Every operator I talk to is asking the same question: where does AI actually move the numbers in a practice? The honest answer is that it moves them in the same places where humans were already losing them. The clearest example I have is from the consultation funnel, which is where most aesthetic businesses leak more revenue than they realize.
I built a workflow for this before there was a product category for it. Tools like Opaline now do a version of what I was solving by hand. The principle has not changed. The execution has only gotten easier.
The consult is won before the patient walks in. Every other lever in the funnel is downstream of how prepared both sides are when they meet.
The bottleneck was never the consult
Most practices treat the consultation as the conversion event. It is not. By the time a patient walks into a consult room, the decision-shaping has already happened. They have either arrived prepared, with a clear sense of what they want and what they can afford, or they have arrived cold, expecting the provider to do all the discovery work in twenty minutes.
When a patient arrives cold, three things go wrong at once. The provider spends most of the appointment on history-taking instead of treatment planning. The patient feels rushed when pricing comes up at the end. And the conversion rate from consult to booked treatment falls, because there was no time to build a real plan.
The leak is in the prep, not the room. That is the part I set out to fix.
The build: a four-part workflow
I designed the workflow around one operating principle: every lead arrives with information the provider needs, and every provider has a response style that can be modeled. If I could capture the first and codify the second, the AI could do the connective work in the middle.
The workflow had four pieces, in order:
1. A structured pre-consultation intake form
Every inbound lead, whether from paid ads, Instagram DMs, or the website, was routed to a single intake form before any provider response. The form was designed not as a data-collection exercise but as a qualification gate. Patients who completed it were demonstrating intent. Patients who did not were exactly the leads that would have churned anyway.
The form captured:
- Name, preferred contact method (text or email), and best time to reach
- Budget tier, with examples of what each tier typically maps to so the patient could self-orient
- Upcoming events: weddings, anniversaries, reunions, professional milestones
- Treatment history at any med spa or aesthetic practice
- Booking timeline: ready now, looking in the next month, browsing for later
- Front and profile photos in natural lighting
- How they wanted to feel when they looked in the mirror or saw a photo of themselves
- Any bad experiences with treatments or med spas in the past
- Recovery window available, framed as: I have a quiet two weeks coming up versus no downtime ever
- Preferred consultation format: virtual, in-person, or phone
The two open-text questions did more work than the rest of the form combined. They surfaced sophistication, raised flags about prior bad experiences, and gave the provider a real human entry point before the first message went out.
The budget tier section was structured so that patients could self-orient without feeling priced out. Each tier was paired with the kinds of treatments it usually covered. That single design choice reduced the number of patients who selected a budget that did not match their actual goals.
The two open-text questions did more work than the rest of the form combined. They surfaced sophistication, raised flags about prior bad experiences, and gave the provider a real human entry point before the first message went out.
The budget tier section was structured so that patients could self-orient without feeling priced out. Each tier was paired with the kinds of treatments it usually covered. That single design choice reduced the number of patients who selected a budget that did not match their actual goals.
Budget tier
Typical outcomes
Under $500
Regimen refresh, glow, hydration, gentle resurfacing
$500 – $1,500
Smoothing, subtle enhancement, brightening, refined definition
$1,500 – $3,500
Texture refinement, tone correction, mild tightening, restored radiance
$3,500 – $7,500
Facial balancing, contouring, deeper resurfacing, structural support
$7,500+
Comprehensive treatment custom plans, full-face transformation, body contouring
2. AI triage by intent and budget band
Once the form was submitted, an AI agent processed the response and assigned each lead a triage label. The labels were simple and operator-grade: ready to book within two weeks, researching within four to six weeks, browsing without a timeline, or flagged for clinical review.
The triage logic combined three signals: stated booking timeline, budget tier, and whether there was an upcoming event. A patient with a wedding in five weeks and a $3,500 budget did not get sorted into the same queue as someone browsing skincare options without a timeline. The provider’s hour was concentrated where it converted.
3. A pre-drafted clinical reply for every lead
This is the piece most practices try to skip and then wonder why their conversion never moves. Every lead, regardless of triage tier, received a pre-drafted response from the AI agent. The drafts were modeled on templates the providers had previously approved, including their voice, their treatment philosophy, and their standard language for pricing transparency.
