How bots help businesses establish omnichannel communications


How bots help businesses establish omnichannel communications

Thanks to the automation of communications, you can achieve impressive results: from saving employee time to increasing customer loyalty. The main thing is to apply the omnichannel approach and modern AI-based bots

Chatbots are "getting smarter" and are being used more and more often

Technology is changing the way businesses communicate with customers right now: more and more companies are adopting chatbots. And these are not just quickly made "on the knee" bots like those that each of us met in Telegram. These are serious self-learning solutions based on artificial intelligence, behind which there is a huge work of many people and a serious infrastructure.

According to Gartner, the total volume of the global chatbot market will reach $ 1.3 billion by 2024. Around the same time, 47% of companies will use chatbots continuously, and 40% will deploy more functional virtual personal assistants. In Russia, the scale is more modest, but the growth rate does not lag behind. According to Accenture, in 2019, the volume of the domestic chatbot market amounted to ₽ 1.5 billion, and the researchers estimated the annual increase at 30%. They also found that 60% of the surveyed representatives of 100 companies from 18 different industries already use chatbots or test them as part of pilot projects. At the end of 2021, the company Just AI estimated the market volume of the entire conversational AI in Russia at $ 80 million Researchers noted that by 2025 it could grow 7 times - up to $ 561 million.

The main customers of chatbots in Russia are retail and the banking sector, but the solution is suitable for any company that communicates with its customers in one form or another and has its own contact center. From the introduction of technology, both business and those with whom it communicates benefit. For the company, chatbots reduce the burden on call center employees, reduce the processing time of the request, allow you to manage all the dialogues from a single system that simultaneously collects a lot of useful statistics. The client receives a response as quickly as possible. And if the bot is made qualitatively, then the answers will be accurate and filled with useful recommendations: the one who asks will be satisfied with the result. It is also extremely important that the chatbot does not need to sleep and go on vacation – it works on the line 24 hours a day and 7 days a week. For Russia with its 11 time zones, this is especially true. And this is valuable for users: Oracle notes that more than 50% of customers want the business to be available online around the clock.

You can create a simple chatbot in a few minutes. It is also possible to implement a third-party solution of any level quickly enough, but the independent development of a serious self-learning solution based on AI from scratch is a matter of at least several years. Teaching a bot to respond is easy. But to create a kernel and an engine, the entire intellectual part that makes it "smart" and "human", is very difficult. And then all this needs to be maintained and updated regularly. If the bot is firmly rooted in the company, then over time the number of tasks that it solves will begin to grow significantly. In short, development is a difficult obstacle for many, but it can be circumvented: subscribe to a bot.

One of the largest providers of such a service in Russia is MegaFon. Because the solution is cloud-based, you can connect it in just a few days. The cost directly depends on the number of dialogues: the basic tariffs include 5 thousand, 20 thousand or 40 thousand communications, and each additional open dialogue in excess of the package will cost ₽ 5-12. In the case of the maximum package, the monthly budget fits within the framework of ₽200 thousand, and the minimum starts from ₽60 thousand At the same time, the web interface can be used by an unlimited number of employees, it has many standard templates for the knowledge base, plus the service provider is always ready to help with technical issues. The basic packages include integration with e-mail and Telegram, as well as a widget for the site, but the customization options are flexible - on request, you can add almost anything.

As an example of successful integrations, MegaFon cites the cases of an online school and a jewelry store. In the first case, the customer was able to successfully close 25% of calls without the participation of contact center specialists and speed up responses by an average of 37%. In the second case, the problem was that the specialists of the call-center during sales managed to quickly respond to only a third of incoming requests. With chatbots, the company managed to cover them all with a peak load of 100 thousand calls per day, without increasing the staff. From 30% to 45% of them the bot closes on its own.

Voice bots are already selling, hiring and not only

The next stage in the development of chatbots is voice bots. The reason why they are gaining popularity lies in the growing number of voice requests of people. Back in 2018, Google reported that 27% of all users use voice search on mobile devices. And PWC researchers found that 71% of users prefer to use voice queries instead of text queries. The popularity of "smart" speakers in the US reinforces this theory: according to eMarketer, by the end of 2022, 4 out of 10 residents of the country will have such a gadget.

