
Intelligent Consult Intake and Routing for a Body-Contouring Clinic Chain
Let's Connect
Overview
What we built
A 20-clinic body-contouring chain let hot advertising leads sit for hours in one call centre's queue. We built a conversational intake layer that qualifies and routes them the moment they arrive.
In plain terms: every enquiry generated by the chain's advertising, across 20 clinics, funnelled into a single call centre for manual triage. New leads routinely waited 6 hours or more before anyone spoke to them, agents re-asked the same qualifying questions on every call regardless of what a prospect had already said, and enquiries were regularly routed to the wrong clinic. By the time a lead reached the right person, many prospects had already gone cold and moved on.
We built a conversational intake layer across web chat and SMS that gathers treatment interest, basic eligibility and preferred location the moment an enquiry arrives. An NLP classifier scores each conversation's intent and urgency, high-intent leads route straight into the nearest clinic's consult calendar with automated reminders, and lower-intent enquiries enter nurture sequences instead of the call centre queue. Median first response fell from over 6 hours to under 3 minutes, and the lead-to-booked-consultation rate climbed from 22% to 34% within the first quarter.
The Problem
Hot leads left waiting
The chain generated a high volume of advertising-driven consultation enquiries, and every single one, across all 20 clinics, funnelled into one call centre for manual triage. That worked when volume was modest, but the chain's own advertising success became the very thing straining the system: more leads arriving meant longer queues, and every lead in that queue was a prospect actively comparing options elsewhere.
Once a call finally connected, prospects endured questions the chain already had answers to. Agents asked the same qualifying questions on every call, treatment interest, basic eligibility, preferred location, regardless of what the advertising campaign or intake form had already captured, adding friction to conversations that should have moved straight to booking.
Routing was its own failure point. Enquiries were regularly sent to the wrong clinic, meaning a prospect who had already waited hours for a response then had to be transferred, re-explain themselves, or wait again while staff sorted out where the appointment actually belonged, by which point many had gone cold.
Single-queue bottleneck
Every advertising-driven enquiry across all 20 clinics funnelled into one call centre for manual triage, so the chain's own marketing success created its own backlog.
Hours before contact
New leads routinely waited 6 hours or more for a first response, plenty of time for an interested prospect to book with a competitor instead.
Repetitive qualifying calls
Agents asked the same qualifying questions on every call, regardless of what a prospect had already shared, adding friction before any real conversation began.
Misrouted enquiries
Enquiries were regularly routed to the wrong clinic, forcing prospects to wait again while staff worked out where the appointment actually belonged.
What it was costing them
Every enquiry stuck in the single-queue bottleneck was a prospect the chain had already paid to attract through advertising. With new leads waiting 6 hours or more for a first response, agents re-asking the same qualifying questions on every call, and enquiries regularly landing at the wrong clinic, a meaningful share of that paid-for interest cooled and looked elsewhere before anyone at the chain even answered the phone.
The Solution
Conversational intake with smart routing
We built a conversational intake layer across web chat and SMS, meeting prospects on whichever channel they had already used to reach out. The moment an enquiry arrives, it gathers treatment interest, basic eligibility and preferred location, the same details agents had been re-asking on every call, so nothing needs repeating once a human does get involved.
An NLP classifier scores every conversation for intent and urgency as it happens, distinguishing a prospect ready to book from one still comparing options. High-intent leads route straight into the nearest clinic's consult calendar, with automated reminders following automatically, removing the manual triage step that had been sending enquiries to the wrong clinic in the first place.
Lower-intent enquiries are not abandoned, they enter nurture sequences instead of competing for call-centre attention with prospects who are ready now. Every interaction, from first message to booked consultation, syncs to the chain's CRM, so agents inherit full context rather than starting each qualifying conversation from zero.
Key decisions
Capture interest immediately
Treatment interest, basic eligibility and preferred location are gathered the moment an enquiry arrives, so nothing waits for a human to ask the same questions again.
Score intent before routing
The NLP classifier scores each conversation's intent and urgency first, so routing decisions reflect how ready a prospect actually is rather than arrival order.
Route to the nearest clinic
High-intent leads go straight into the nearest clinic's own consult calendar, replacing the manual triage step that had been sending enquiries to the wrong location.
Automate the reminders
Booked consultations carry automated reminder sequences by default, addressing no-shows without adding another manual task for call-centre agents.
Nurture instead of queue
Lower-intent enquiries move into nurture sequences rather than the call-centre queue, freeing agents to focus on complex enquiries that need judgement.
Measurable Impact
What changed after launch
The queue that used to swallow every enquiry no longer does. Median first response fell from over 6 hours to under 3 minutes, and the lead-to-booked-consultation rate climbed from 22% to 34% within the first quarter, evidence that prospects who used to go cold are now booking while their interest is still fresh.
The automated reminder sequences pulled the no-show rate down from 28% to 18%, and call-centre manual triage workload fell by roughly 45%, freeing agents to focus on the complex enquiries that genuinely need a person rather than repeating the same qualifying questions on every call.
First response
Over 6 hours to reach a new lead
Under 3 minutes from enquiry to response
Booking rate
22% of leads converted to consultations
34% converted within the first quarter
No-shows
28% consultation no-show rate
18% with automated SMS reminders
Call-centre workload
Manual triage on every single enquiry
Workload down roughly 45%, complex calls only
Headline results
Median first response to new enquiries cut from over 6 hours to under 3 minutes
Lead-to-booked-consultation rate improved from 22% to 34% within the first quarter
Consultation no-show rate down from 28% to 18% with automated SMS reminder sequences
Call-centre manual triage workload reduced by roughly 45%, refocusing agents on complex enquiries
Tech & Tools Used
What powered the build
Every tool below earned its place in this engagement. Here is the part each one played.
Node.js
Runs the real-time conversation service behind the web chat and SMS intake, coordinating the classifier, routing logic and calendar writes as each enquiry arrives.
Python (FastAPI)
Serves the intent and urgency scoring as an internal API that the conversation service calls the moment treatment interest and eligibility are captured.
Hugging Face Transformers
Provides the models behind the NLP classifier, trained to distinguish high-intent prospects ready to book from lower-intent enquiries still comparing options.
Twilio SMS
Carries the SMS side of the intake layer and delivers the automated consult reminders that helped bring the no-show rate down.
PostgreSQL
Stores every enquiry, its scored intent and its routing outcome, giving the chain one consistent record across web chat and SMS.
Redis
Holds live conversation state so a prospect can move between intake questions without losing their place, on either channel.
React
Builds the web chat interface where prospects share treatment interest, eligibility and preferred location before ever speaking to an agent.
HubSpot CRM API
Syncs every interaction, from first message to booked consultation, into the chain's CRM so agents pick up with full context.
AWS Lambda
Executes the event-driven steps in the routing pipeline, including scheduling the automated reminder sequences once a consultation is booked.
Ready to Build your Medical Aesthetics Business with Intelligent Intake & Routing
Ask Byte
Ask Byte
Typically replies instantly
just Now
Hi! I'm OrganByte's assistant. How can I help you today?
AI-generated content may be incorrect

