Which Channel Actually Brought the Patient? Inside BELSA's Attribution Fix

Ask a founder which marketing channel is working and they will answer in under three seconds. Ask them to prove it, and the confidence disappears. BELSA Estétic ran on exactly that gap for years — until a line on an intake form made it impossible to ignore.

The question nobody could answer honestly

Before RIVEL, BELSA Estétic — the aesthetic medicine clinic in Barcelona led by Consolación Sánchez — ran patient relationships the way most boutique clinics do: conversations living in WhatsApp threads, the schedule in one shared spreadsheet, and a steady stream of new enquiries arriving from Instagram, Google and word of mouth. The clinic wasn't failing. Revenue was steady. Patients loved the team. Nothing was broken in a way anyone could point at.

But ask the founder which channel actually produced a paying patient, and the honest answer was a shrug dressed up as an opinion. "Instagram is doing well lately." "I think the Google reviews are bringing people in." Nobody was lying. Nobody had the data to lie about, because a message arriving from Instagram or Google became a patient record only if somebody remembered to copy it across — and the field that said where it came from rarely survived the copy. First response took hours instead of minutes. Follow-ups had no owner. Revenue was steady because the founder coordinated it by hand, every day, from memory.

That's not a BELSA problem. That's the default state of every clinic, studio and consultorio that grew on WhatsApp and a spreadsheet: real revenue, real reputation, and zero defensible answer to the one question that decides where the next euro of marketing spend goes.

Why "I think it's Instagram" is a more expensive sentence than it sounds

Attribution sounds like a reporting problem. It is actually a money problem, and it shows up in three places at once:

• You can't tell which channel pays for itself. Without a source field attached to every enquiry, ad spend and referral effort get judged by vibes — and vibes have a well-known bias toward whatever channel the founder personally enjoys checking.
• You can't defend the marketing budget to your own numbers. When someone asks "where did most of our bookings come from this quarter," the honest answer is a guess with confidence attached. That's not a system. That's a founder doing mental math under pressure.
• You can't double down on what's actually working — because "what's working" is a feeling, not a number, and feelings don't survive a bad month.

Caos con buena reputación sigue siendo caos — chaos with a good reputation is still chaos. It just takes longer to notice, because the clinic is busy enough that nobody stops to ask the question. Busy is not the same as measured, and a founder can run both at once for years before the gap becomes expensive.

What actually got installed, and in how long

When RIVEL's Strategy Lab read BELSA's operation, the direction was simple: one place where a patient exists once, and the question of which channel produced which patient stops being an opinion. Not more tools stacked on top of WhatsApp — fewer tools, connected, with WhatsApp wired in as a channel instead of the record itself.

The install ran 90 days from the first session to the team running it on its own: two weeks reading the clinic and agreeing the direction, then build, validation and training on the live system — not a slide deck, the actual system the front desk would use on a Tuesday morning. Three pieces did the real work:

• Marketing Hub attaches the source to every single enquiry — Instagram, Google, referral, walk-in — the moment it arrives, before anyone has the chance to forget.
• CRM holds the full patient record: conversation, source, stage and treatment history, in one place instead of scattered between a phone and a spreadsheet. Contacts are labeled patients, because that's what the clinic calls them — the system was built around BELSA's language, not the other way around.
• Integrations connect WhatsApp Business, Kommo and Meta as channels into the system. Patients keep writing exactly where they always wrote. The record simply stopped living there.

Nothing about how BELSA talks to patients changed. The compliance context of aesthetic medicine in Barcelona — the tone, the consent steps, the boundaries around what a clinic can and can't claim — stayed exactly where it was. What changed is that every conversation now arrives with a receipt attached.

The 90 days, broken down

Nobody at BELSA sat through a demo and called it done. The timeline had three real phases, not a launch date:

1. Weeks 1–2 — reading the clinic. Strategy Lab mapped where enquiries actually came from, where they actually died, and which of the three operational problems — communication, calendar, lead capture — was costing the most. The direction got agreed before anything got built.
2. Weeks 3–10 — build and connect. Marketing Hub, CRM and the integrations layer went in around BELSA's existing services, tone and compliance context, not a generic template with the clinic's logo pasted on top.
3. Weeks 11–13 — validation and training on the live system. The front desk learned the system by running it, on real enquiries, before the training wheels came off. Ninety days after the first session, the team was operating it alone.

The objection every founder raises first

"My team won't adopt another tool" is the sentence that kills more installs than any budget conversation, and it's a reasonable objection — reception staff at a boutique clinic did not sign up to learn software. BELSA's install answered it by refusing the premise: nothing new got handed to the patient, and almost nothing new got handed to the team.

