Skip to content
India Ecom Website
Choosing the Right Chatbot: Verifast or LimeChat?
Problem Context :
A significant portion of our website visitors come with specific health or wellness concerns but often struggle to find the right product. Our current product discovery relies heavily on manual browsing or keyword-based search, which only matches terms from product descriptions, ingredients, or listed benefits.
However, users typically search or think in natural, conversational language — for example:
Our existing system fails to understand such queries, leading to zero search results, confusion, and higher drop-offs — especially from first-time users unfamiliar with Ayurveda.
To address this, we’re exploring AI-first conversational tools that not only instantly resolve user questions but also guide them to purchase by providing personalized product recommendations, checkout flows within chat, and continuous learning to improve responses over time.
Customer PainPoints :
Lack of Guided Discovery
Users often don’t know what product to look for. They come with specific concerns ("hair fall after pregnancy", "acidity during sleep") but struggle to find relevant solutions due to the absence of a guided, symptom-based journey.
Search Doesn’t Understand Natural Language
Our keyword-based search engine doesn’t recognize real-world phrases like “mujhe gastric issue hai” or “my skin feels dull”. This results in either irrelevant suggestions or zero results, frustrating users.
Lack of Instant Support
Many high-intent users abandon the session when they don’t get quick answers. Waiting for a human response or having to dig through FAQs increases bounce rates.
Language & Accessibility Gaps
A large portion of our audience prefers Hindi or regional languages. The absence of vernacular support creates a disconnect and excludes users who might otherwise convert.
Trust Gap in Choosing the Right Product
Without contextual explanation or personalized recommendations, users may hesitate to buy — fearing side effects, dosage confusion, or product mismatch.
Limechat - Status Quo
First response time, wait time currently for customer is 2.5 Hrs - source Limechat dashboard
March 14 to April 14
Comparison
table
Feature / Criteria
Verifast.ai
LimeChat
Notes
Language Support
Supports English, Hindi, and multiple vernacular languages (Tamil, Malayalam, etc.)
Only English and Hindi
We have 13% of our sessions coming from Maharashtra & Karnataka each, 8% each from TN, Delhi & Kerala
Speed & Responsiveness
Web-based chatbot is fast and responsive
WhatsApp bot observed to be slow (as seen with Kapiva); website bot might be better
Speed is important to cut down drop-offs
AI-First Features
Enables live training of the bot,
We can review chats & have option to give extra context to the bot - so that it gets trained and picks it from next time
Lacks AI-first capabilities — more basic - seems like its not their priority
long term it should keep up with the pace - ai is growing
Conversion Attribution & Analytics
Robust dashboard: Tracks top interactions, query types, traffic source, clicked links, etc.
Dashboard is weak, lacks deep insights or attribution features
will talk about the report in the section below
Maturity & Stability (Based on Our Use)
Currently live, with positive performance data available
Yet to be tested in our ecosystem
More confidence in Verifast
Cost
199$ (17k INR) for 2500 conversation - currently we do approx → 1200 conversations a month
Roughly For similar conversation → per conversation cost is 2.5 Rs → 6250 INR + extra cost for whatsapp consumables
Lead gen
leads are only collected if needed
but we can set up small offers to ask the user share their contact number
every user has to go to whatsapp - becomes a way to get user’s whatsapp
There are no rows in this table
Current metrics
Success Metrics
Metric
Definition
Target
Verifast
Why it Matters
1. Query Understanding Rate (NLP Accuracy)
% of natural language queries accurately interpreted and responded to with relevant product recommendations
>85%
90%+
Measures how well the bot understands free-form queries like “gastric problem after meals” or “hair fall after delivery”
2. Session Engagement Rate
% of sessions where user engages meaningfully (clicks, adds to cart, asks follow-up)
>60%
63%
Indicates quality of interaction and relevance of recommendations
3. Chat-to-Conversion Rate
% of chat sessions that result in a purchase or add to cart
>3–5%
16%
Direct measure of the tool’s commercial impact on sales
4. Live Agent Handoff Rate
% of chats escalated to human support
<5%
1.6%
Measures how independently the bot resolves queries
5. Time to First Response
Avg. time for bot to reply after user message
<2 seconds
approx 3-5 seconds
Faster response = better CX and reduced bounce
There are no rows in this table
Recommendation & Questions in front of us
Based on the current analysis and real-world performance data, Verifast.ai is better aligned with our immediate goals of improving product discovery, enabling natural language interactions, and driving faster, AI-led conversions on our website.
Verifast has demonstrated:
LimeChat brings strength in WhatsApp automation. However, for our business context — where users need in-the-moment guidance during product discovery — verifast holds good.
Should we keep limechat & verifast both in website?
or
we should experiment by keeping only verifast and taking limechat off the website & see how it goes?
time taken to response on Limechat AI vs Verifast AI
25-40 seconds
3-6 seconds
Want to print your doc?
This is not the way.
This is not the way.
Try clicking the ⋯ next to your doc name or using a keyboard shortcut (
CtrlP
) instead.