Customer

Podcast 2 — In conversation

Speaker A: I keep seeing customer and client used interchangeably at work. Are they actually different?

Speaker B: They are. A customer buys a product or service in a straightforward transaction. A client receives professional advice or bespoke services from an expert. You are a customer at Tesco but a client of your solicitor. The distinction is about the nature of the relationship — transactional versus advisory, standard versus personalised.

Speaker A: So a SaaS company would have customers rather than clients?

Speaker B: Generally yes. If everyone gets the same product at the same price, they are customers. If each account receives tailored strategy, personalised onboarding, and a dedicated account manager, they start to feel more like clients. Many businesses deliberately use client to make customers feel special — it implies a higher level of care, even if the actual service is standardised.

Speaker A: What about the informal use — like he is a tough customer?

Speaker B: That is the colloquial sense — customer meaning a person of a particular type, usually implying they are difficult to deal with. A tough customer is someone who will not be easily persuaded or intimidated. An awkward customer causes problems. An ugly customer is someone you would not want to cross. It is mildly old-fashioned but still perfectly understood.

Speaker A: And how does customer differ from consumer?

Speaker B: A consumer is the person who actually uses or consumes the product. A customer is the person who buys it. Often they are the same person — you buy and eat the sandwich. But sometimes they differ — a parent is the customer who buys the toy, but the child is the consumer who plays with it. In business strategy, you might optimise for the customer — who pays — or the consumer — who uses. Understanding which one you are serving at any moment is crucial.

Speaker A: Is the customer always right?

Speaker B: That phrase — the customer is always right — was coined by Harry Gordon Selfridge in the early nineteen hundreds as a marketing principle. It does not mean customers are literally infallible. It means businesses should treat customer perception as reality for the purposes of service. If a customer feels wronged, resolve it regardless of who is technically correct. The principle is about commercial pragmatism, not factual accuracy.

Podcast 3 — Prompt engineering

Speaker B: Let us explore customer in development prompts. This is perhaps the most fundamental business domain word in software development. When you write customer in a prompt, you anchor the entire system around the paying user — their data, their journey, their satisfaction, and their lifetime value.

Speaker A: So it tells the AI the system is commercial rather than internal?

Speaker B: Exactly. A CRM prompt first.

Speaker B: PROMPT: Build a customer relationship management dashboard showing each customer's contact details, purchase history, communication timeline, and lifetime value score — with a customer health indicator that turns amber when engagement drops below average and red when a customer has not interacted for thirty days, plus a one-click customer outreach button.

Speaker B: Customer relationship management dashboard immediately tells the AI to build a sales-focused tool. Customer health indicator creates a retention-focused metric. The thirty-day inactivity threshold for red status produces an early warning system. One-click customer outreach bridges insight and action. The word customer five times ensures every feature serves the commercial relationship rather than internal operations.

Speaker A: Sales teams can see which customers need attention before they churn.

Speaker B: Right. A database schema prompt.

Speaker B: PROMPT: Design a database schema for a customer data platform — a central customer table with unique customer ID, contact information, and acquisition channel, related tables for customer orders, customer support tickets, and customer preferences, with a materialised view calculating customer lifetime value updated nightly and indexes optimised for customer lookup by email and phone.

Speaker B: Central customer table establishes the single source of truth. Related tables for orders, support, and preferences create a three-sixty-degree customer view. Materialised view for customer lifetime value means the expensive calculation happens asynchronously rather than on every query. Indexes on email and phone optimise the most common customer lookup patterns. The word customer throughout ensures the schema is designed around the buyer's identity rather than internal entities like transactions or products.

Speaker A: That gives every department a unified view of the same customer.

Speaker B: A segmentation prompt.

Speaker B: PROMPT: Create a customer segmentation dashboard for our e-commerce platform — automatically group customers into segments based on purchase frequency, average order value, and recency of last purchase, display each customer segment as a card with size, revenue contribution, and growth trend, with a drill-down to individual customer profiles within each segment.

Speaker B: Customer segmentation based on frequency, value, and recency is the classic RFM model that marketing teams rely on. Displaying segments as cards with revenue contribution gives instant business context. Drill-down to individual customer profiles within segments allows both macro strategy and micro targeting. The AI builds an analytics tool that marketers can immediately use because customer segmentation is a well-defined domain with established patterns.

Speaker A: Marketers can see their best customers and their at-risk customers at a glance.

Speaker B: A support portal prompt.

Speaker B: PROMPT: Build a customer support portal with a customer-facing ticket submission form, a customer knowledge base with search, a customer satisfaction survey triggered after ticket resolution, and an internal agent view showing the customer's full interaction history and account health score alongside each open ticket.

Speaker B: Customer-facing ticket form tells the AI to build for the external user rather than internal staff. Customer knowledge base with search creates self-service deflection. Customer satisfaction survey after resolution provides feedback data. The internal agent view with full customer history means support staff never ask customers to repeat information. Each use of customer distinguishes the external-facing experience from the internal tools.

Speaker A: Customers get self-service options and agents get full context — both sides win.

Speaker B: A loyalty program prompt.

Speaker B: PROMPT: Design a customer loyalty program engine for our retail application — track customer points earned per purchase, display a customer tier progression bar showing progress toward the next loyalty level, send automated customer milestone notifications when they unlock a new tier, and provide a customer rewards catalogue where accumulated points can be redeemed for discounts or exclusive products.

Speaker B: Customer loyalty program engine establishes the gamification framework. Customer tier progression bar creates visual motivation. Automated customer milestone notifications generate delight at achievement moments. Customer rewards catalogue provides the redemption mechanism that closes the value loop. The word customer throughout ensures the AI builds for the buyer's experience and motivation rather than the business's back-office tracking needs.

Speaker A: That turns transactions into an engaging long-term relationship.

Speaker B: One final prompt — onboarding.

Speaker B: PROMPT: Create a customer onboarding flow for our SaaS platform — a guided customer setup wizard that adapts steps based on the customer's selected use case, a customer progress checklist showing completed and remaining setup tasks, personalised customer tips appearing contextually during the first seven days, and a customer success milestone email sequence celebrating first value moments.

Speaker B: Customer onboarding flow tells the AI this is for paying users, not trial visitors. Adapting steps based on customer use case means branching logic. Customer progress checklist creates accountability and momentum. Contextual customer tips during seven days means time-triggered intelligence. Customer success milestone emails celebrate moments when the customer derives real value — not just when they click buttons. The word customer ensures the entire flow optimises for business outcomes rather than feature tours.

Speaker A: New customers reach their first value moment faster because the experience adapts to what they need.

Speaker B: That is customer in prompts. It anchors every system around the paying user — their identity, their journey, their satisfaction, and their lifetime value. One word that makes the AI build commercially-oriented, user-centred software rather than abstract internal tools.