everyday routines in mobile phones. For example, Apple's Siri and Google Now (that many Android phones have) are virtual AI-based assistants. They help to find information and control the phone by voice commands.

Machine learning algorithms that bots use were developed decades ago. But why have bots risen to surface right now? There are two reasons for this.

Technological development

The amount of computing power has risen exponentially during the past ten years.

Digital change

Digital service channels and e-commerce shopping are here to stay. Worldwide e-commerce sales are predicted to be almost 4 billion euros in 2021.

Are you interested in using chatbots to enhance your business? Book now time for a free demo!

What are the benefits of a chatbot?

Do you want more out of your customer service resources? Or do you want to boost e-commerce sales or leads?

Our client Telia was chosen as the Digital Winner of the Year 2019. Our chatbot is part of Telia's successful digital client management solution.

Chatbot will help your company e.g. to:

  • to boost sales
  • enhance customer experience
  • prevent churn
  • divide your customer service rush hour traffic more evenly
  • put your resources into more effective use

Do you want to learn everything about chatbots? Check Chatbot Academy!

More sales with chatbot

Chatbot will grow the sales of your digital channels. Bot can recognize the buying customers' needs and help them to complete the purchase processes. Chatbot can also used to gather data from the buying process and frequently asked questions.

Customer satisfaction and chatbot

Serve your clients even better around the clock. Consumers' individual needs create challenges to the e-commerce sites customer service. If a customer cannot find what they are looking for, they switch to another website. With a bot you can guarantee that the customer can find what they are looking for in the middle of the night.

Preventing the churn

Churn minimization is the key success factor in subscription-based services. Chatbot can be used to gather information on behaviour that is related to ending a subscription.

Using customer service resources more effectively

Discussion related to machine learning and artificial intelligence is often circling around man versus machine -topict. It's very important to understand that a useful chatbot needs also the human and the machine to function properly.

So far the customer service chat is one of the key elements in online customer service. Bots offer new possibilites to online customer service.

"Bots take over customer service jobs."

This is one of the most common misconceptions related to chatbots. In most cases the bots free the customer service agents time to handle the challenging customer service situations. Best results are a combination of a chatbot and customer service chat. Chatbot can ask the routine info questions and validate the need and then the discussion can be handed over to a customer service agent in the same chat.

"Even better customer service online exactly when the customer needs it."

CASE: What are the Nordics's biggest telecommunications operator Telia's experiences about chatbot?

Chatbot and customer experience

The clients' needs should be the starting point of a chatbot project. After a chatbot, the customer experience should be same or even better than previously. We recommend testing the chatbot with a small group of customers first. Read more about chatbot and customer experience.

Interested in a free demo? Book now time for a chatbot demo!

How to start planning and building a chatbot?

We divide the chatbot construction to three different main phases:

  • Current situation analysis and goal-setting
  • Technical application
  • Continuous development

The chatbot project can be a big process for the company connecting sales, marketing and customer service departments.

The current status analysis

Chatbot requires a lot of internal work from the organization. Highly functioning bot has to be connected to different departments' daily routines. If the bot is the new customer service aide, the customer service needs to clarify the goals for the bot.

If the aim of the bot project is to cut the customer service costs, we analyze the current situation and find development goals based on customer service calls and categorization information.

At the same time it's useful to consider the project resourcing: do the company silos and siloed budget prevent the effective use of the chatbot? If every department has their own goals, they might not want to spend their budgets to build a company-wide chatbot.

A chatbot project has various "hidden" costs that are inevitable but not easy to spot. The price of a chatbot is not just the sum paid to the chatbot provider. Internal planning, integrations to other systems and tools take time and resources from the company as well. That's why the chatbot project requires dedicated personnel to manage the project.

On the other hand, the chatbot makes it possible to offer the customers totally new possibilities. Chatbot provides customer service 24/7.

Of course, the chatbot is not always more effective tool in customer service. If the company IT systems are old-fashioned and slow down the customer service process, can the benefits of a chatbot stay on a neglible level.

What kind of a bot might suit your company? Book now time for a free chatbot consultation!

Project needs to have the management's mandate but space to operate

Are you a leader? Your role is to encourage others and create an accessible atmosphere. You got to highlight the importance of the project but give free reins to the operative level. Don't set conditions that prevent innovations to happen during the project.

The next step is the most challenging thing in the chatbot project: the problem that chatbot is set to solve has better to emerge from real business needs. But there should be room for innovation and improvisation because we've learnt that there will always be surprises along the way.

