I Set Up a CRM Before I Had a Pipeline. Here's Why Attio.

Part 5 of a short series on formalizing my consulting business. Part 4: setting up the books.

I have three relationships that might become engagements: a friend whose VA claims practice I've been helping for months, a childhood friend who runs a behavioral services firm I moved off paper and onto Notion, and my own CPA, who I'm trying to convince to let me modernize his practice. That's not a pipeline. It's three people I text.

I set up a CRM anyway. Here's why, and why it was Attio.

Why now

Because "I'll set up a CRM when I have clients" is how you end up with a year of decisions scattered across texts, email, and meeting notes you never wrote down. The whole value of a system of record is that it existed before the thing you needed to look up.

Also — and this is the real reason — one of those three friends needs a CRM himself, and I'm advising him on it. If I'm going to recommend a tool, I'd better be running it.

Why Attio

The shortlist was HubSpot, Notion-as-CRM, and Attio. Attio won on four things:

• The data model is honest. Companies, People, Deals are objects. Lists are views onto those objects. That sounds like a nerdy distinction; it's the thing that makes everything else work (more below).
• Free tier that's actually usable for one person. Gmail and Calendar sync in, so every email and meeting with a contact is already on their record without me doing anything.
• Granola feeds it directly. I wrote in May about why Granola won my note-taker bake-off. Now every call in my "Client calls" folder syncs to the matching company in Attio automatically. Phone call ends, structured notes appear on the record. That was the moment the whole stack clicked.
• Claude can write to it. Attio has a native connector. I say "create the company, link this person, open a deal, add these three follow-ups from Tuesday's call," and it's done. No form-filling. The CRM stopped being a chore the day I stopped typing into it.

A CRM you have to type into is a CRM you'll stop using in six weeks. The bar now is: notes and email flow in automatically, and you talk to it to make changes.

The lesson that took an hour to learn

I turned on Deals, then also made an "Active Engagements" list — and immediately had two competing sets of stages. One on the deal, one on the list. Same client, two statuses, guaranteed to disagree within a month.

Here's the distinction I was missing. A field on an object (like deal stage) belongs to that record everywhere it appears. A field on a list exists only inside that list. So stage belongs on the Deals object, once, and it should run through delivery, not stop at "closed": Lead → Discovery → Proposal Sent → Active → Completed / Lost. Lists are for things that aren't deals — referral partners, people I'm nurturing informally.

I deleted the redundant list. One pipeline, one truth.

What's in it right now

Three companies, a handful of people, two deals in early stages and one marked Won (the friend I've been coaching, whose engagement was real before the LLC was). Every record has the Granola notes from our calls and the email thread behind it. And Clarent itself has a record — a fully filled-out company profile with the stack and how the workspace is organized — because it doubles as the demo I'll show a prospective client. "Here's how I run my own operation" is a better pitch than a slide.

The relationship-first thing

You'll notice all three leads are people I already know, and all three got real help before anyone discussed money. That's not an accident and it's not softness. For a new consultancy, the cheapest way to prove the work is to do the work, and the people most likely to let you are the ones who already trust you. The CRM's job is to make sure the trust turns into a documented engagement instead of a favor nobody can invoice.

If you're a solo operator picking a CRM, ask me what I'd check first — the objects-versus-lists thing above is the question I wish someone had asked me. And if you're on HubSpot or Notion and happy, tell me why. I'm still early enough to be wrong.