Blog " Will 2023 be the year of AI chatbots in our country?
During the first edition of Sales Club Conference 2022During the first edition, Elitsa Stoilova presented to the audience why businesses need AI chatbots and explained everything about them from A to Z in her 90-minute workshop.
The workshop made it clear how conversational AI chatbots become a key tool for businesses by automating communication, extending company working hours, optimizing employee workflows, serving customers—especially night owls—increasing sales, and overall supporting the digital transformation of the business.
The digitalization of business is a continuous process, involving the ongoing transformation and updating of companies’ digital tools to remain competitive in their environment. The rapid pace of technology not only presents challenges but also offers numerous opportunities for growth.
By the end of 2022, the global tech landscape began revolving around OpenAI’s ChatGPT following its launch in November of that year. It also became clear that Microsoft plans to invest 10 billion dollars in this artificial intelligence tool.
You have been working with business chatbots for a long time. Umni.bg provides businesses from various sectors the opportunity to acquire this solution, adapt it to their needs, and make it a real assistant. Do global technological trends affect us as a market, or in this regard are we still lagging behind geographically? What is the future of this technology here?
I first heard the word “chatbot” in 2017. in a YouTube video about chatbot paralegals (assistants for lawyers) in the U.S. Since I had worked 17 years on American soil, I knew very well what a paralegal was, and a chatbot… hmm… obviously some kind of software. It was love at first sight: if it can be a paralegal, then it can work for anyone! Let’s see what this is! That initial curiosity sparked Umni.bg.
I started talking to people about chatbots—at business meetings, conferences, and other events. Back then, the word “chatbot” was truly unfamiliar in our country. I had learned it like a mantra and quoted it hundreds of times.that “a chatbot is a software solution, with or without conversational AI, which is integrated into a website or elsewhere, simulates human conversation in a chat format, providing instant information in the form of text, audio, images, and other media, and is used for marketing, sales, customer service, automation of routine questions and requests, and other purposes to meet the specific needs of a business.”
Today, it’s different. I rarely need to use the chatbot mantra anymore. Usually, the people I talk to have at least heard of or used a chatbot at least onceand they have some idea of what it is, although sometimes it’s superficial, incorrect, or outdated, because the technology evolves very quickly. Some people, for example, confuse a chatbot with live chat. The big difference today, however, is that whereas before I had to reach out to business owners and managers to present the idea of a chatbot for their business, now they contact our team themselves and are proactive in seeking information on how a chatbot can be useful for their team, clients, and business.
The pandemic also contributed to this, making people in our country more digital. At a slower pace, we are following global trends – it’s inevitable; we live in a global digital village where information knows no boundaries.
Since you mentioned ChatGPT, this AI chatbot is being called a revolution. Google and Meta promise to launch their own similar products very soon. Can businesses take advantage of it for their needs, and how?
ChatGPT is a language model with a “chat window,” and at first glance it looks like an extremely well-trained super AI chatbot, but it is not an AI chatbot. Simply put, it is a “creation” that can engage in conversation. on hundreds of thousands of topics, generating responses independently based on a vast database with billions of parameters, current as of 2021. Tools like it are still in the experimental and development stage—far from perfect and not yet reliable for business communication, unlike AI chatbots designed specifically for business use.
Language models are designed to generate an appropriate response, but they are not programmed to verify whether what they say is true. In practice, no one knows what response they will produce, and the best way to check the answer is to know in advance what result you expect. They can do many things incredibly well – for example, to generate text on various topics and in different styles, which is why they could be used as drafts for posts, advertisements, articles, emails, and reports and other types of materials—after being verified for accuracy, further developed, and finalized by the business team. This leads to significant savings in resources and allows staff to focus on the more critical aspects of creating a piece of content. I read about a student in Moscow who, within 24 hours, wrote his thesis with the help of ChatGPT and successfully defended it. The originality check of the final material, after the student’s revisions, showed over 80% originality. The text directly from ChatGPT was over 90% original.
Can ChatGPT be used as a business chatbot?
ChatGPT could be used for “casual conversations” to entertain users, and it can be trained to recognize and answer questions about a specific business, but it would still be only chat for FAQsand not a fully functional business AI chatbot. There also remains the issue of so-called “AI hallucinations” – the language model can generate a response that has nothing to do with reality but sounds extremely convincing as a statement. Other deviations are also possible, as warned by the developing companies – such as offensive content, discrimination, personal data issues, and more. A business has no control over the model’s responses and cannot prevent a “hallucination” or other undesirable effect. For example, and we asked ChatGPT in Bulgarian how to make taratorand the answer was that we need a large space where a car can be parked, and metal profiles for its construction. We can only guess what made the model respond this way, and whether it confused tarator with tractor. But there is definitely no need for such surprises in a business chatbot. A lot of work still lies ahead in this area.
