{"id":181,"date":"2026-08-04T18:41:00","date_gmt":"2026-08-04T18:41:00","guid":{"rendered":"https:\/\/www.totaliweb.com\/ai-chatbots-in-2026-why-your-website-needs-a-trained-agent-not-a-script\/"},"modified":"2026-08-04T18:41:00","modified_gmt":"2026-08-04T18:41:00","slug":"ai-chatbots-in-2026-why-your-website-needs-a-trained-agent-not-a-script","status":"publish","type":"post","link":"https:\/\/www.totaliweb.com\/it\/ai-chatbots-in-2026-why-your-website-needs-a-trained-agent-not-a-script\/","title":{"rendered":"AI Chatbots in 2026: Why Your Website Needs a Trained Agent, Not a Script"},"content":{"rendered":"\n<p>Here&#8217;s the uncomfortable truth most agencies won&#8217;t tell you: <strong>the chatbot you installed three years ago is probably costing you customers right now.<\/strong> Scripted bots with decision trees and canned replies have become a trust-killer \u2014 visitors recognize them instantly, and 67% of users in a 2026 Drift\/Salesforce survey said they&#8217;d abandon a site that gave them a &#8220;robot loop&#8221; instead of a real answer.<\/p>\n\n<p><strong>TL;DR:<\/strong> In August 2026, a website without a well-trained AI agent is the equivalent of a store with an unhelpful front-desk worker who can&#8217;t answer a single product question. The gap between scripted bots and data-trained AI agents is now so wide it directly impacts revenue.<\/p>\n\n<nav class=\"totaliweb-toc glass-panel border border-white\/5 rounded-2xl p-6 shadow-xl my-8 is-empty wp-block-totaliweb-toc\" data-toc=\"true\" aria-label=\"In this article\">\n\t<h4 class=\"font-bold text-sm mb-4 text-white uppercase tracking-wider flex items-center gap-2\">\n\t\t<i class=\"fa-solid fa-list-ul text-primary\" aria-hidden=\"true\"><\/i>\n\t\tIn this article\t<\/h4>\n\t<ul class=\"space-y-3 text-sm font-medium\" data-toc-list><\/ul>\n<\/nav>\n\n\n<h2>The Scripted Bot Era Is Over \u2014 Here&#8217;s What Killed It<\/h2>\n<p>For years, the chatbot market was dominated by flow-based platforms: <em>if user says X, reply with Y<\/em>. These tools were cheap and easy to set up, but they optimized for <strong>deflection<\/strong>, not <strong>resolution<\/strong>. Fast-forward to mid-2026, and the landscape has been fundamentally disrupted by large language models capable of holding genuinely contextual, multi-turn conversations.<\/p>\n<p>Users have been exposed to powerful AI assistants in their personal lives \u2014 ChatGPT, Gemini Advanced, Claude \u2014 and their expectations have permanently shifted. Landing on a website and getting a scripted &#8220;How can I help you today? (a) Pricing (b) Support (c) Other&#8221; widget now feels jarring, even insulting. The contrast is too stark.<\/p>\n\n<div class=\"my-8 glass-panel border rounded-2xl p-6 flex items-start gap-4 border-warning\/30 bg-warning\/[0.06] shadow-[0_0_30px_rgba(255,215,0,0.06)] wp-block-totaliweb-callout\">\n\t<div class=\"text-2xl flex-shrink-0 mt-0.5\">\n\t\t<i class=\"fa-solid fa-triangle-exclamation text-warning\" aria-hidden=\"true\"><\/i>\n\t<\/div>\n\t<div class=\"min-w-0\">\n\t\t<div class=\"font-bold mb-1 text-warning\">The silent drop-off you&#039;re missing<\/div>\n\t\t\t\t\t<p class=\"text-gray-300 text-sm leading-relaxed m-0\">Most analytics platforms don&#8217;t track chat abandonment separately from page bounces. If your bot can&#8217;t answer a question, visitors leave silently \u2014 and that lost conversion never shows up in your chatbot reports.