A claim status chatbot that cuts call-centre load for insurers

Updated 2026-09-19 · 5 min read · By the Convo360 team

Why claim status is a call-centre magnet

Nothing generates repeat contact like an open insurance claim. Customers message or call every few days asking for an update, often because they were never told what stage means what, or how long each stage typically takes.

Insurers running large retail health or motor claim volumes out of cities like Pune or Hyderabad can deflect a large share of this by giving customers a way to check status themselves, any time, on WhatsApp.

The anxiety behind repeat contact is usually specific and fixable: a health claim customer worrying whether hospital discharge will be delayed, a motor claim customer wondering if their car will be ready before a planned trip. A status update that answers the underlying worry, not just the technical stage name, reduces the follow-up messages much more than a status label alone.

This matters especially for cashless health claims, where the customer is often still in the hospital and anxious about whether they'll need to arrange payment themselves if approval is delayed. A clear, current status genuinely changes what they do next, not just how they feel about waiting.

The status lookup flow

An Ask a Question block collects the claim number, and an HTTP/API request block queries your claims system for the current stage. A Condition block then picks a clear, specific reply for each stage — submitted, document verification, survey scheduled, under approval, settled, or rejected.

Wording each stage this specifically, rather than a single generic "under process" for everything, is what actually reduces repeat contact: vague status labels are exactly what prompts a customer to call and ask what it really means.

Automatic document reminders

When a claim is stuck at "pending documents", the flow can proactively message the customer with the exact list still needed, rather than waiting for them to ask. This is triggered by the same status lookup — if the API returns "pending docs", the flow sends the checklist automatically.

A useful pattern here is a short escalating cadence rather than one message: an initial checklist the day the claim moves to pending-docs, a reminder after 3 days if nothing arrived, and a note to the servicing officer after 7 days so a human follows up before the claim stalls entirely.

Setting expectations honestly

Each status reply should include a realistic timeframe for the next stage, sourced from your own historical averages: not a promise, just an expectation. This single addition reduces the volume of "any update?" follow-ups significantly, since the customer already knows roughly when to check back.

Where a claim is taking longer than the usual timeframe for a genuine reason, such as a complex case needing additional review or a surveyor's schedule, say so directly rather than repeating the same generic stage message across multiple check-ins. Customers tolerate a longer wait far better when they understand why it's happening than when the same unexplained status repeats every time they check.

For health claims specifically, the policyholder is sometimes not the one actively checking status. A family member handling the hospital admission often needs the same information. Where your policy terms allow, let an authorised family member's number also be registered against the claim for status checks, rather than assuming only the primary policyholder will ever ask.

Building it on Convo360

On Convo360's Flow Builder, this whole flow is drag-and-drop, no code required.

1. Add a keyword trigger for "claim status", "track claim" alongside New Conversation.
2. Use an Ask a Question block for the claim number, then an HTTP/API request block to your claims system.
3. Branch with Condition for each status, sending a tailored message, and include a document checklist reply where relevant.
4. For rejected or disputed claims, route straight to live chat with a human claims officer.
5. Log every status check to your CRM or analytics for volume tracking.

See webhooks docs and live chat docs.

If your claims system doesn't expose a direct API, a simpler version can run off a regularly updated Google Sheets block instead. It is less real-time than a live API, but still far faster for the customer than waiting for a callback, as long as the sheet is refreshed at least once a day.

A mistake to avoid, and what to measure

The single biggest mistake in claim-status bots is a status reply that sounds more certain than it is. "Your claim will be settled tomorrow" is a promise the bot can't actually guarantee; "typically takes 4-5 working days based on similar claims" sets an honest expectation without over-committing the insurer to a specific date it doesn't control.

Track two things monthly: how many status-check conversations end without escalation (deflection), and the average number of times a single claim is checked before it's settled. A claim checked eight or ten times before resolution usually means the stage descriptions or timelines shown aren't reassuring enough, which is worth revisiting even if the underlying claims process itself is running normally.

What it costs

This runs on the Starter plan (₹999/month, ex-GST) for the flow, API lookup and WhatsApp channel. If you also want the bot to answer free-text questions about claim procedure using an AI agent grounded in your policy documents, that needs the Pro plan (₹2,999/month, ex-GST).

Example conversation

Frequently asked questions

Related reading

Available across India — see WhatsApp & AI calling automation by state and city.