Drafts included specific recommendations on treatment approach, timing relative to the patient’s event window, and skincare prep. They never included a final clinical recommendation without provider review. The AI was doing the writing. The provider was doing the medicine.
4. Provider review, batched once a week
Before this workflow, the providers were reviewing inbound consults daily — often in the middle of clinical hours, often on their phones between patients. The new cadence was a single one-hour review block once a week. The provider opened the queue, reviewed the AI-drafted replies in priority order, made edits where needed, and approved the batch to send.
Anything flagged for clinical review (medical history concerns, contraindications, mismatched expectations) was pulled out of the batch and handled separately. The batched cadence preserved provider focus during clinical hours and made the lead response feel less reactive and more deliberate.
If a patient did not book after the first reply, an email drip sequence triggered automatically — also pre-drafted, also reviewed, also branded in the provider’s voice. The drip ran for four weeks before the lead moved to long-term nurture.
What it produced
The workflow ran for several months before I began tracking the lift in a structured way. The numbers below are from a single practice implementation, measured against the prior twelve months of the same funnel before the workflow was in place.
- Lead-to-appointment conversion: moved from approximately 35% to 80%. The single biggest driver was speed: every lead received a thoughtful, personalized response within a defined window, regardless of when they submitted the form.
- Consultation-to-treatment conversion: lifted meaningfully, with a notable shift in patient behavior — more patients moved directly from the virtual consultation form to a booked treatment, skipping the in-person consult step entirely.
- Lifetime value per patient: moved from a baseline of $800–$1,200 to approximately $3,500. The increase came from better-matched first treatments, more accurate pricing expectations on the first visit, and a stronger cross-sell into skincare and maintenance.
- Provider triage time: shifted from daily ad-hoc review to a single one-hour batch session per week. That time went back into clinical hours.
- Added monthly revenue: approximately $40,000 to $60,000, attributable to the combination of higher conversion, higher first-visit ticket, and recovery of leads that would have otherwise gone cold.
One workflow. Four KPIs moving. The leverage was not in the AI itself. It was in the structure the AI made possible.
The guardrails that kept it clinical
AI in a clinical context only works if the guardrails are non-negotiable. The workflow held to four:
- Clinical review on every outbound reply. No AI-drafted message was sent without provider sign-off. The AI compressed the time required to review. It never replaced the review itself.
- Escalation rules for medical complexity. Any intake response that surfaced a medical history concern, contraindication, or expectation mismatch was pulled from the standard batch and handled in a separate workflow.
- No AI handling of dissatisfied patients. If a patient ever expressed dissatisfaction or concern post-treatment, the AI was completely out of the loop. Those responses were always human.
- Voice fidelity, not voice substitution. The AI was trained on the provider’s actual templates and language patterns. The patient was reading the provider’s voice with AI’s speed. Not a generic chatbot pretending to be a clinician.
Where the category is now
When I built this workflow, there were no purpose-built tools for it in the aesthetic space. I was working with a structured form, an AI assistant, and a set of provider-approved templates. Today the category has caught up. Tools like Opaline now do a version of this work — AI-powered intake quizzes, intent-driven lead triage, and downstream treatment recommendations, all designed for aesthetic practices specifically.
I will be writing a deeper breakdown of the Opaline platform in a separate post. For now, the relevant point is that the workflow logic is the same whether you build it or buy it. Personalize before the consult. Triage by intent. Keep clinical review intact. The tools just got better.
The principle
If your conversion from lead to appointment is sitting in the 30% range, the leak is almost never the consult itself. It is the cold start every lead is forced to make. Fix the prep on both sides — the patient arrives ready, the provider arrives informed — and the rest of the funnel responds.
Whether you build the workflow yourself or implement a purpose-built platform, the operating principle is the same. The consult is won before the patient walks in. Build the prep. Protect the conversion.
• • •
Audrey Campbell, MPH is the founder of The Audrey Aesthetic, a thought leadership platform on AI implementation, practice operations, and scaling strategy for the medical aesthetics industry. Subscribe to the Substack at theaudreyaesthetic.substack.com.