A chatbot on your website is a widget somebody can close.<\/p>\n
A chatbot on WhatsApp is in the same app as their family and friends. It feels personal, and a bad experience there registers differently.<\/p>\n
A customer stuck in a loop on WhatsApp is more annoyed than one stuck on a website.<\/strong> That is the practical reason to scope these tightly.<\/p>\n Four things.<\/p>\n Answer a defined set of factual questions.<\/strong> Opening hours, order status, returns policy, delivery areas. Definite answers, definite questions.<\/p>\n Route the conversation.<\/strong> Ask one or two questions and pass it to the right team, with what has been said already attached.<\/p>\n Collect information before a person joins.<\/strong> Order number, account reference, a description of the problem. The agent starts with context instead of asking for it.<\/p>\n Cover out of hours.<\/strong> Acknowledge, capture, and set an expectation \u2014 provided it is honest about being a bot and about when somebody will reply.<\/p>\n FIGURE 1: THE FOUR THINGS THAT WORK<\/p>\n Defined questions<\/p>\n Routing<\/p>\n Out of hours<\/p>\n Anything outside its list.<\/strong> The moment a question is unusual, a bot produces a confident irrelevant answer \u2014 which is worse than no answer.<\/p>\n A customer who is already annoyed.<\/strong> Asking somebody to rephrase when they are frustrated makes it worse, quickly.<\/p>\n Multi-part problems.<\/strong> “My order arrived damaged and I was charged twice” needs a person.<\/p>\n Anything with money attached.<\/strong> Refunds, credits, disputes.<\/p>\n Four. Breaking any one costs you more than the bot saves.<\/p>\n If the bot cannot answer, it says so and offers a person. Immediately.<\/strong><\/p>\n Not after three attempts. Not after “could you rephrase that”. Once.<\/p>\n A customer being asked to reword their question is forming an opinion about your company<\/strong>, and it is not a good one.<\/p>\n Do not give it a human name and photograph. People work it out, and being caught pretending costs more trust than the automation saves.<\/p>\n A simple line at the start is enough.<\/p>\n Write down exactly what it handles. Everything else goes to a person.<\/p>\n A bot with a small confident scope is useful.<\/strong> A bot claiming to handle everything gives confident wrong answers on the questions that mattered most.<\/p>\n No prices it made up. No delivery dates it estimated. No promises about what you can do.<\/p>\n If the answer is not a fact from your system, it should not be given as one.<\/strong><\/p>\n FIGURE 2: BOTS THAT HELP AND BOTS THAT COST YOU<\/p>\n Set up well<\/p>\n Set up badly<\/p>\n The moment the bot passes to a person is where most of the damage happens.<\/p>\n During working hours:<\/strong> the conversation history goes with it. The customer never repeats themselves. If nobody is free, say so honestly \u2014 “everyone is with a customer, we will reply within the hour” beats silence.<\/p>\n Outside hours:<\/strong> be clear it is out of hours, capture their details and question, and give a realistic time. Then meet it.<\/strong><\/p>\n Never leave somebody waiting on a bot with nobody behind it.<\/strong> A conversation with no reply is worse than no bot at all.<\/p>\n Two approaches, and they suit different situations.<\/p>\n Menu-driven.<\/strong> Numbered options \u2014 “Reply 1 for order status, 2 for returns”. Predictable, and it makes the scope obvious. It also constrains the customer.<\/p>\n Free text.<\/strong> The customer types what they want, and the bot works out what they mean. Feels natural, and it fails less gracefully.<\/p>\n A menu is usually the safer choice on WhatsApp<\/strong>, because it sets expectations. The customer sees what the bot can do and does not waste a message on something it cannot.<\/p>\n A useful pattern:<\/strong> a menu, plus “or type your question and we will pass you to someone”.<\/p>\n Five numbers. One matters more than the rest.<\/p>\n Conversations started.<\/strong> Volume.<\/p>\n Resolved by the bot.<\/strong> How much it actually handled.<\/p>\n Escalations.<\/strong> How often a person was needed. Not a failure metric<\/strong> \u2014 a well-scoped bot escalates often, and that is correct.<\/p>\n Abandonment.<\/strong> How often somebody stops replying mid-conversation.<\/p>\n Blocks and reports.<\/strong> Directly threatens your quality rating.<\/p>\n Watch abandonment most closely.<\/strong> It tells you whether the bot is helping or driving people away, and it rarely appears in vendor demonstrations.<\/p>\n FIGURE 3: WHAT TO WATCH<\/p>\n Abandonment<\/p>\n Blocks and reports<\/p>\n Resolved by the bot<\/p>\n Escalations<\/p>\n Three cases where the answer is no.<\/p>\n Low message volume.<\/strong> If you get twenty conversations a day, a person handles them. A bot adds a step and no capacity.<\/p>\n Complex or high-value products.<\/strong> If every enquiry needs a real conversation, a bot delays the conversation people want.<\/p>\n No capacity to respond.<\/strong> If messages already sit unanswered, generating more of them is not an improvement.<\/p>\n Weeks 1\u20134.<\/strong> No bot. Handle conversations manually and record what people actually ask.<\/p>\n Week 5.<\/strong> Build a bot for the five most common questions only.<\/p>\n Week 6.<\/strong> Add routing and information collection.<\/p>\n Week 8.<\/strong> Review abandonment and escalation. Adjust the scope.<\/p>\n Then expand slowly<\/strong>, based on what people actually ask rather than what you assumed.<\/p>\n Point one matters most.<\/strong> Building a bot before you know what people ask means guessing, and guessing produces a bot with the wrong scope.<\/p>\n A WhatsApp bot works when it is narrow, honest about being a bot, and one message from a person.<\/strong><\/p>\n Escape after one failure. Never invent a fact. Watch abandonment.<\/strong><\/p>\n And run manually for a month first. What customers actually ask is rarely what you would have guessed.<\/p>\n Getting the same questions on WhatsApp all day?<\/p>\n Get in touch. We will look at what people actually ask before building anything \u2014 that decides whether a bot helps or annoys.<\/p>\n<\/div>","protected":false},"excerpt":{"rendered":" Why the stakes are higher here A chatbot on your website is a widget somebody can close. A chatbot on WhatsApp is in the same app as their family and friends. It feels personal, and a bad experience there registers differently. A customer stuck in a loop on WhatsApp is more annoyed than one stuck […]<\/p>\n","protected":false},"author":3,"featured_media":9411,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[27],"tags":[],"class_list":["post-9410","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-automation"],"yoast_head":"\nWhat a WhatsApp bot can do well<\/h2>\n
\n
\n
\n
What it does badly<\/h2>\n
The rules that decide whether it helps<\/h2>\n
Escape after one failure<\/h3>\n
Say it is a bot<\/h3>\n
Keep the scope small<\/h3>\n
Never invent a fact<\/h3>\n
\n
\n
The handover<\/h2>\n
Menus and free text<\/h2>\n
Measuring it<\/h2>\n
\n
\n
\n
\n
When not to have one<\/h2>\n
A sensible rollout<\/h2>\n
The short version<\/h2>\n