Positive surprises offer a new ideas to serve the clients in ways that were beyond imagination in the beginning of the project. So it's important to have a flexible organisation and an open atmosphere. This requires trust from the management to personnel and chatbot provider.

This is often hard for corporations who need predictability. For example buying a chatbot and buying a CRM system are totally different kind of projects even though they are often handled in the organization according a same process. Lue lisää chatbot editorista. 

Unpleasant surprises are often related to data and the time that it takes to gather that data. Usually chatbot projects take more time than organizations expect.

But preparing to gather the data is to be one step closer to a ready bot.

Project resourcing

Staffing the chatbot project is not an easy task because there are various stakeholders present. There should be at least experts from business, technology and customer service. Customer service personnel might have fears that the chatbot will dispose of their jobs.

But chatbot will never totally replace the human work force. Chatbot frees the human employers time to handle the challenging customer encounters. Customer service personnel usually has a lot of tacit knowledge related to company customers, products and services. It's beneficial to use this information when building a chatbot.

When the project starts the management has to keep an eye on the workload so that the project participants are able to use the time there is allocated to the chatbot project. If the employers are asked to fulfill other duties as well, will the bot never be finished.

Chatbot and integrations

Chatbot requires data to function. But what is this data? First it's important to consider if there is data already or should it be gathered. There might also be different systems in use so that the data needs to be compiled from various sources.

In this phase it's useful to research the requirements for the integrations. Most common integrations for a chatbot are customer and product information systems. Does integration work in two-way or just one way?

When choosing the chatbot provider it's relevant to ask for integration references too.

Chatbot scenarios

A chatbot is safe because it will be programmed to offer certain answers so you don't need to be afraid of trolls. Chatbot can be used to open for certain website visitors - e.g. those with the best revenue porential. That is called chatbot scenario work.

Chatbot scenarios can be build for example to show the chatbot only for the visitors from certain sources or keywords.

Book now time for a free consultation! Book a demo!

Chatbot planning

  • Decide the business problem that the chatbot solves
  • Involve expert from different departments
  • Understand what integrations chatbot needs to function properly
  • Plan scenarios

Technical application and training

When analysis and background work are completed is the technical application of the chatbot much faster.

Chatbot editor

Chatbot editor is used to build the chatbot flows. We offer our clients their own chatbot editor account that the clients can use to modify the chatbot. Lue lisää chatbot editorista. 

Natural Language Processing

The menu bot is the simplest way to set up a chatbot. More advanced bots utilize natural language processing. This means that the chatbot can search database for answers to match a query the user has written.

On the other hand the chatbot can be programmed to recognize certain words or phrases from the text. If a bot asks for a website user's home town and they answer Helsinki, can the bot offer contact details for company's retail stores in Helsinki.

Then again there are peculiarities that have to given thought to. From example Helsinki can also be "helsinki", "Hki", "hesa", "stadi", "Helsingfors" etc. The website user can also misspell "helsiki".

The chatbot whisperer

A chatbot needs teaching. Bot does not learn without human's help. Like a customer service agent, a chatbot needs continuous training.

Text recongnition is usually implemented so that the chatbot is taught different phrasings. There are also conditions how to react to these phrasings. The person who teaches a chatbot is a chatbot whisperer. The chatbot whisperer is a link between customer service agents and the algorithm.

Roadmap for development

Typically the chatbot project is launched with a menubot that is later developed to utilize natural language processing. This way the bot is added a search query field that is used to search for more complex information.

What kind of companies and communities benefit the most from a chatbot?

Chatbot can be tailored to fit the needs of different industries and companies. It can also be beneficial for company internal use.

Of course, it's important to define the goal of the chatbot - it's not useful to build it just for the sake of it.

Also, public sector can build a chatbot even though no product would be sold through the bot. Chatbots are experts when the customer resources are scarce. Chatbots can be developed to answer the most frequently asked questions when it comes to public services. Interested in learning more about chatbots in internal company use?

Read more from our blog:

The chatbot project homework

Ask these 10 questions from your chatbot provider

How Telia has utilized chatbot

Artificial Intelligence and Chatbot for Real

How to Build a Chatbot

Chatbot Q&A

How Public Sector Can Benefit from a Chatbot?

Are Chatbots Safe?

Interested in a free demo? Book now time for a free consultation!

 
Scroll to Top