Short answer: A CRM is not a contact list. It is the operating model of your company written down: who your customers are, what stage they are in, what happens next and who owns it. I have been building them for a decade, and the lesson that has held across every tool is this: design the pipeline first, then pick the software.

The mistake I made first

My first CRM was a spreadsheet with ambitions. Names in one column, a status in another, notes wherever they fit. It worked until it didn’t, which is the life story of every spreadsheet CRM. The day a second person needed to use it, it broke, because the rules only lived in my head.

That was the real lesson. The software was never the problem. The problem was that the business had no written-down definition of a customer’s journey. Every tool I have used since, from purpose-built CRMs to all-in-one marketing platforms to relational databases I designed myself, has only been as good as that definition.

Design the pipeline before the tool

Before I open any software now, I answer five questions on paper.

  1. What are the stages? Not “lead, prospect, customer.” The real ones, in the customer’s words. What do they believe at each step, and what do they need to move?
  2. What moves someone forward? One clear event per stage change. A signed agreement. A paid invoice. A completed form. If you cannot name the event, the stage is fake.
  3. Who owns each stage? One name per stage, never “the team.”
  4. What happens automatically? The confirmation, the reminder, the follow-up after silence, the internal alert when something stalls.
  5. What must a human always touch? The moments where tone, money or judgment matter.

Answer those and the tool almost picks itself. Skip them and no tool will save you.

The automations that actually matter

After a decade of building these, I have a short list of automations I put in every pipeline, whatever the business.

  • Instant acknowledgment. Every inbound inquiry gets a real, specific response within minutes. Speed to first reply is the single most underrated conversion lever in services and commerce alike.
  • The silence trigger. If a thread goes quiet past a set number of days, the system drafts a follow-up and puts it in front of the owner. Most lost deals are not lost. They are forgotten.
  • Stage-change handoffs. When a customer moves stages, the next owner is notified with everything they need, in one message, so nobody has to ask “where are we with this?”
  • The weekly exceptions report. Not a dashboard of everything. A list of what is stuck, overdue or missing, with a name next to each item.
  • Clean data at the door. Required fields, consistent formats and duplicate checks at the moment of entry, because cleaning data later costs ten times as much.

Where AI changes the CRM

For years the CRM’s weak point was the human typing into it. Notes were thin, fields were skipped, follow-ups depended on someone remembering.

AI changes that. Agents now read the inbox, attach the conversation to the right record, update the stage when the trigger event happens and draft the next message in the right voice. The CRM finally reflects reality without anyone doing data entry.

But the same rule applies as everywhere else in my companies: agents draft, people decide. An agent can prepare the follow-up. A person sends it. An agent can flag a billing question. A person answers it.

What I would tell a founder setting one up today

  • Write your pipeline on one page before you buy anything.
  • Choose a system that lets you own your data and connect it to everything else. A CRM that cannot talk to your database, your task system and your inbox will become an island.
  • Automate the reminders, not the relationships.
  • Review your exceptions every week. The pipeline tells you the truth about your business if you let it.

Ten years in, I still think the CRM is the most honest document a company has. It shows exactly how you treat people, step by step. Build it like it matters, because it does.

Lo Dixon builds ecommerce brands and the technology behind them, including CRMs, automation and AI agent systems. She has spent a decade in technology, building companies and the systems that run them. More at lodixon.com.