So how is a business AI chatbot different then?
Business AI chatbots – chatbots with a conversational AI module – are tightly focused on the specific needs of a business, its staff, and its customers. They “live” within and operate as part of the business’s digital ecosystem. Globally, about half of them are implemented on company websites. They are created and developed by the business team in a controlled environment —The training topics, the responses of the conversational module, and other content and functionalities are 100% controlled by the business and are linked to the company’s products and services. For every single question, it is known exactly what the answer will be, down to the last comma—because it is defined by the business.
Of course, it takes time for the conversational AI in a chatbot to be trained to recognize all customer questions, but that is just a small part of all the necessary functionalities in a business chatbot. Business chatbots can sell, integrate with other business platforms to automate routine tasks, and be actively used for 24/7 customer support, marketing, as well as collecting first- and zero-party customer data, and customer behavior data, enabling the business to make more intelligent decisions. Another major difference is that the responses in the chatbot’s conversational AI module can redirect users to other parts of the chatbot, allowing the business team to create communication/marketing funnels within the chatbot that are triggered by a customer’s question. Additionally, each part of the chatbot has his own unique link and QR codewhich allows the business to personally direct users to a specific page on the website with a specific element in the chatbot from anywhere online or offline. Truly very different from what ChatGPT does.
You work with many and diverse businesses. Share your observations about their needs. What challenges do they face on a daily basis that could be easily addressed with the help of artificial intelligence and chatbots as assistants?
There is statistics that shocks almost everyonewhich we share, even though it’s on the surface and obvious. Ordinary businesses operating from 9 to 6 are, on average, idle 75% of the time throughout the year! That amounts to over 840 work shifts in 2022 – nights, weekends, and holidays. The same will apply in 2023 – 847 shifts, and this is the more expensive working time. Over 70% of companies struggle to find employees for the hours they operate. At the same time, up to 80% of employees’ time is lost on routine questions, tasks, and requests These are the first problems: how to extend the company’s working hours and get more done with fewer staff and lower investments.
These needs are also driven by the customers themselves – 72% of clients expect businesses to be available 24/7 for instant communication We see this in all chatbots – nighttime communication accounts for up to 45% of total interactions. Customers themselves have become more impatient and have higher expectations for a business’s technological adequacy – that is, to provide accessible modern digital tools, including self-service options. This puts businesses in a dilemma: how to meet these expectations without hiring more staff or paying external teams for customer support, while keeping clients satisfied and the company— competitive in the market The AI chatbot is one of the tools for reducing or solving this pain point.
The pandemic and the ongoing crisis only reinforce the need for digitalization, automation, process optimization, and cost savings… If I were to summarize, the needs of businesses remain classic, even though they are now situated in the digital world: to sell more, save more, while simultaneously improving customer care, and to grow. By the way, this is also the motto of Umni.bg: “Sell. Save. Care. Grow – with an AI chatbot.”
There are probably businesses—such as in goods manufacturing—for which an AI chatbot may not be suitable.
In my opinion, behind the statement “the chatbot is not for us” often lies a lack of in-depth analysis of business processes, unfamiliarity with the technology, and sometimes simply a personal dislike of the technology itself. Whether it will be for customer support, business partners, employees, marketing, or collecting feedback from end users — chatbot has his own place.
Recently, a business owner in the service sector told me that a chatbot would only work for 10% of his clients, using that as a reason to doubt the need for one. But is it really insignificant to free up staff from 10% of the workload and improve service for 10% of the clients? What else can employees do with that 10% of time for the other clients and for the company—tasks with higher added value? It’s simply a matter of calculation: the costs of a chatbot versus the savings, added value, and increased customer satisfaction.
Here’s another example from my practice: we were discussing the possibility of a chatbot with a machine manufacturer. At first glance, it really seemed unnecessary—the machines are sold to other manufacturing companies. However, during the conversation, it became clear that the machines operate 24/7, and when a warning light comes on at 2 a.m. and the machine stops, contact with the manufacturer’s technical support can only be made the next morning. We identified the problem and came up with the idea of a chatbot as an assistant for the client’s employees! When the warning light comes on, the employee scans a QR code on the machine’s sticker to access the chatbot. Inside the chatbot, a section opens with common malfunctions and instructions for that specific type of machine, so the employee can potentially solve the issue independently. There are also forms for preventive maintenance requests and repair requests. These forms send an email to the manufacturer, which is automatically captured by the CRM and opens the corresponding technical support ticket. The manufacturer has access to the information in the chatbot for each machine type at any time and can update and expand it with texts, images, videos, downloadable documentation, and more, while the QR sticker on the machine remains permanent. Access to important information for the end user – the machine operator, is just one scan away. The benefits extend to all parties.
