There has been so much talk about chatbots lately as it has become obvious – they are dramatically changing the business world. Just think about how greatly the ways of communication between businesses and customers have changed over the past few years!
Five years ago people mostly communicated by email and – from time to time – by phone. Now we are more likely to use Facebook Messenger or WhatsApp. In the near future, even this would undergo drastic changes due to the increased popularity of chatbots. Soon it’ll be enough for customers just to open a messenger to enjoy a completely new channel of instant communication with your company. Chatbots have the potential to revolutionize presale processes and customer service through giving instant answers anytime they want.
More and more companies become interested in building their own chatbots. 80% of the respondents (chief strategy officers, senior marketers and sales executives) said they already used chatbots or planned to use them by 2020 (according to a survey by Oracle).
Determined to implement a chatbot into the work of their company, a person in charge immediately faces the following questions:
- What needs to be done prior to building a chatbot?
- How does the chatbot development process look like?
- How much would it all cost?
In this article, we’ll disclose the price on chatbots while looking into the phases of the development process. We’ll reveal cost-effective factors and name the approximate costs for each phase, as well as for the whole development.
It’s needless to say, that the main factor influencing the project price is the work being done to develop a product. The development work is usually split into several phases. Having a general idea of the project phases may help you make a rough estimation of cost. The following roadmap has been found suitable for developing chatbots, and is adopted by Azati.
1. Discovery Phase
A Discovery Phase is the first move on the way to success.
As a first step, we define the project purpose and value proposition; set goals, expectations and metrics. Having a proper understanding of what you want a chatbot to do and the ways it can improve your business is the very first thing to do. Afterwards, these findings would play an important role when defining the technologies to use. Next questions are a good starting point:
- What is the size of your business – small, medium, enterprise?
- Do you want a chatbot to entertain the clients or to provide a full customer support with reduced human involvement?
- Which of the repeatable processes can be automated with a chatbot in your case (presale customer communication, customer support or any other)?
- How many messaging platforms are you going to communicate through? What are they?
- How many users do you expect your chatbot to have? What is the expected conversational flow?
Although the questions are fairly simple, this step is crucial and in-depth analysis is highly required since the outcomes of this phase determine the following product development processes.
This part of the discovery phase is about learning what business has, wants and needs. The effort here goes a long way to having everyone happy in the end.
The next move during the discovery phase is to identify the requirements that cover the functional units of a chatbot, such as:
- Conversational Flow
- Control Panel
- Interaction Channels
- Natural Language Processing
Can’t but mention that you also need to decide on how the chatbot will be build. There are two major options in front of you:
– Building a custom solution
It’s great to get a chatbot that is built according to your needs, preferences and plans to scale. This way you are not limited to any platform abilities; although, the support and maintenance of such a bot becomes your responsibility.
– Building a bot on top of a platform
As a rule, creating a chatbot using an existing platform is a more cost-efficient option. Such services work on a subscription basis and charge monthly payments (sometimes basic functionality is provided for free). However, there are always fees for customization and oftentimes the sum you pay sooner or later equals to the sum you’d pay for a custom solution.
All in all, this option means that:
– a chatbot is build a bit faster (as basic functionality is already offered by a platform);
– you get support and maintenance provided by the platform;
– you can still get a customized chatbot.
There is no right choice here. The complexity of the future chatbot doesn’t matter when choosing between the options – any of them might be suitable.
At last, the project managers plan how to manage time, energy and budget, and reflect it all in the corresponding documentation.
At the end of the phase, the parties define the scope of work needed for the next move (which is the prototype development). If all the required data is accessible, the phase takes 5-7 working days on average.
At Azati, we usually go through the discovery phase for free.
2. Prototype Implementation and Evaluation Phase
A prototype is an early sample that does not have all the functionality of the final product. It can be be code-based or simply in a form of a mockup (the choice depends on the chatbot complexity). A prototype helps to learn many essential things. It is built with the intent to verify some key functional aspects of the intended design, in other words, to conduct a feasibility study.
When the prototype is build, it is then discussed with the client. The client and the developers are now able to validate the requirements and design choices, and make adjustments, if needed. The risks and costs associated with software implementation are dramatically reduced, as the requirements are well-discussed early on, prior to development.
We are striving to make this phase as inexpensive as possible. Typically, the prototype costs are about $3,000 – $5,000, including the next phase estimation.
3. Minimum Viable Product
Bearing the findings from the prototype phase in mind, we develop a minimum viable product (MVP). An MVP is a product with basic functional features, a minimal first version of the ultimate product solution that already works with actual data. An MVP is exposed to a small group of real end users in order to gather the vital feedback. It is way less expensive to modify the system at this stage, than when it is fully developed.
The cost for an MVP ranges between $10,000 – $20,000, depending on the total project size and complexity.
4. Chatbot Release
As the result of this phase, the product with a complete set of predefined features is developed and then launched onto the market. During the preceding phases, we elicit and validate the exact requirements, therefore the end product is made with minimal risks.
The cost of the release phase is usually estimated during previous phases.
Chatbot Continuous Training
Just as an employee’s effectiveness depends on their experience, the effectiveness of a chatbot depends on its knowledge base and training. The AI-powered chatbots are able to continuously learn from the data submitted by end users. The process of ongoing training never ends, and much attention should be paid to it. With such an outstanding ability, you’ll witness your chatbot grow more intelligent, able to answer more detailed queries and communicate in a more human-like manner, raising the level of customer satisfaction.
So, how much in numbers?
The common misconception about AI-driven chatbots is that they must cost a fortune, but it’s not true nowadays. Thanks to the emergence of various tools, libraries, frameworks for building chatbots, the technology is becoming affordable for many companies.
What concerns costs – prices are calculated for each case in particular. Most often software companies charge from $15,000 to $30,000 for a custom bot.
The companies located in Silicon Valley set a minimum price of $30,000 for a pretty simple bot that is able to reply to various users’ questions.
The price for a very sophisticated AI-chatbot for customer support starts from $40,000 and can be as high as $100,000. Such bots are built on Natural Language Processing and Machine Learning technologies and are endowed with more powerful capabilities of learning and absorbing information. They can target customers through marketing campaigns and media, handle personalized news and content distribution and more. If combined with payment services, such chatbots become complete business-enabling solutions.
One may start with a simple chatbot that automates most frequent flows of your company and add AI-based bot later on, as soon as the first one shows its efficiency.
Hardly anyone would refuse to recognize the chatbots’ great capabilities in identifying and satisfying the users’ needs. Gone are the days when your customers had to call during office hours. Luckily, today many businesses can enable their clients to appreciate the easiness of a self-serving tool, that would inform them about your products and services 24/7/365.
Do you need any help with a chatbot? Call us at +1 (973) 597-1000 or fill out the form below for a free consultation.