Hold The Line: AI Customer Support Tools
Generative AI is emerging as a groundbreaking solution to a conundrum at the heart of customer support: the need to simultaneously scale and personalize interactions. Traditional customer service, predominantly a human-centric endeavor, demands extensive training, faces cultural and linguistic barriers, and often grapples with scalability. First-generation automations, such as chatbots and IVRs, were primarily rule-based, offering scalability but lacking the finesse of personalization. Generative AI can bridge this gap and the impact is measurable already.
A recent study by NBER showed that AI implementation can reduce Average Handling Time (AHT) of support tickets significantly, while maintaining NPS scores. JetBlue's implementation of ASAPP’s generative AI solution saved their contact center an average of 280 seconds per chat, resulting in 73,000 hours of agent time saved in just one quarter.
Player Archetypes: Incumbents and established startups have led AI innovation in customer support
Customer support is an extremely competitive market with 50+ established players specializing across customer segments, communication channels and interaction types
- Customer segments: Solutions span from those tailored for individuals or SMBs with a handful of CS agents (e.g., Ada, Ushur) to comprehensive, 360 degree suite solutions designed for enterprise clients with thousands of CS agents (e.g., Salesforce, Zendesk).
- Channels served: Platforms accommodate diverse communication channels, such as chat or SMS (e.g., Gupshup, Twilio), social media interactions (e.g., Yellow.ai, Haptik), and embedded web-based communications (e.g., Intercom).
- Interaction types enabled: Tools may focus on one-way communication (e.g. Hubspot’s automated emails, Twilio’s SMS) or offer two-way interactions (e.g. EliseAI).
Incumbents and unicorns driving AI innovation in this sector
Tech giants have already hopped on the generative AI bandwagon – Salesforce's EinsteinGPT, Freshworks' Freddy, Intercom's Fin, Zendesk AI and Zoho's Zia are notable examples.
Compared to markets like AI Copywriting and AI Video Tools where new AI-native startups led innovation and were able to scale substantially (e.g. Jasper, Synthesia etc.), the AI innovation in customer support is being driven by incumbents and established startups. The set of new players is nascent in terms of adoption and funding. This trend is driven by -
- Distribution & switching costs: Incumbents, often horizontal SaaS providers, have substantial customer bases locked in by high switching costs. This gives them a ready market to develop, sell and improve their AI offerings.
- Limited lower-end use-cases: The customer support market, does not have a use case for individuals or “Prosumers”, unlike AI Copywriting or Video tools. This limits the available market for new tools, which typically build in those lower-end markets first and eventually scale into Enterprise.
- Database advantage: Incumbents, due to their extensive market presence, have amassed substantial data assets. This data is crucial for AI models, which require large datasets for effective training and refinement.
Incumbents focussing on end-to-end product suites, startups building for specific point solutions
Most incumbents in this sector are offering comprehensive, 360-degree solutions that address a wide range of support needs – multichannel conversational support, ticketing and workflow automation, agent support and productivity tools, and customer experience insights. Startups (both new and established) are typically focussed on point solutions such as AI chatbots, response drafting, query summarization etc. and focus on specific channels / modalities.
Market: Bustling space with Conversational AI and Agent support emerging as key use cases
Conversational AI and Agent Assist are key use-cases attracting innovation
Both Conversational AI and Agent Assist tools are serving use-cases that involve direct, customer-facing tasks; hence, generative AI solutions can potentially diminish the dependency on manual effort, improve TAT and enable efficient scaling of customer support operations. Conversational AI, in particular, is relevant across businesses of all sizes and industries, given the need for quality customer support in a digital-first world. Agent support tools are typically more geared towards contact centers and large enterprises.
A bustling market with several startups carving out niches in different use cases, industry-specific solutions
Given the ubiquity of AI-based conversational and agent support tools / features, new startups in the space are trying to differentiate in unique use cases like agent performance management, feedback analysis, intelligent build tools etc. Some startups have also focused on industry-specific customer support tools like MeetElise for real estate, Prodigal and Kasisto for banking, and Yuma for e-commerce, which can be trained on customer support data specific to the respective industry.
Building from low-cost geographies, for the World - is a common theme among startups
Customer service is a price-sensitive function for companies, since operational costs grow with the scale of the organization. Startups located in low-cost geographies like India are well positioned to offer tools at lower price points, while offering competitive features. Several funded startups are originating from India, such as Gupshup, Haptik, Verloop, Kommunicate, Rezo, and Alltius – leveraging price-competitiveness as well as acquired knowledge from historical outsourcing of customer service to India.