The voice bots themselves will get smarter literally before our eyes. A good illustration of this is Amazon's Alexa voice assistant. It is not represented in Russia, but in English-speaking countries it occupies a significant market share. In 2017, Alexa in its home American market could perform about 7 thousand different actions. In 2018 - about 26 thousand In 2019 - more than 55 thousand In 2020 - more than 70 thousand, and at the beginning of 2021, the meter overcame the mark of 80 thousand actions.

The basis of a voice bot in the service of business is exactly the same as that of a chatbot: an extensive knowledge base, a core, an engine and complex machine learning algorithms with which the bot works with the available information. There is only one addition: it gives answers to users not in the form of text, but in the form of a voice. Two approaches are possible. Artificial speech synthesis is cheaper, easier and faster, but the result is less "human" - in words there seem to be several accents, and intonation accents arise suddenly. The second option is pre-recording, when all possible phrases are immediately read entirely by a professional announcer. This approach is used by MegaFon for its bots.

Another important difference between a voice bot and a chatbot is that the second waits for people to come to him with a question, and the first can be proactive – to go to customers himself. The "out-of-the-box" subscription solution from MegaFon, for example, provides for about half a dozen different scenarios. A voice bot can conduct surveys, engage in "cold sales", inform or remind about something, as well as verify delivery. The connection is as simple and fast as in the case of text scripts, and operators get a ready-made interface in which everything can be very quickly and simply customized to the needs of the business. Moreover, some typical scenarios can be customized: for example, call the bot Angelina or Konstantin and, based on this, re-voice the scripts.

The developers of MegaFon place special emphasis on the fact that the bot is not a replacement for a living person, but his assistant. For example, a voice bot can do something that contact center specialists almost never have time for: measure the consumer loyalty index (NPS). The realities of business are such that they are often made according to the residual principle: operators conduct surveys when there are few calls to call centers. But for a qualitative assessment, a large sample is needed, and a voice bot can easily cope with this. Since the questions in this scenario are typical, and the communication itself is short, very few users will be able to distinguish it from a living person.

Another scenario for the use of a voice bot is the automation of part of the tasks of the HR department in the process of hiring line personnel. For example, you need to hire a few dozen cashiers and call a lot of people for this. According to the internal data of MegaFon, about 85 calls out of 100 will be made "idle". If you imagine that each needs 3 minutes, this is 5 hours - more than half of the working time of the HR-manager for the day. The robot can call simultaneously tens, hundreds, thousands of people. He will solve a similar problem in just a few minutes: he will conduct an initial survey, better understand the peculiarities of dialects or, perhaps, even communicate with applicants from other countries in their native language. For those who are suitable, he will immediately appoint a full interview based on the dates available in the HR manager's calendar, and then remind them of them and, if necessary, adjust the time and date of the meeting. This frees up a huge number of hours for the hiring employee to perform other higher-level tasks, which usually abound. By the way, according to Gartner, in 2022, 35% of companies will use such scenarios to one degree or another. And job seekers are more likely to be ready for this than not. In the Allegis survey, 66% of respondents said that they are comfortable communicating with a bot during the preparation for the interview.

The third interesting scenario is "cold selling." In the vast majority of cases, the caller is communicated rudely and disrespectfully. Employees of call-centers react emotionally to this, quickly burn out and leave. The robot is not only completely devoid of emotions, but can literally make tens or hundreds of times more contacts in a day. The main thing is to put in it a really interesting and useful presentation of the product being sold. Experts of the American Forbes notes that chatbots are able to raise sales by 67%. True, in such a scenario, people in 87% of cases still prefer to communicate with their own kind, and not with a machine.

Intelligent Call Handling: Exclusive Voice Assistant

The next level of development of voice bots in MegaFon is called intelligent call processing. The general idea is that this is a much deeper, very personalized level of customization in relation to the customer company. The scope of application is the same: informing, reminding, service, quality assessment and sales. However, within them, the solution provider can develop whatever the customer wants.