<\/p>\n\t\t\t<\/div>\n<\/div>\n\n\n<h2>What &#8220;Trained on Your Data&#8221; Actually Means<\/h2>\n<p>A <strong>data-trained AI chatbot<\/strong> is an AI agent that has been fed your specific business knowledge: your service pages, FAQs, product catalog, pricing structures, case studies, policies, and even past support tickets. Instead of matching keywords to scripted replies, it understands <em>intent<\/em> and synthesizes answers from everything you&#8217;ve told it about your business.<\/p>\n<p>The practical difference is enormous:<\/p>\n\n<div class=\"my-10 glass-panel border border-white\/10 rounded-2xl overflow-hidden overflow-x-auto custom-scrollbar wp-block-totaliweb-comparison-table\">\n\t<table class=\"w-full text-left border-collapse text-sm md:text-base min-w-[480px]\">\n\t\t<thead>\n\t\t\t<tr class=\"bg-white\/5 border-b border-white\/10\">\n\t\t\t\t<th class=\"py-4 px-5 font-bold text-white uppercase tracking-wider text-xs\">Caratteristica<\/th>\n\t\t\t\t<th class=\"py-4 px-5 font-bold text-primary uppercase tracking-wider text-xs text-center\">Opzione A<\/th>\n\t\t\t\t<th class=\"py-4 px-5 font-bold text-accent uppercase tracking-wider text-xs text-center\">Opzione B<\/th>\n\t\t\t<\/tr>\n\t\t<\/thead>\n\t\t<tbody>\n\t\t\t\t\t\t\t\t\t\t\t<tr class=\"border-b border-white\/5 last:border-0 hover:bg-white\/[0.03] transition-colors\">\n\t\t\t\t\t<td class=\"py-4 px-5 font-semibold text-white\">Velocit\u00e0<\/td>\n\t\t\t\t\t<td class=\"py-4 px-5 text-gray-300 text-center\">Alta<\/td>\n\t\t\t\t\t<td class=\"py-4 px-5 text-gray-300 text-center\">Media<\/td>\n\t\t\t\t<\/tr>\n\t\t\t\t\t\t\t\t\t\t\t<tr class=\"border-b border-white\/5 last:border-0 hover:bg-white\/[0.03] transition-colors\">\n\t\t\t\t\t<td class=\"py-4 px-5 font-semibold text-white\">Costo<\/td>\n\t\t\t\t\t<td class=\"py-4 px-5 text-gray-300 text-center\">\u20ac\u20ac<\/td>\n\t\t\t\t\t<td class=\"py-4 px-5 text-gray-300 text-center\">\u20ac<\/td>\n\t\t\t\t<\/tr>\n\t\t\t\t\t<\/tbody>\n\t<\/table>\n<\/div>\n\n\n<h2>The Human Handover Problem (And Why Most Bots Fail at It)<\/h2>\n<p>Even the best AI agent will encounter a moment where a human needs to step in \u2014 a complex complaint, a high-value negotiation, a sensitive customer situation. This is where nearly every off-the-shelf chatbot solution falls apart. The handover is cold: the human agent receives a ticket number and no context. The customer is forced to repeat themselves. Frustration spikes.<\/p>\n<p>In a well-architected 2026 AI agent, human handover is a <strong>warm, intelligent transition<\/strong>. The AI detects escalation signals (repeated confusion, negative sentiment, high-value intent), summarises the conversation in structured form, and routes to the right human with full context intact. The customer barely notices the switch.<\/p>\n<p>This isn&#8217;t a feature \u2014 it&#8217;s a <strong>revenue safeguard<\/strong>. High-intent visitors who would have been lost to a scripted dead-end are now captured by a human who already knows what they need.<\/p>\n\n<div class=\"my-8 glass-panel border rounded-2xl p-6 flex items-start gap-4 border-success\/30 bg-success\/[0.06] shadow-[0_0_30px_rgba(0,255,163,0.06)] wp-block-totaliweb-callout\">\n\t<div class=\"text-2xl flex-shrink-0 mt-0.5\">\n\t\t<i class=\"fa-solid fa-lightbulb text-success\" aria-hidden=\"true\"><\/i>\n\t<\/div>\n\t<div class=\"min-w-0\">\n\t\t<div class=\"font-bold mb-1 text-success\">Pro tip<\/div>\n\t\t\t\t\t<p class=\"text-gray-300 text-sm leading-relaxed m-0\">Train your AI agent on your case studies and testimonials, not just your service pages. When a visitor asks &#8216;can you handle a project like mine?&#8217;, the agent can cite a real example \u2014 that&#8217;s genuine sales performance.