¿Cómo detecta Kelmia Signals que un cliente se está yendo a la competencia?

Kelmia Signals construye el patrón de comportamiento habitual de cada cliente — frecuencia de pedido, importe medio, familias de producto. Cuando ese patrón cambia de forma progresiva y sostenida, Signals genera una alerta. En este caso, la combinación de caída de frecuencia, reducción de pedido medio y abandono de categorías específicas permitió identificar una migración a un proveedor alternativo en los primeros dos meses de funcionamiento.

¿Qué hace el equipo comercial con las alertas de Kelmia Signals?

Cada alerta incluye el contexto necesario para actuar de inmediato — el patrón esperado, la desviación detectada, la magnitud y el historial reciente del cliente. El comercial no necesita buscar datos ni hacer análisis: recibe la señal interpretada y puede contactar al cliente con información concreta sobre qué ha cambiado y cómo abordarlo.

¿Puede Kelmia Signals identificar qué familia de producto está causando una caída de ventas?

Sí. Kelmia Signals analiza el comportamiento por categoría de producto para cada cliente. Cuando un cliente reduce su compra de forma global, Signals identifica qué familias específicas están cayendo — permitiendo al comercial tener una conversación concreta sobre ese segmento en lugar de una conversación genérica sobre la relación.

Signs the sales team would never have spotted on their own

The context

A portfolio that no sales rep
can fully monitor.

This B2B distributor’s sales team does an excellent job. The problem isn’t their capability—it’s that with hundreds of active customers and thousands of SKUs in the catalog, it’s humanly impossible to detect all behavioral changes in time.

Before Kelmia Signals, alerts came too late: a customer who had already reduced their order for weeks before anyone noticed, a missed order that wasn’t recovered until the next cycle, a product family that was declining without anyone knowing why.

The goal was to transform that reactive approach into something different: a system that monitors the entire portfolio every day and alerts when something changes—so the sales rep can act before the problem escalates.

The Client

Industry: B2B Distribution

Portfolio: Hundreds of active clients

Catalog: Thousands of SKUs

Previous Situation: Reactive portfolio management

Solution: Kelmia Signals

Time to first signals: First week

Signals detected

What Signals saw
before anyone else.

These are the specific signals that Kelmia Signals detected in the first two months of operation. None of them would have arrived in time through the usual channels.

Signal 1 — Ongoing supplier migration

A customer with a consistent purchase history began to reduce the frequency of their orders. Then their average order size decreased. Then they stopped buying certain categories they had always ordered. The combined pattern is recognizable: the customer was actively testing a competitor.

The alert reached the sales team with a complete analysis: what had changed, since when, in which categories, and the magnitude of the deviation from historical behavior.

→ The sales rep was able to have a proactive conversation with the right context—not a generic follow-up call, but a specific conversation about what had changed.

Signal 2 — Customers who had simply forgotten to order

Several customers with regular purchasing patterns had gone longer than usual without placing an order. It wasn’t dissatisfaction—it was forgetfulness. Without Signals, that order would have arrived late or not at all until the next regular cycle.

The alert allowed the sales rep to reach out with a specific message: “I see you haven’t ordered in X days—is everything okay? Do you need us to prepare anything?”—without coming across as intrusive, and with a genuine reason to call.

→ Several orders recovered before the customer sought an alternative out of urgency.

Signal 3 — The product family that was underperforming

A customer was buying less overall — but Signals’ analysis identified that the drop was concentrated in a specific product family, not across the entire relationship. The rest of the catalog continued to perform normally.

That completely changed the nature of the conversation. Instead of discussing the relationship in general, the sales rep was able to focus specifically on that category—availability, price, alternatives, and competition.

→ The root cause of the decline in that product family was identified, and a conversation was initiated that would have taken months to reach through the usual channels.

What changed

From reactive management
to context-driven conversations.

The impact of Kelmia Signals isn’t measured solely in recovered sales—it’s measured in the type of conversation the sales team can now have with its customers.

Before Signals

"Everything okay? It’s been a while since you placed an order."

"I noticed sales have dropped a bit this month..."

Routine follow-up call without concrete data.

With Signals

"It's been 18 days since your last order—do you need anything?"

"I noticed that category X has dropped by 40%—is there anything we can do to improve that line?"

A targeted conversation, backed by data, at the right time.

Alerts designed to monitor drops in frequency and average order value are very useful for detecting dissatisfaction and correcting it in time, identifying missed orders, and knowing exactly which product families are causing the decline.

The system evolves

Signals are the beginning.
Actions are the goal.

Kelmia Signals is designed to integrate with the rest of the Kelmia ecosystem. Signals alerts can directly trigger workflows in Kelmia Offers—generating a personalized proposal for the at-risk customer—or in Kelmia Flow, to automatically manage recovered orders.

Next step: from alert to automatic action

We are working on integrating Signals with Kelmia Offers so that, when Signals detects a customer at risk, the system automatically generates a personalized proposal featuring products from that category—ready for the sales representative to review and send.

See Kelmia Offers →