Conversational assistants that resolve 60-70% of enquiries without an operator and properly route those that need a human. 24/7 service with judgement, not the chatbot of the 2010s. The difference lies in how it's designed and how it's governed.
Order status, returns, sizes, availability. 70% of enquiries are resolved with information already available — the chatbot frees the team for the complex ones.
Consultancies, law firms, clinics, real estate. The chatbot pre-qualifies the lead and books an initial session, saving the sales team manual filtering.
Hotel with 24/7 enquiries (check-in, services, bookings, recommendations). Human service in critical hours + chatbot for the rest. A realistic and profitable combination.
Technology platforms with many recurring configuration questions. The chatbot integrated with your documentation resolves issues without friction for users who otherwise drop off.
Audit of the real enquiries of the last year (email, phone, chat, social). Patterns, volumes, complexity, languages. Without this analysis, the chatbot is designed blind.
Construction of conversation flows with judgement: when to answer, when to route to a human, how to handle edge cases, what tone to use. Good conversation design is UX, not technology.
Technical build with platform (Voiceflow, Botpress, custom GPT, ManyChat) + integration with your CRM, ERP, e-commerce, booking system. Without integration the chatbot is decorative.
Monitoring of real conversations, identification of gaps (what the chatbot doesn't resolve), iterative improvement. Without this, the chatbot degrades quickly.
The chatbot is neither magic nor threat — it's a tool. Well-designed:
Resolves 60-70% of enquiries automatically. Your support team focuses on the complex 30% — more professionally satisfying and more relevant to the customer.
Hospitality, international e-commerce, urgent services. The chatbot covers gaps that can't be filled humanly without night shifts.
The chatbot converses, asks, qualifies and only routes leads with judgement to sales. Saves sales time — and improves conversion ratio.
Every conversation is data: what people ask, what's missing on your site, what to improve in the product. The well-governed chatbot is a source of insight not detectable any other way.
Same customer, same query, same answer. The chatbot guarantees consistency that's never humanly achieved, especially in large teams.
A modern chatbot manages 50+ languages at quality close to human. International markets viable without hiring local support.
Online store with saturated 3-person team. Chatbot with integration to order status + returns + FAQ. 62% of enquiries self-resolved. Team focused on complex cases.
El marketing del cerebro es más predictible que el marketing de la opinión. — Ángel Ortega Castro
Canary Islands resort. Chatbot answering on services, local recommendations, schedules. 8 languages. Reception workload reduced significantly during peak hours.
B2B company with acquisition site. Chatbot pre-qualifying leads through natural conversation. Sales receives only valid leads. Conversion to session +40%.
CRM, events, content, campaign history.
Trained or generative algorithm guided by rules.
Lead prioritised, content published, decision taken.
Deploying a chatbot makes sense in these four scenarios. Outside them it's usually a questionable investment — and it pays to know that:
If your team spends daily hours answering the same things (order status, hours, prices, conditions), the chatbot pays back quickly.
Hospitality, international e-commerce, urgent services. If 24 hours can't be covered humanly and the customer demands it, the chatbot is the only profitable route.
If your site receives forms without qualification and sales loses time filtering, the chatbot pre-qualifies better and leaves only what merits human attention.
A modern chatbot handles languages that humanly require hiring natives. For small or exploratory markets, it's the pragmatic route.
They hate bad chatbots — the rigid decision-tree ones from 10 years ago. Modern chatbots with well-designed conversational AI are accepted, especially if the exit to a human is clear and fast when the case requires it. The difference lies in design.
No, and they shouldn't. Good practice is to transparently declare that it is an automated assistant, offering the option to speak with a human when preferred. Deceiving the customer about the nature of their interlocutor is ethically questionable and commercially counterproductive.
It depends on the case. For a simple chatbot: Voiceflow, Botpress, ManyChat. For complex cases: integration with LLM (ChatGPT, Claude) + RAG with your documentation. For large volumes and strict compliance: enterprise solutions like Salesforce Service Cloud Einstein.
A simple chatbot (FAQ + basic integration) falls within reasonable budgets. A complex assistant (conversational AI, RAG with documentation, integration with several systems) scales proportionally. Monthly running cost depends on platform and volume.
A real risk. That's why the design includes: limited confidence in answers (easy human handoff), conversation monitoring, restriction of sensitive information (specific prices, legal clauses). A well-designed chatbot knows what it doesn't know.
First 45-minute session, free of charge and no commitment. If we fit, I send you a detailed proposal within 5 days. If we don't, you take away a useful initial diagnosis.