Funding Landscape: Majority of recent funding has gone to incumbents, established players
Recent funding dollars have mostly gone towards Incumbents & Established players
~95% of the funding dollars in 2022-23 has been directed towards incumbents and mature startups. Established players are safe bets in this space as explained in the Player Archetypes section - with large, sticky customer bases and substantial data, they are very well positioned to drive value from AI.
There are a large number of well-funded players in the market, making this a very competitive space
New, AI-native startups have attracted smaller, speculative bets
Investors haven’t shied away from demonstrating interest in new, AI-native startups; albeit with smaller, speculative bets. Although new startups represent only ~3% of total funding for 2022-23, they comprise >40% of the funding deals made during that period.
Website Visits: Signs of initial AI hype settling down; Sapling, Verloop seeing considerable traction
Market going through a “trough of disillusionment” as initial AI hype settles down
AI customer support tools, like other AI markets, saw strong interest during the peak of the AI hype in Q4’22 and Q1’23. However, about ~70% of startups that experienced growth in Q4’22 and Q1’23 have reported de-growth in Q2’23. This trend is common in emerging tech markets, where the initial interest surges, but settles down into a more realistic level as the realities of implementation and market adoption come into focus. ChatGPT experienced a similar pattern, noting around a 10% decline in global traffic for two successive months in Q2’23. Going forward, we expect steadily growing interest in tools that can prove successful value creation for customers, while less differentiated tools will struggle - in other words, we expect customer attention to be driven by results vs by hype going forward.
Sapling, Verloop and MonkeyLearn showing strong traction
Sapling.ai recorded impressive consumer interest at >4M visits in Q2’23. Sapling’s AI-copilot comes as a web extension and also integrates with CRMs or traditional communication platforms such as Gmail or Slack to optimize agent communication by retrieving relevant responses, rephrasing and enhancing language. The platform offers a comprehensive snippet library and actionable analytics from customer conversations.
Verloop is an India-focused agent-assist platform that offers 24/7, multilingual AI-powered support across chats, calls, and soon, emails. The platform has been extremely focussed on certain verticals, including ecommerce, banking and FSI. Verloop has worked with brands like WhatsApp Business, Nykaa (which saw a 2.2x rise in engagement post-automation with Verloop), and Ninjacart (addressing 689% more queries post-automation).
MonkeyLearn – used by computer tech giant Dell to save 400+ employee hours on feedback analysis – offers a no-code AI toolkit for customer support teams. It simplifies AI adoption, enabling teams to automate support tickets, analyze customer feedback, gauge sentiment and NPS, and visualize the results graphically.
Product Reviews: Boost.ai and Kommunicate have built strong customer franchisees
Boost.ai has earned acclaim for its advanced conversational AI capabilities – it was recognized in Gartner® Magic Quadrant™ for Enterprise Conversational AI Platforms for two consecutive years (2022 & 2023). Boost.ai deploys advanced virtual agents tailored to manage thousands of industry-specific intents with a 90% resolution rate. Their conversational AI adapts swiftly to customer feedback, supported by a user-friendly no-code builder and AI trainer, empowering enterprises to scale and refine customer service effortlessly.
Kommunicate – the chatbot that resolved 75% of GAP’s customer interactions in Chile – stands out for its blend of bot and human-driven interactions. Users laud its dashboard for clarity and the software's knack for streamlining communication.
- Highly competitive market with established players best-placed to deploy AI solutions at scale
- Transition away from human agents could be only a few years away
- Despite established players’ dominance, some great new startups emerging
- Large deals and M&A activity will be prominent in the next 2 years
Large incumbents and unicorns have been quick to adopt AI and results are showing (e.g. JetBlue’s implementation of ASAPP’s AI solution). Expect established players to compete for large AI-focused transformations as more and more enterprises overhaul their customer support teams.
The nature of customer support is on the precipice of change - while current solutions have shown strong value in improving agent productivity, there is a lot of innovation in fully self-serve solutions. The role of support agents is likely to shift towards more strategic activities like relationship building, handling the largest clients etc.
While AI innovation has predominantly been steered by industry giants, new startups such as Sapling, Kommunicate, and Boost.ai are gaining ground. Established players like Yellow.ai and Verloop are evolving from point solutions to comprehensive product suites.
The market is primed for M&A activity with multiple large well-funded players, fast-paced innovation and interesting use cases emerging. We have already seen some M&A - Ramp’s acquisition of Cohere, Dstny’s acquisition of Tactful are examples.