<\/p>\n\t\t\t<\/div>\n<\/div>\n\n\n<h2>Where Data-Trained AI Agents Win on Real Websites<\/h2>\n<p>The ROI from a properly trained conversational AI agent shows up across several parts of the funnel simultaneously:<\/p>\n<ul>\n  <li><strong>Top of funnel \u2014 discovery:<\/strong> Visitors from organic search or social ads land with varied intent. A trained agent can qualify, segment, and guide each visitor toward the right offer without a human on call 24\/7.<\/li>\n  <li><strong>Mid-funnel \u2014 objection handling:<\/strong> &#8220;What&#8217;s included in your plan?&#8221; &#8220;How long does the project take?&#8221; &#8220;Do you have experience in my industry?&#8221; These micro-objections kill conversions when left unanswered. A trained agent answers them instantly, at 2 a.m., on mobile, without delay.<\/li>\n  <li><strong>Bottom of funnel \u2014 lead capture:<\/strong> Rather than a static contact form, a conversational agent collects lead data progressively \u2014 name, company, budget, timeline \u2014 in a natural dialogue that feels helpful, not interrogative. Conversion rates on progressive conversational forms are consistently 2-4\u00d7 higher than static forms.<\/li>\n  <li><strong>Post-sale \u2014 support deflection:<\/strong> Trained on your documentation and policies, the agent handles the majority of tier-1 support questions, freeing your team for work that actually requires human judgment.<\/li>\n<\/ul>\n\n<h2>What Makes an AI Agent &#8220;Good&#8221; in 2026<\/h2>\n<p>Not all AI chatbots are equal. By August 2026, the market has fragmented into low-quality widget plugins and genuinely powerful custom agents. Here&#8217;s what separates them:<\/p>\n<ol>\n  <li><strong>Knowledge depth:<\/strong> The agent should be trained on <em>all<\/em> your content \u2014 not just one FAQ page. This requires a deliberate indexing and chunking strategy, not a five-minute plugin install.<\/li>\n  <li><strong>Persona and tone alignment:<\/strong> The agent&#8217;s language should sound like your brand, not like a generic LLM. This requires prompt engineering and real testing.<\/li>\n  <li><strong>Fallback intelligence:<\/strong> When the agent doesn&#8217;t know something, it should say so clearly and offer next steps \u2014 not invent an answer. Hallucination in customer-facing AI is a brand-risk, not just a tech problem.<\/li>\n  <li><strong>Integration with your stack:<\/strong> A useful agent in 2026 connects to your CRM, booking system, or e-commerce backend \u2014 it doesn&#8217;t just answer, it <em>acts<\/em>. This is the boundary where a chatbot becomes a true <a href=\"\/services\/ai-agent\/\">autonomous AI agent<\/a>.<\/li>\n  <li><strong>Observability:<\/strong> You need dashboards showing what questions were asked, which ones the agent struggled with, and where humans stepped in. Continuous improvement requires data.<\/li>\n<\/ol>\n\n<div class=\"my-10 glass-panel border border-white\/10 rounded-2xl p-6 md:p-8 wp-block-totaliweb-chart\">\n\t\t\t<h4 class=\"font-bold text-lg text-white mb-6 flex items-center gap-2\">\n\t\t\t<i class=\"fa-solid fa-chart-simple text-primary\" aria-hidden=\"true\"><\/i>\n\t\t\tWhere AI Agents Outperform Scripted Bots (Impact Score)\t\t<\/h4>\n\t\t<div class=\"space-y-5\">\n\t\t\t\t\t\t\t\t<div>\n\t\t\t\t<div class=\"flex items-center justify-between mb-1.5 text-sm\">\n\t\t\t\t\t<span class=\"font-semibold text-gray-200\">Lead Qualification<\/span>\n\t\t\t\t\t<span class=\"font-bold text-white tabular-nums\">91<\/span>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"w-full h-3 rounded-full bg-white\/5 overflow-hidden\">\n\t\t\t\t\t<div class=\"h-full rounded-full transition-all duration-700\" style=\"width:95.79%;background:linear-gradient(90deg,#9D4EDD,#9D4EDD);box-shadow:0 0 18px rgba(157,78,221,0.35);\"><\/div>\n\t\t\t\t<\/div>\n\t\t\t<\/div>\n\t\t\t\t\t\t\t\t<div>\n\t\t\t\t<div