Patients kept writing on WhatsApp. The front desk kept working from one screen. The only thing that moved was where the record lived — out of a phone and a spreadsheet, into a system that remembers the source automatically instead of asking a human to. Adoption wasn't a training problem to solve after the fact; it was designed out of the install from week one, which is why the team was running it alone by day 90 instead of still fighting it.

The six numbers that replaced four screens and a memory

Ask most founders for their key numbers and you get a pause, then a guess. BELSA's founder now opens one dashboard, once a month, and reads six:

• Booking pace — how fast enquiries are turning into appointments, by week, not by feeling.
• Attendance — who actually shows up, which is the number that makes every other number honest.
• Conversion to treatment — how many consultations become paying treatments, and which source produces the patients who convert, not just the ones who message.
• Average ticket — what a patient is actually worth, by channel, which is the number that ends the "but Instagram is free" argument.
• Retention — whether patients come back, which referrals and reputation depend on more than any single ad.
• Open pipeline — enquiries sitting unanswered right now, the thing a spreadsheet hides until someone complains.

None of these six numbers existed as numbers before the install. They existed as impressions — strong ones, earned over years of running the clinic by hand, but impressions all the same. The difference between a founder's instinct and a founder's dashboard isn't that one is smarter. It's that one survives a bad month and the other doesn't.

The result, documented over the first 90 days with the clinic's own operating data: +40% online bookings, +25% conversion rate, and 3x engagement on the new site. Those are the headline numbers, and they are published with the window they cover because a figure without a method reads like marketing even when it's true. The number that matters more doesn't show up on a dashboard — it's that the question "which channel is actually working" now has an answer instead of an opinion.

"The team finally trusts the data. That's what changed everything." — Consolación Sánchez, founder of BELSA Estétic

The second-order effect: the founder stops being the attribution model

Before the install, Consolación was the attribution system. She was the one who remembered which campaign ran in March, which patient came from a referral in June, which post actually converted versus which one just got likes. That's not a system — that's a founder's memory doing a job a system should do, and memory doesn't scale and doesn't survive a bad week, a holiday, or a second location.

Today, the founder opens the same dashboard the team operates from, once a month, and reads the numbers instead of assembling them from four different places. Reception answers enquiries from inside the system in minutes instead of at the end of the day, and every new patient arrives pre-labeled with where they came from. The six numbers the founder actually opens — booking pace, attendance, conversion to treatment, average ticket, retention and open pipeline — sit in one view instead of four screens and a memory.

That's the real difference between a clinic with good intentions and a clinic with a system: not that the tools changed, but that the founder stopped being one of them.

The same decision, two different inputs

Why this isn't a BELSA-sized problem

Scale this down to a two-room boutique clinic, or up to a multi-location practice, and the shape of the problem doesn't change — only the cost of not fixing it does. Casa KiGua, RIVEL's wellness studio case study in Cancún, hit the same wall from a different angle: interest arriving from social media that never became a record, indistinguishable from someone who had never written at all. Different vertical, same leak — a channel doing real work with nobody able to prove it.

If you're running a consultorio, a studio or a clinic on WhatsApp and a shared spreadsheet, you already have the symptom, even if nobody has named it yet: you can tell a patient exactly which treatment they need in thirty seconds, but you can't tell your own business which channel paid for the lead who booked it. That gap is precisely what RIVEL's leak check is built to surface, before you spend another euro guessing.

Comprar una herramienta de marketing no resuelve esto — buying a marketing tool doesn't fix this. A dashboard bolted onto a broken intake process just produces prettier guesses dressed up as insight. What actually fixes it is the same move BELSA made: a system that attaches the source at the moment the enquiry arrives, not a report that tries to reconstruct it three months later from memory and good intentions.

What to check before you trust your own answer

Three honest questions, before you decide your attribution is fine the way it is:

1. Can you pull, right now, the conversion rate by channel for last month — without asking anyone or opening four tabs?
2. If a new enquiry arrived five minutes ago, does it already know where it came from, or does that depend on someone remembering to write it down?
3. When you say "Instagram is working," is that a number, or a feeling you've repeated enough times to trust?

If any of those answers made you pause, you're running the same system BELSA ran before the install — and it worked, right up until the day it stopped being free to not know.

WhatsApp and Excel got BELSA here. They didn't get it past here. If your own answer to those three questions is a guess, the next move isn't a better spreadsheet — it's a Strategy Lab session, or a straight look at what a RIVEL OS plan would show you about where your leads actually come from. The channel debate ends the same way it ended at BELSA: not with a better opinion, but with a system that doesn't need one. Browse more installs on the RIVEL blog.