From WhatsApp to a Custom CRM in 15 Days: How I Fully Digitalized a Real Business
Published: (February 18, 2026 at 09:46 AM EST)
7 min read
Source: Dev.to
Source: Dev.to
# Hello dev community,
It’s been a long time since I last wrote. I always say I’ll do it more often, and then work absorbs me again. This time I decided to sit down and document something that completely changed the way I think about generating ideas for a client.
In 15 days, I faced the full digitalization of a real business. Not a theoretical side project. Not a portfolio demo. A business that moves money, inventory, people, and hard deadlines.
---
## The Starting Point
My wife and I run a virtual business selling personalized arrangements, flowers, and balloons. It operates mainly through Instagram, Facebook, and WhatsApp—a very common model across Latin America.
The flow was simple:
- PDF catalog sent via WhatsApp.
- Customer selects a product.
- Manual questions for personalization.
- Bank‑transfer payment.
- External delivery coordination.
It worked. It generated around **15 orders per week**. It was extra income, but it was completely sustained by manual processes.
We had spreadsheets with customers and orders, but without clear structure. There was no real traceability. We didn’t know with precision:
- Where orders were coming from.
- Which products rotated the most.
- Which ads were working.
- Which areas were ordering more.
- Which customers were returning.
We were operating on intuition.
---
## The Trigger
On **February 1st**, we decided we needed a change. Valentine’s Day was approaching, and the previous year we had losses from flowers that weren’t sold on time. This year had to be different.
The context had also changed: today we have AI tools that drastically accelerate development. What used to require months of planning and execution can now be approached with strategic speed.
I finally did something I had postponed for years: **build our own system**.
---
## The Paradigm Shift
I didn’t start by thinking about code. I started by thinking about the **system**, the same way I would if this were a client and not my wife’s business.
I asked myself basic questions:
- Where are the orders coming from?
- How are customers arriving?
- What frictions exist in the process?
- Which tasks are repetitive?
- What data are we not capturing?
The answer was clear: we needed to **digitalize the entire flow** and, most importantly, **measure everything**.
---
## The Solution: A Business‑Oriented Custom CRM
Instead of adapting the business to generic software, I decided to build a system aligned with the business.
I designed a CRM capable of managing:
- Orders
- Products
- Catalog
- Inventory
- Customers
- Delivery routes
- Campaign tracking
- Advisor system with commissions
---
## Advisors and Structured Organic Growth
Instagram had approximately **15 000 followers**, which was the main source of organic sales.
The advisor system did not come directly from that audience. We shared a registration link in different communities we belong to, inviting people interested in becoming advisors.
We created a small training system focused on selling the product using **HIGH‑TICKET methodology** principles: value‑based positioning, personalization, experience, and consultative closing. It wasn’t just about sharing a link; it was about structuring a better way to sell.
That’s what truly made the difference.
- We created personalized links for advisors.
- Each sale is associated with the referring advisor.
- We offered a **10 % commission**.
- More than **20 people** actively participated.
The result was immediate: a significant increase in orders and a clear attribution system.
---
## Advertising with Real Tracking
We launched quick campaigns, and each ad was connected to the CRM. For the first time, we were able to precisely measure which ads were generating real conversions, without relying solely on Meta’s dashboard.
We stopped guessing.
---
## Logistics: The Silent Problem
One of the biggest pain points was delivery. We use external couriers, which is common for this type of business.
The recurring issue was **location accuracy**—sometimes drivers couldn’t find the exact address.
We solved this by integrating maps directly into the order process. Each order now generates:
1. A structured location.
2. A clear route.
3. A delivery‑cost calculation based on logic similar to Uber’s dynamic pricing, using the **Google Maps API** to determine the best route.
This was crucial because previously our agreement with couriers was to pay whatever Uber estimated. Uber’s estimate didn’t always match the real sending time, creating inconsistencies in costs. By structuring the logic within our system, we eliminated that friction.
Operational friction dropped dramatically.
---All headings, lists, and image markdown have been cleaned up while preserving the original content and structure.
# Replacing guesswork delivery pricing with structured logic and Google Maps API calculations
## System Modules Overview
To give a clearer picture of the scope, here’s a breakdown of the core modules currently running inside the system:
- **Main Dashboard** – Operational overview with real‑time metrics: daily orders, revenue, pending deliveries, and performance indicators.
- **Analytics** – Data‑visualisation layer to analyse product performance, campaign attribution, advisor activity, and geographic demand.
- **Orders** – Centralised order management including status tracking, personalisation details, payment confirmation, and fulfilment state.
- **Supplies (Insumos)** – Raw‑material control to track flowers, balloons, packaging, and other inputs to prevent inventory loss.
- **Final Products** – Structured product definitions with pricing logic, cost structure, and margin visibility.
- **Public Catalog** – Customer‑facing catalogue dynamically connected to the backend instead of static PDFs.
- **Members** – Internal team management and role‑based access.
- **Advisors** – Referral‑tracking system with unique links, commission calculation, and performance monitoring.
- **Routes** – Delivery route planning with map integration and optimised logistics visualisation.
- **Drivers** – Courier registry with delivery‑assignment tracking and cost control.
- **Interested Countdown** – Lead‑capture system for high‑demand campaigns (e.g., Valentine’s Day), allowing pre‑registration and demand forecasting.
- **Margins** – Profitability tracking per product and per campaign.
- **GMD Users** – Administrative user management.
- **View Management** – Interface customisation and operational visibility controls.
This wasn’t just a website – it became operational infrastructure.
---
## The Result
In 15 days we built a fully functional MVP.
- **Last year:** ~20 Valentine’s‑Day orders.
- **This year:** >180 orders in a single day.
**Sales didn’t just increase:**
- Inventory loss was avoided.
- Advisors generated income.
- Routes were efficiently organised.
- Real data was obtained for future decisions.
---
## The Learning
1. **The world changed.**
Technology that used to be accessible only to medium‑ and large‑scale companies is now available to small businesses.
2. **AI amplifies developers, it doesn’t replace them.**
It allows execution in days what previously required months.
3. **The real advantage isn’t the code; it’s the ability to structure data and eliminate operational friction.**
Digitalising isn’t about “having a website.” It’s about designing a system that captures information, enables measurement, and supports optimisation.
We moved from operating on intuition to operating on evidence.
And it all started with one decision: stop treating the business as informal and start treating it as a system. That is the real paradigm shift.
---
## What’s Next
I’ll be working more deeply on this project to optimise it and bring it to a 100 % operational level. This business will grow this year. The difference is clear: today we have tools that allow us to build with speed and precision. Thank you, AI.
**See the result:**
**Get in touch:**
- Email:
- Portfolio:
- Hugging Face:
- GitHub:
I’ll be glad to connect.