class=\"flex items-center justify-between mb-1.5 text-sm\">\n\t\t\t\t\t<span class=\"font-semibold text-gray-200\">Objection Handling<\/span>\n\t\t\t\t\t<span class=\"font-bold text-white tabular-nums\">87<\/span>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"w-full h-3 rounded-full bg-white\/5 overflow-hidden\">\n\t\t\t\t\t<div class=\"h-full rounded-full transition-all duration-700\" style=\"width:91.58%;background:linear-gradient(90deg,#00F0FF,#00F0FF);box-shadow:0 0 18px rgba(0,240,255,0.35);\"><\/div>\n\t\t\t\t<\/div>\n\t\t\t<\/div>\n\t\t\t\t\t\t\t\t<div>\n\t\t\t\t<div class=\"flex items-center justify-between mb-1.5 text-sm\">\n\t\t\t\t\t<span class=\"font-semibold text-gray-200\">24\/7 Availability<\/span>\n\t\t\t\t\t<span class=\"font-bold text-white tabular-nums\">95<\/span>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"w-full h-3 rounded-full bg-white\/5 overflow-hidden\">\n\t\t\t\t\t<div class=\"h-full rounded-full transition-all duration-700\" style=\"width:100%;background:linear-gradient(90deg,#00FFA3,#00FFA3);box-shadow:0 0 18px rgba(0,255,163,0.35);\"><\/div>\n\t\t\t\t<\/div>\n\t\t\t<\/div>\n\t\t\t\t\t\t\t\t<div>\n\t\t\t\t<div class=\"flex items-center justify-between mb-1.5 text-sm\">\n\t\t\t\t\t<span class=\"font-semibold text-gray-200\">Human Handover Quality<\/span>\n\t\t\t\t\t<span class=\"font-bold text-white tabular-nums\">82<\/span>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"w-full h-3 rounded-full bg-white\/5 overflow-hidden\">\n\t\t\t\t\t<div class=\"h-full rounded-full transition-all duration-700\" style=\"width:86.32%;background:linear-gradient(90deg,#FFD700,#FFD700);box-shadow:0 0 18px rgba(255,215,0,0.35);\"><\/div>\n\t\t\t\t<\/div>\n\t\t\t<\/div>\n\t\t\t\t\t\t\t\t<div>\n\t\t\t\t<div class=\"flex items-center justify-between mb-1.5 text-sm\">\n\t\t\t\t\t<span class=\"font-semibold text-gray-200\">Post-Sale Support<\/span>\n\t\t\t\t\t<span class=\"font-bold text-white tabular-nums\">78<\/span>\n\t\t\t\t<\/div>\n\t\t\t\t<div class=\"w-full h-3 rounded-full bg-white\/5 overflow-hidden\">\n\t\t\t\t\t<div class=\"h-full rounded-full transition-all duration-700\" style=\"width:82.11%;background:linear-gradient(90deg,#FF1053,#FF1053);box-shadow:0 0 18px rgba(255,16,83,0.35);\"><\/div>\n\t\t\t\t<\/div>\n\t\t\t<\/div>\n\t\t\t<\/div>\n<\/div>\n\n\n<h2>The Hidden Cost of Doing This Badly<\/h2>\n<p>There&#8217;s a real risk in deploying a poorly trained AI agent that&#8217;s just as costly as leaving the old scripted bot in place. A hallucinating agent that quotes wrong prices, promises services you don&#8217;t offer, or gives legally sensitive advice is a liability. So is an agent trained only on a handful of pages \u2014 it will confidently confabulate when questions venture outside its narrow knowledge.<\/p>\n<p>This is why architecture matters as much as the AI itself. The knowledge base needs to be curated, chunked correctly, and updated in sync with your actual offerings. The prompt layer needs guardrails. The escalation logic needs to be tested against real conversation paths. These aren&#8217;t one-afternoon tasks \u2014 they&#8217;re the craftsmanship behind a system that actually works.<\/p>\n<p>If you&#8217;re curious what this looks like in practice, our <a href=\"\/case-studies\/\">case studies<\/a> show how we&#8217;ve built and deployed these systems for real businesses \u2014 including how a dental group boosted appointment bookings significantly with a fully trained, human-handover-ready agent (<a href=\"\/case-studies\/dentiweb\/\">see the Dentiweb case study<\/a>).<\/p>\n\n<h2>Is Your Current Chatbot Helping or Hurting?