Voice AI platform DialNexa has released findings from an analysis of more than one million AI-assisted business calls in India, highlighting how response speed, language handling, retry strategies and conversation design can influence the performance of automated calling systems.
The study, titled The State of AI Voice Calling in India, is based on live calls handled through DialNexa for its customers. Around 84% of the calls analysed were outbound and 16% were inbound. The dataset covered English, Hindi and Hinglish conversations, with pre-sales lead qualification and webinar or event attendance among the main use cases.
According to the company, new outbound calling numbers recorded a first-attempt pickup rate of close to 48% across their first 1,000 leads. Pickup rates declined as numbers were used more extensively, but structured retry sequences helped some campaigns achieve cumulative connectivity of more than 70%.
DialNexa's analysis also found that fewer than 3% of conversations ended because the customer disconnected first. The company said factors including natural-sounding voices, the opening of the conversation, response speed and transparency about the use of AI influenced whether users remained engaged.
Latency emerged as another factor affecting conversations. Median AI response latency remained below one second, while the 95th percentile was approximately 2.1 seconds. The company said longer pauses could interrupt the natural flow of a phone conversation even when average response times remained low.
Language behaviour was particularly relevant in the Indian market. The dataset found frequent switching between Hindi and English, sometimes within the same sentence, highlighting the importance of supporting Hinglish and code-switching rather than evaluating voice systems only on English-language performance.
Pre-sales lead qualification emerged as the largest and strongest-performing use case in DialNexa's dataset, followed by webinar and event attendance. EdTech and BFSI were the leading sectors represented in the analysis.
Inbound calls showed different performance characteristics. They accounted for around 16% of the dataset but recorded an 89% goal completion rate, according to DialNexa. Inbound conversations averaged 13 minutes, while the user-initiated disconnect rate was reported at 0.01%.
The study also identified 10 am to noon, 4 pm to 6 pm and 8 pm to 9 pm as the periods generating the highest connectivity for campaigns targeting working professionals.
The findings are based solely on calls handled through DialNexa's platform and should not be treated as an industry-wide benchmark. The company did not disclose the number of businesses represented or detailed performance breakdowns by region and industry. Its analysis nevertheless offers a snapshot of how voice AI is being used in live business interactions in India.
Disclaimer: This article may include information derived from interviews, press releases, public statements, research, company communications and other publicly available or third-party sources. Such material may be summarised, paraphrased or contextualised for journalistic and editorial purposes. All rights in third-party content remain with their respective owners.