Market Guide for Conversational Platforms
Gartner predicts that by 2021, 15% of all customer service interactions will be completely handled by AI, an increase of 400% from 2017
Gartner’s report “Market Guide to Conversational Platforms” aims to help enterprises to understand the emerging market of conversational platforms. So we wanted to summarize the article to show the tools to build multiple chatbots and assistants and the dynamics of the growing market by a brief reading.
The main takeaways:
- According to a recent Gartner survey, 4% of enterprises have deployed conversational interfaces, but 38% planning to or actively experimenting
- It is a highly crowded and fragmented market — not easy to understand the combination
- Every type of vendor is in an evolving phase and has its own strengths and weaknesses
- In the future, it is expected that the market will be split into two parts:
- conversational platforms built around strong proprietary NLP engines
- conversational middleware focusing on orchestration and product lifecycle
The recommendations of Gartner for the application leaders:
- Select your vendor tactically: The vendor selection for conversational platforms is currently tactical — selecting a vendor based on the current and immediate needs, not a prediction of your future needs is the best way
- Put in place an exit strategy: In the very near future, the vendor landscape will change drastically. So make it easy to switch vendors by avoiding lock-in of training data, responses and integrations
- Focus on the middleware functionalities: Understand the functionality of analytics, scalability and orchestration, in order to ensure that you can keep improving the implementation
The strategic planning assumption that Garner uses is that “by 2021, 15% of all customer service interactions will be completely handled by AI, an increase of 400% from 2017 (Gartner Predicts 2018: CRM Customer Service and Customer Engagement)
What is a conversational platform?
These are platforms that can be used by developers to build conversational user interfaces, chatbots and virtual assistants for a variety of use cases.
- They offer integration into chat interfaces such as messaging platforms, social media, SMS, website, chat etc.
- A conversational platform has a developer API, so that third parties can extend the platform with their own customizations and additions
What are the popular use cases of conversational platforms?
- Virtual customer assistants — Used for the automation of call-centre-operations; chat or voice-based communication; commonly includes escalation mechanisms for humans to take over
- Virtual enterprise assistants — Conversational interfaces for employees to simplify their access and engagement with the enterprise and its systems
- Virtual personal assistants — Generalist assistants for users that broker first-, second- or third-party services and knowledge; commonly deployed on consumer or dedicated devices
- Chatbots — Narrowly focused chat agents; typically for a specific task
Which direction does the market go?
It is a fast evolving and confusing market — consisting of a large number of vendors that focus on any one of four different use cases mentioned above.
According to the 2018 Gartner CIO Survey;
- 4% of enterprises have currently deployed a conversational interface, such as a chatbot or virtual assistant
- 17% are actively experimenting or planning to do so in the short term
- 21% have it in their medium or long-term planning
A tremendous growth is expected for the next years as projects will evolve from POC/plan to implementation. To see large negative media coverage, failed projects and abandonment of the technology is an expected case because the technologies involved in the conversational platforms enter the “Trough of Disillusionment” on the Hype Cycle.
The main evolutions that will occur in the market:
- Vendors with simple platforms may be active in the geographies with less-widely used languages
- Many vendors with simpler platforms will not be able to keep up with the demand for capabilities. The majority of future acquisitions are likely to be by enterprise software vendors. It is highly likely that ERP, CRM, business process management (BPM) and robotic process automation (RPA) vendors will look to make acquisitions to gain conversational capabilities
Informational vs Transactional
- As opposed to the generally-known, informational use cases need a large number of intents
- Transactional use cases, however, usually rely on fewer intents. The conversations can be longer and need to be more flexible, with multiple paths to the endpoint. Contextual information from user profiles or past conversations can greatly enhance transactional conversations
- Platforms with capabilities of both informational and transactional use cases are likely to be preferred
Starting With Self-Service
- There is a clear trend toward self-service becoming more common. Vendors offering a self-service sign-up process that focusses on onboarding of developers and business users have an advantage
Evolving Toward Middleware Capabilities
- There is an increased focus on the analytics, maintenance and operation of chatbots and assistants, as well as the orchestration of multiple chatbots
- We see the support for one master assistant routing requests to several back-end chatbots covering different domains. It requires a middleware layer to manage the maintaining training data and dialogues
- This move toward middleware capabilities might split the market in two (1) conversational platforms built around a core NLP engine (2) conversational middleware built around orchestration and lifecycle management
Accelerators and Vertical Offerings
- There is an increasing focus on vertical offerings — ready-made intents and marketplaces where intents can be traded. Banking, telecom, insurance, healthcare, retail, human resources and help desk seem to be the top domains for these intents
- Many of the vendors already have a voice as a capability. Voice support is, however, is not as sufficient as text-based chatbots
- But voice will be a necessary capability for future conversational platforms because the potential savings of automating voice communication are greater than text-based chat
- Platform vs product: Most of the vendors in the market started out creating products to solve a specific use case, and then evolved these into a platform. Application leaders should be aware of the legacy of their chosen platform — if it is started as a platform or started a product and evolve to be a platform
- Cloud vs. on-premises: It is a cloud-dominated market with several vendors offering on-premises or hybrid installations. A cloud-only strategy can limit a vendor’s ability to grow in industries such as finance, healthcare and government, where security and privacy compliance is regulated
- Pricing model: The most popular pricing model is the intent resolution-based model. Other models include charging flat license fees, which is most common on on-premises installations. There is also some difficulty in the comparison of pricing models. As the market matures, the pricing models are expected to become more transparent and to normalize. For now, vendors offering multiple pricing models have the advantage.
Read the full article: https://www.gartner.com/doc/3879492?ref=mrktg-srch