<\/h2>\n<p>Ask yourself four direct questions:<\/p>\n<ul>\n  <li>Can your current chatbot answer a nuanced question about a specific service, using language from your own website?<\/li>\n  <li>When it fails, does it hand over to a human with full context \u2014 or just drop the visitor on a contact page?<\/li>\n  <li>Do you have data on which questions it couldn&#8217;t answer last month?<\/li>\n  <li>Has it been updated every time your pricing, services, or policies changed?<\/li>\n<\/ul>\n<p>If any answer is &#8220;no&#8221; or &#8220;I&#8217;m not sure,&#8221; you have a leaking funnel. The good news: this is a solvable problem, and solving it is one of the highest-leverage investments a growing business can make on its website right now.<\/p>\n<p>Our <a href=\"\/services\/ai-chatbot\/\">AI chatbot service<\/a> is purpose-built for exactly this: a conversational agent trained on your data, aligned to your brand voice, connected to your stack, and equipped with intelligent human handover \u2014 deployed on any website, including WordPress.<\/p>\n<section class=\"tw-faq glass-panel\" style=\"margin:2.5rem 0;padding:1.75rem;border:1px solid rgba(255,255,255,0.08);border-radius:1.5rem;\"><h2 style=\"margin:0 0 .5rem;\">Frequently asked questions<\/h2><div class=\"tw-faq-item\" style=\"border-top:1px solid rgba(255,255,255,0.08);padding:1.25rem 0;\"><h3 style=\"font-size:1.15rem;margin:0 0 .5rem;color:#fff;\">What&#039;s the difference between an AI chatbot and a scripted chatbot?<\/h3><p style=\"color:#9ca3af;margin:0;line-height:1.7;\">A scripted chatbot follows pre-defined decision trees and can only respond to questions it was explicitly programmed for. An AI chatbot uses a large language model trained on your business data to understand intent and generate contextual, accurate answers \u2014 handling questions it has never seen before.<\/p><\/div><div class=\"tw-faq-item\" style=\"border-top:1px solid rgba(255,255,255,0.08);padding:1.25rem 0;\"><h3 style=\"font-size:1.15rem;margin:0 0 .5rem;color:#fff;\">How is an AI chatbot trained on my business data?<\/h3><p style=\"color:#9ca3af;margin:0;line-height:1.7;\">Your website content, product pages, FAQs, policies, case studies, and documentation are indexed and structured into a knowledge base. The AI agent is then configured to retrieve and synthesise answers from this knowledge base, keeping responses grounded in your actual information rather than generating generic replies.<\/p><\/div><div class=\"tw-faq-item\" style=\"border-top:1px solid rgba(255,255,255,0.08);padding:1.25rem 0;\"><h3 style=\"font-size:1.15rem;margin:0 0 .5rem;color:#fff;\">What does &#039;human handover&#039; mean in an AI chatbot?<\/h3><p style=\"color:#9ca3af;margin:0;line-height:1.7;\">Human handover is the moment when the AI agent recognises it can&#039;t resolve a query \u2014 or detects a high-value or sensitive conversation \u2014 and transfers the user to a real team member. In a well-built system, the human receives the full conversation transcript and an intent summary, so the customer never has to repeat themselves.<\/p><\/div><div class=\"tw-faq-item\" style=\"border-top:1px solid rgba(255,255,255,0.08);padding:1.25rem 0;\"><h3 style=\"font-size:1.15rem;margin:0 0 .5rem;color:#fff;\">Can an AI chatbot work on a WordPress website?<\/h3><p style=\"color:#9ca3af;margin:0;line-height:1.7;\">Yes. AI chatbots trained on business data can be deployed on any website, including WordPress, via lightweight JavaScript embeds or purpose-built plugins. Integration with CRM tools, booking systems, and email platforms is also possible regardless of the underlying CMS.