That is why the cycle of implementation and implementation here is longer. We need more stages of testing, it takes more time to record the answers of the voice assistant. And it is very important to comply with the requirements of the laws on personal data. If you want to work with customer response records, it is important that customers consent to their processing by third parties. This subtle nuance often forces you to rewrite the answers again and lengthens the implementation of the project by 3-4 weeks. Payment in the market of such solutions is almost always tied to the installation payment, because we are talking about individual development. However, MegaFon has a different approach: all payments go to pay for future traffic. As for a regular voice bot, there is a grid of tariffs, where the cost depends on the total duration of calls per month: from about ₽6.7 to ₽8.5 per minute.

Omnichannel instead of multichannel

When developing bots and in general everything related to the activities of the contact center, it is extremely important for business not to forget: more than 60% of customers use different communication channels for communication. A PWC study found that nearly two-thirds of customers say goodbye to a company after a few instances of negative communication, and 19% leave after the first such experience. That is why it is necessary to move from multi-channel communication to a more modern omnichannel model. Its essence is in unification: on the side of the call-center, all points of communication with the user are aggregated in a single system. If the communication began through VK, but then the person wrote again through Telegram, then the support employee will see all this and will be able to continue the dialogue with a convenient point for the client, and not start it again and find out all the details. This solves an important problem: 9 out of 10 customers are annoyed when a support specialist begins to re-find out the details of his problem. And if the business itself wants to appeal to a client with whom there has already been a dialogue, then it will immediately be clear which communication channel is most convenient for a person.

The effectiveness of the omnichannel approach to communication is confirmed by research. According to Omnisend, it allows you to increase the customer retention rate (Retention Rate) by 90%. Omnichannel also affects the customer's life value indicator (LTV): for companies applying the approach, it is 30% higher, and this is already directly related to revenue.

The main problem is that the independent development of a single system for monitoring all communications of the company with users is a task on the border of the impossible. The fact is that in addition to deep expertise in the field of development, you will need the expertise of many communication specialists: each channel has a lot of its own special nuances. According to MegaFon, russia is unlikely to have more than ten corporations capable of coping with such a challenge. An alternative to a subscription simply cannot be found if you need a comprehensive solution, and not its disparate components. The best experts, of course, are from telecom operators.

The omnichannel solution can be installed both on cloud servers and in the customer's IT circuit. The budget depends on the needs for customization of the platform and the need to create non-standard scenarios for working with appeals (from ₽ 5 million and above to a call center with 20 accounts). If there are no problems with this, then from the confirmation of the project to the start of the system takes an average of about 3 months.

As a result, on its side, the business receives a convenient and adapted to specific tasks interface: an automated workplace of a call center employee with a very wide range of opportunities and unique scenarios for working with appeals to solve narrow-profile tasks. There are answer templates, you can work simultaneously with several sessions, all information about the client is displayed, the history of interactions with him is stored and his entire path from the first contact is shown, from here you can control voice calls and the capabilities of voice and chatbots. The supervisor of the call-center has its own separate dashboard for monitoring the load, distribution of queues and reloading of ordinary employees. Plus, of course, there are many tools for analytics to eliminate inefficient load on contact center employees, for example, filtering spam calls and tools for working with quality of service assessment.

On the client side, it looks like magic. For example, he reports on WhatsApp about a problem and gets a primary solution. In the process, he realizes that it is inconvenient for him to communicate here and continues to communicate already in Telegram - for the client this happens imperceptibly, because the employee sees the client's appeals in all channels. Then the person thinks about the decision and realizes that he needs some kind of file: he calls on the phone, asks to send it and receives it immediately in Telegram. At the same time, a large document with detailed information on his request comes to the mail. It is quite natural that with this level of service, the loyalty rate will grow very seriously. A company that does not just solve problems, but does it in a convenient format for you, you always want to come back again.


Popular posts from this blog

"Investors need a minimum." When Bitcoin Will Begin Recovery

When Bitcoin Will Be More Expensive Than $100 Thousand: A Selection of Predictions

In the US, launched mining at a nuclear power plant