If someone identifies with the category of “personal dislike of chatbots,” I have one question: what is more important—the subjective feeling of one person, or what the business, its clients, and/or staff actually need?
In your 2022 year-end review, you mention that Umni.bg doubled its client base compared to 2021 and also set several records. This indicates a significant change in business behavior regarding AI chatbots. What volumes do you expect for the upcoming year?
Let’s mention 2–3 records so that people don’t wonder what kind of record a chatbot can have. On our platform, we registered 70,000 users in a single chatbot in one day and 700,000 user interactions in a single chatbot in one month. The bots on our platform saved over 15,000 working hours for businesses last year.with some chatbots working the equivalent of 1 to 3 full-time employees at a company. There is also a significant increase in chatbot usage, with an average of 75% of users interacting with the chatbots two or more times.
Indeed, once businesses see such results—not somewhere in the U.S. or Asia, where chatbot usage has long been established as a standard, but here in Bulgaria, across different industries and businesses of various sizes—it becomes a convincing argument for investing time and resources into creating and using a chatbot. Moreover, the technology today is accessible both in terms of usability and financially, even for the smallest businesses—whether it’s a chatbot with conversational AI or without, depending on the specific needs of the business. In other words, the barriers to using the technology are mostly in the minds of thosewho decide whether there is a place for a smart digital assistant in the business or not. If someone is looking for a reason to say “no,” they will always find one.
I am optimistic about Bulgarian businesses and expect at least a doubling of interest in chatbots and their usage again this year.
What is the focus for Umni.bg in 2023?
With the Umni.bg team, we have accumulated 5 years of experience in the world of chatbots. While other platforms focus on the needs of corporations, developers, or messenger-based chatbots, we help companies transform their own website into an intelligent 24/7 communication hub with the help of an AI chatbot—without requiring any technical knowledge or skills. So they can have their own digital employee who handles marketing, sales, support, and data collection efforts, while “learning” from every conversation with a customer.
Although we also work with large companies, our focus this year will continue to be on the needs of small and medium-sized businesses. Our goal is to bring together in a single chatbot the most important tools and functionalities for the business and its clients, in order to achieve maximum benefits for all parties through the use of the chatbot.
Taking into account the accumulated experience of our team and the needs of businesses for active, personalized, and flexible use of their chatbots, we are preparing to launch a new version of our platform. The focus is on upgrading chatbots with artificial intelligence in the new platform into a business super-app by adding key functionalities for this purpose. We want the chatbots on our platform to become a tool for success for as many Bulgarian companies as possible. The technology is here, and we will continue to develop it. The ball is in the businesses’ court, and they shouldn’t delay. I will quote Peter Diamandis, the American entrepreneur, author, and futurist: “By the end of this decade, there will be two types of companies: those that use artificial intelligence, and those without a business.”
What advice would you give to businesses planning to use an AI chatbot this year?
Get familiar with the technology, plan carefully and for the long term, assign a person or a team to take care of your chatbot, and be patient: a chatbot is not a one-time effort, but a process of “growing” a personalized digital tool for your business, where the main voice belongs to the users of the chatbot. Listen to them, adapt it, and you will have a successful chatbot!
Elitsa Stoilova is the co-founder and manager of the no-code platform Umni.bg for AI chatbots for websites.
Elitsa has over 20 years of experience in hospitality and tourism, including roles as Marketing Director and General Manager of hotels in Turkey, Tunisia, Morocco, the Mariana Islands in the Pacific Ocean (17 years), and in Bulgaria. She has also spent more than 10 years as a marketing consultant for the destination of the Mariana Islands Tourism Organization (MITO), for which she received two government awards from MITO.
Elitsa’s experience in tourism—an industry highly dependent on timely and high-quality communication between staff and clients—motivates her today to help businesses across various industries implement conversational AI chatbots in areas where human interaction was previously required. She co-founded Umni.bg during her participation in one of Silicon Valley’s largest and most challenging startup accelerators, The Founder Institute, in 2017 in Sofia, with a project for Umni.bg focused on chatbots for the hospitality industry.
Today on the platform Umni.bg Every business can create a digital version of its best employee—a conversational AI chatbot for its website: for customer service, sales, marketing, 24/7 support, data collection, answering frequently asked questions, and automating routine tasks—without any technical skills, as easily as playing with LEGO.
Elitsa is a guest in various media—radio and TV programs, podcasts, and other online publications—as an expert in AI chatbot automation, and a speaker at conferences on topics such as business digitalization, the use of conversational AI for customer communication, marketing, and sales.
