¿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.

Signals the sales team nunca habría detectado solo

The context

A portfolio no salesperson
can fully monitor.

The sales team of this B2B distributor does excellent work. The problem isn't their capability — it's that with hundreds of active clients and thousands of catalogue references, it's humanly impossible to detect all behaviour changes in time.

Before Kelmia Signals, signals arrived late: a client who had already reduced their order for weeks before anyone noticed, a forgotten order that wasn't recovered until the next cycle, a product family declining with nobody knowing why.

The goal was to turn that reactive management into something different: a system that monitors the complete portfolio every day and alerts when something changes — so the salesperson can act before the problem escalates.

The client

SectorB2B Distribution

PortfolioHundreds of active clients

CatalogueThousands of references

Previous situationReactive portfolio management

SolutionKelmia Signals

Time to first signalsFirst week

The signals detected

What Signals saw
before anyone else.

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

Signal 1 — Supplier migration in progress

A client with a consistent purchase history started reducing their order frequency. Then lowered their average order. Then stopped buying certain categories they had always ordered. The combined pattern is recognisable: the client was actively trying out a competitor.

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

→ The salesperson could have a proactive conversation with the right context — not a generic follow-up call, but a specific conversation about what had changed.

Signal 2 — Clients who had simply forgotten to order

Several clients with regular buying patterns had gone more days than usual without placing an order. It wasn't dissatisfaction — it was forgetfulness. Without Signals, that order would have arrived late or not until the next usual cycle.

The alert allowed the salesperson to make contact with a specific message: "I see you haven't ordered in X days, everything OK? Do you need us to prepare something?" — without seeming intrusive, with a real reason to call.

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

Signal 3 — The product family that was failing

A client was buying less overall — but Signals' analysis identified that the decline was concentrated in a specific product family, not in the relationship as a whole. The rest of the catalogue continued with normal behaviour.

That completely changed the type of conversation. Instead of talking about the relationship in general, the salesperson could talk specifically about that category — availability, price, alternatives, competition.

→ Root cause of the decline in that family identified, opening a conversation that would have taken months to arrive through usual channels.

What changed

From reactive management
to conversations with context.

The impact of Kelmia Signals is not measured only in recovered sales — it's measured in the type of conversation the sales team can now have with their clients.

Before Signals

"Everything OK? You haven't ordered in a while."

"I've seen that sales have dropped a bit this month..."

Routine follow-up call without concrete data.

With Signals

"You haven't ordered in 18 days — do you need anything?"

"I've seen category X has dropped 40% — is there anything we can improve in that line?"

Specific conversation, with data, at the right moment.

The alerts created to monitor frequency drops and average orders are very useful for detecting dissatisfaction and correcting it in time, identifying forgotten 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 flows in Kelmia Offers — generating a personalised proposal for the at-risk client — or in Kelmia Flow, to automatically manage recovered orders.

Next step: from alert to automatic action

We are working on the integration of Signals with Kelmia Offers so that, when Signals detects an at-risk client, the system automatically generates a personalised proposal with products from that category — ready for the salesperson to review and send.

Ver Kelmia Offers →