Long-form field reports from inside live WhatsApp AI deployments, CRM automation builds, and paid-media campaigns. Written for operators, not consultants.
SPF has told WhatsApp, Telegram, Facebook, Instagram, TikTok and the major marketplaces to lock scammers out by 31 January 2027: warning labels on unknown senders, identity checks on advertisers. What the codes require, and the practices that keep a legitimate business on the right side of the new filters.
Read the guide
Most AI deployments answer FAQs, escalate the hard cases, and stop, exactly where a considered purchase begins. Why the follow-up window decides higher-priced sales, and how we structure agents, sequences and handoffs to work it.
Read the guide
One ad account, 76,000+ conversations started, more than S$18 million in client topline revenue. Why a phone number beats an email address, the follow-up frameworks that close 350% more than email, and how AI agents, tool calling and the CRM run as one system.
Read the guide
Meta grades your number on how people react to it: blocks, reports, reads. The playbook from 15M+ broadcasts - never buy lists, always give an out, watch the ratio, and send utility messages before you ever send marketing.
Read the guide
Most enquiries that go quiet were never going to reply to silence. The 24 hour window, the three Meta template types (utility, marketing, authentication), business verification, and the approval wait - everything it takes before automated follow-ups can send.
Read the guide
Most WhatsApp AI agents can talk about appointments; very few can make one. The full architecture of a booking agent on Cliniko: six narrow endpoints, one source of truth, the production rules that keep a live calendar safe, and why a purpose-built API replaced make.com.
Read the build story
Both answer WhatsApp messages automatically. One walks customers through a menu; the other holds the conversation, checks the calendar, and books the job. A side-by-side comparison, where each one fits, and a 60-second test to tell what you're talking to.
Read the post
Veneer patients message four clinics at 11pm and book with whoever made it easiest. How a veneers practice runs its WhatsApp with AI employees: instant answers, follow-ups that bring price-shoppers back, reminders that protect the appointment book, and review requests timed to the happiest moment of the journey.
Read the post
An AI that knows your policies but not your records spends its day saying "let me check with the team". What HTTP Request steps are, what they cost on respond.io, what changes when your AI can finally look things up, and how we build them into service-business deployments.
Read the guide
Meta is rolling out the BSUID, a new identifier for every WhatsApp user, and letting customers hide their phone number behind a username. What changes, what quietly breaks if you do nothing, and how to prepare across GHL, respond.io, Wati, SleekFlow, Twilio, Bird and 360dialog.
Read the guide
Six months with Singapore's premium window-tint specialist. Daily installs went from 4 to 6 up to 9 to 12 cars, customer acquisition cost dropped 38%, and online reviews grew from 400 to 1,060+, all without expanding the team.
Read the case studyMore posts coming. We publish when we have something concrete to say, not on a content calendar.
Tell us about your WhatsApp setup. We'll tell you where the real leverage is.