Analysis Of The Aragon Conversational AI Report
And how the Aragon report compares to the Gartner Magic Quadrant…
◾️ The Aragon Report covers 20 technology providers.
◾️ Gartner covers 21 technologies.
◾️ Aragon mixes chatbot capability with voicebot capability.
◾️I really hope Gartner brings out separate reports for voicebots and chatbots in 2023, with focus on voicebot enablers (ASR/Synthesis).
◾️ In the image below is the Gartner Magic Quadrant (left) and the Aragon Research Globe (right).
◾️ Aragon rightly observes that the market is flooded with technology providers.
◾️ Aragon Research claims to be tracking > 300 providers in conversational AI.
◾ Below, The technology providers marked in a red block on each report are excluded in the other.
◾️ The Gartner report seems more balanced and the only critique I have on the Gartner report is the exclusion of Microsoft, Nuance (Mix) and NVIDIA.
◾️ The Gartner report seems more balanced and representative, with 28% of technology providers in the leading quadrant.
◾️ The Aragon report is very unbalanced with 55% of technology providers in the leader quadrant which really devalues the report, or points to weak judging criteria.
◾️The maturity model provided in the report is shown below, with 2022 marked in red by me. The model does seem premature with unsupervised ML and unlabelled data seeming far off at this stage.
◾️The Aragon report does allude to the growing importance of voicebots, but this subject needs more discussion and detail.
◾️ The frameworks mentioned but which were excluded, include:
- Aivo
- Boost.ai
- Drift
- Engati
- Flow XO
- Haptik
- Inbenta
- LiveChat
- Pypestream
- SAP Conversational AI
◾️ The Aragon inclusion criteria included chatbots, voicebots, or both.
- A minimum of $5 million in primary revenue for conversational AI solutions.
- Or a minimum of $10 million in revenue in a related market (team collaboration / messaging, RPA, or other related software market).
◾️ The Aragon research document on Conversational AI for 2022 can be accessed here.
Conclusion
Comparing Conversational AI platforms in a like-for-like approach is becoming unfeasible, unpractical and unfair.
- Solutions should be tiered, very much like in the case of the Gartner Peer Reviews.
- Solutions should be compared instead of companies or complete offerings.
- Segmentation is required according to voicebots, chatbots and enabling technologies like ASR, Speech Synthesis, etc.
- Additional criteria can be Conversation Design, NLU Design, LLMs, Semantic Search, etc.
I’m currently the Chief Evangelist @ HumanFirst. I explore and write about all things at the intersection of AI and language; ranging from LLMs, Chatbots, Voicebots, Development Frameworks, Data-Centric latent spaces and more.