Conversational AI platforms market – According to Gartner, Conversational AI platforms market is extremely varied in terms of vendors’ positioning and enterprise needs.
Selecting the best solution is strategic to succed.
Gartner extimates 2.000 CAIP vendors on the global market that are different in terms of capabilities, supported language and business strategies.
Decision makers need tooling to segment and evaluate the market, in order to take decisions in terms of choices, investments and skills.
Gartner suggest to limit the vendors’ selection to those able to provide the necessary sophistication’s level to succed and to assure return on investments.
Solutions’ sophistication level may vary according to key factors, such as:
- Number and complexity of back-end systems’ integrations, such as authentication platforms, CRM, ticketing systems that may increase complexity;
- Sophistication of customers’ domain model;
- Number and type of communication channels with customers, as well as interaction’s type (for example: developing a chatbot that handles text only is less complex than a chatbot that manages images or vocal interaction);
- Orchestration of multiple chatbots (or virtual assistant) for different use cases can greatly increase the architectural complexity;
- Number of utterances and intents to support and related tasks of generation, tuning and training data management, that increase complexity;
- Number of languages and language variations to be handled, that impact on solutions’ sophstication’s level.
As reported by Gartner:
“Tools assist in mitigating the complexity of designing, developing and maintaining solutions. As solution requirements become more demanding, the importance of strong tools increases”
Gartner
Gartner lists three ways to categorize vendors: the first one is considering who will be in charge of building and maintenance of the solution. The market has three broad categories that most vendors fall into:
1. Application development — Define if CAIP require NLP development skills or make available API for developers.
Vast scale implementations require constant monitoring: dedicating internal developers may be prohibitive in terms of scalability and maintenance.
2. Line of business (LOB) — Verify if CAIP provide low-code/no-code tools for the training and maintenance of intents and dialogues design, that will allow LOB people to do the build and mantain.
3. Managed by vendor — Verify if the building and maintenance of the platform is managed by the vendor. An appropriate positioning would be when the development and maintenance activity of the vendor overlaps with business people autonomy in managing the solution.
The second categorization takes in consideration the purpose for which the CAIP should be used for:
- Cost reduction: aims to reduce the number of allocated resources in particular processes to pursue cost cutting policies;
- Empowering resources: increase the quality of the work, by employing conversational AI to support employees, decision making and compliance;
- Growth and innovation: increase business opportunities through innovation and opening new communication channels (for example, integrating a chatbot to handle social media instant messages like Facebook Messenger, Whatsapp, Telegram etc.)
Finally, the third categorization takes in consideration the intended audience of the implementation.
- Customers— Conversational AI solutions aimed to customers should correctly handle wide variety of languages, short hand terms, different vocabolaries and spelling that customers may use in products or services description.
- Employees—Business partners and other users require specific tasks and languages that will include specific terms, acronims, abbreviations, in a more transactional interaction.
- Consumers— Consumers is usually a synonim of customers, except for the media business for example, where customers are advertising buyers and consumers are media audiences. In this case, typical integrations are with web or mobile applications, speakers etc.
Conversational AI platforms market: gartner Sophistication Continuum
Conversational AI platforms market – Gartner Sophistication Continuum summarizes the evolution process of conversational AI solutions, from chatbot to virtual assistant (or Conversational AI agents as we like to say in crafter.ai):
Depending on interaction complexity, conversational AI solutions evolve along the “sophistication line” : interactions with low complexity where “the user tells the bot what to do” concern Q&A handling, have a limited domain and require simple integrations may be handled by chatbots.
Interactions where “we tell the bot what info we need” refer to complex dialogues, multiple integrations and larger scopes: solutions evolve from chatbots to “virtual assistants”.
In conclusion, evolved virtual assistants handle complex and contextual integrations where the bot can even “anticipate customers’ needs”, going beyond conversations and understanding context in advanced architectures.
Source:
You may like:
https://crafter.ai/en/2021/06/24/conversational-ai-for-business/
