Automation
WhatsApp enquiries into the CRM, on their own
Meta ad / form / WhatsAppDe-duplicate, tag sourceCRM + assign ownerFirst reply in under a minute
BeforeOne person copying every enquiry into a sheet, then the CRM, then typing the reply. Evening leads wait till morning.
AfterEvery lead in one list with its source. First reply automatic. A 9 am summary for the founder.
The task. A sales team receiving enquiries from Meta ads, the website form and WhatsApp. One person copies each one into a Google Sheet, then into the CRM, then sends a first reply. Leads arriving after 7 pm wait until the next morning, and a few slip through every week.
What gets built. An n8n workflow with the official WhatsApp Business API. Every enquiry lands in one place, is de-duplicated, tagged by source and assigned to the right person. A first reply (approved template, in English or Hinglish) goes out within a minute. A daily summary reaches the founder at 9 am.
What changes. Nobody types the same reply forty times a day. The team sees every lead in one list, with the source. The "who replied to this?" question stops.
Time to live. Around ten working days, most of which is waiting for Meta's WhatsApp approval and testing edge cases. Runs on the client's own n8n and Meta accounts.
Automation
The Monday report that writes itself
Monday 8 am triggerSheets + ads + CRM pulledAI writes the summaryWhatsApp + email to the founder
BeforeTwo to three hours every Monday, three tabs open, numbers that never quite match last week's.
AfterReport waiting at 8 am, same definitions every week, and a warning instead of a wrong report if a source fails.
The task. Every Monday someone spends two to three hours pulling last week's numbers from Google Sheets, the ads account and the CRM into one document. The founder reads it for two minutes.
What gets built. A scheduled n8n workflow that pulls each source at 8 am, calculates the week-on-week changes, and has an AI model write a short plain-English summary with the numbers attached. It goes to the founder on WhatsApp and email. If any source fails, the workflow tells you instead of sending a wrong report.
What changes. Monday morning is back. The report is more consistent than the hand-made one, because the definitions never drift.
Time to live. About a week, once the data sources are agreed.
App
From a shared spreadsheet to an internal dashboard
One sheet, five editorsScreens agreed on paperWeb app + Android appOne source of truth
BeforeFormulas break, rows go missing, nobody is sure which version is right.
AfterLogins, roles, history, the reports the sheet never managed, and the field team updating from their phones.
The task. A service business running orders, staff assignments and payments on one Google Sheet edited by five people. Formulas break, rows go missing, and nobody is sure which version is right.
What gets built. A web app with logins and roles, a proper database, change history, and the reports the sheet was never able to produce. An Android app for the field team to update jobs from their phone. Hosted on accounts in the client's name, code in their GitHub.
What changes. One source of truth. The owner can see the day's status without asking anyone. New staff learn the app in an afternoon.
Time to live. Longer than an automation, typically four to eight weeks depending on the scope. The written scope says exactly what is in the first version.