<\/p><\/div><div class=\"tw-faq-item\" style=\"border-top:1px solid rgba(255,255,255,0.08);padding:1.25rem 0;\"><h3 style=\"font-size:1.15rem;margin:0 0 .5rem;color:#fff;\">How long does it take to deploy a properly trained AI chatbot?<\/h3><p style=\"color:#9ca3af;margin:0;line-height:1.7;\">A well-architected, data-trained AI chatbot with human handover typically takes 2\u20134 weeks to build and deploy, depending on the size of the knowledge base and the complexity of integrations. Quick-install plugins take minutes but deliver the generic experience that drives visitors away \u2014 the build time is the investment that makes it effective.<\/p><\/div><\/section><div class=\"tw-article-cta glass-panel\" style=\"margin:2.75rem 0;padding:1.75rem;border:1px solid #9D4EDD55;border-radius:1.5rem;background:#9D4EDD14;\"><div style=\"display:flex;align-items:flex-start;gap:1rem;\"><span style=\"flex:none;width:3.25rem;height:3.25rem;border-radius:1rem;display:flex;align-items:center;justify-content:center;background:#9D4EDD22;border:1px solid #9D4EDD55;\"><i class=\"fa-solid fa-robot\" style=\"color:#9D4EDD;font-size:1.4rem;\"><\/i><\/span><div style=\"flex:1;\"><div style=\"font-size:.8rem;text-transform:uppercase;letter-spacing:.08em;color:#9D4EDD;font-weight:700;\">Fatto per te da Totaliweb<\/div><strong style=\"display:block;font-size:1.3rem;margin:.15rem 0 .4rem;color:#fff;\">Chatbot AI personalizzato<\/strong><p style=\"color:#9ca3af;margin:0 0 .35rem;line-height:1.6;\">Un chatbot AI addestrato sul tuo sito, sui tuoi PDF e sulle tue FAQ. Costruito e testato prima sui nostri server \u2014 installato sul tuo sito solo dopo la tua approvazione. Un sito incluso.<\/p><p style=\"color:#cbd5e1;margin:0 0 1rem;font-weight:600;font-size:.95rem;\">Da \u20ac499 \u00b7 7 giorni<\/p><a href=\"https:\/\/www.totaliweb.com\/it\/services\/ai-chatbot\/\" class=\"btn-gradient\" style=\"display:inline-block;padding:.7rem 1.4rem;border-radius:9999px;color:#fff;font-weight:700;text-decoration:none;margin-right:.75rem;\">Richiedi un preventivo gratuito<\/a><a href=\"https:\/\/www.totaliweb.com\/it\/case-studies\/\" style=\"color:#9D4EDD;font-weight:600;text-decoration:none;\">Esplora i nostri casi studio \u2192<\/a><\/div><\/div><\/div>","protected":false},"excerpt":{"rendered":"<p>Generic chatbot scripts are losing customers \u2014 learn why AI agents trained on your own business data are the 2026 standard for websites that convert.<\/p>\n","protected":false},"author":0,"featured_media":182,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[21],"tags":[166,22,182,28,185,184,183,5],"class_list":["post-181","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-ai-automation","tag-2026-trends","tag-ai-agents","tag-ai-chatbot","tag-conversational-ai","tag-customer-support","tag-lead-generation","tag-website-automation","tag-wordpress"],"_links":{"self":[{"href":"https:\/\/www.totaliweb.com\/it\/wp-json\/wp\/v2\/posts\/181","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.totaliweb.com\/it\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.totaliweb.com\/it\/wp-json\/wp\/v2\/types\/post"}],"replies":[{"embeddable":true,"href":"https:\/\/www.totaliweb.com\/it\/wp-json\/wp\/v2\/comments?post=181"}],"version-history":[{"count":0,"href":"https:\/\/www.totaliweb.com\/it\/wp-json\/wp\/v2\/posts\/181\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.totaliweb.com\/it\/wp-json\/wp\/v2\/media\/182"}],"wp:attachment":[{"href":"https:\/\/www.totaliweb.com\/it\/wp-json\/wp\/v2\/media?parent=181"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.totaliweb.com\/it\/wp-json\/wp\/v2\/categories?post=181"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.totaliweb.com\/it\/wp-json\/wp\/v2\/tags?post=181"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}