Small-business CRM that starts from your inbox

Small businesses buy CRM the way they buy gym memberships: after a messy month, with a demo that looks like a command center, and a quiet drop-off when nobody has time to log calls. The deals did not move to the new software. They stayed in Gmail, because that is where customers write.

A CRM that starts from the inbox is not “anti-CRM.” It is the honest sequence. First make the live conversations visible as a pipeline. Then, if you hire a salesperson or a marketer who needs a database, you can graduate. Starting with the database and hoping the inbox will sync is how you get two sources of truth, both wrong.

What a small-business pipeline actually is

It is not a funnel graphic. It is a shared answer to: who are we talking to, what is it worth, when might it close, and who owns the next action. In a five-person company, “who owns it” is often the founder’s Gmail. Pretending otherwise with seats and territories does not create a sales team. It creates empty required fields.

If two people sell, you still may not need a full CRM on day one. You need each person to see their own board, and a weekly conversation about the Negotiation column. Shared inboxes and permission matrices can wait until you are actually stepping on each other’s threads.

Inbox first, then the rest of the stack

Invoicing, support, and delivery can live elsewhere. Many small firms already have Holded, Holded-like tools, Stripe, a shop, or a project board. Those systems are good at money after the yes, or work after the yes. They are weak at “this quote is five days old and nobody poked.”

That poke lives in email. So the sales object should be a card tied to a thread: company, value, close date, next step, notes. When the card is Won, hand off to whatever you use to deliver. Do not force the delivery tool to also be the pipeline. Mixed boards become junk drawers.

Labels in Gmail are not this. A label says the thread belongs to a customer. It does not say stage, money, or next step. Nested labels like Sales/Negotiation still make you open five threads to see the month. A board is the list you can read without reconstructing.

A setup a founder can finish this week

1. Pick five columns. Lead, Contacted, Negotiation, Won, Lost is enough. Rename if you sell retainers or projects with a legal pause.
2. Take the last 30 days of mail, not five years. Create a card only for threads that are commercial. Skip suppliers and newsletters.
3. Add value when you know it. Add a close date you would defend in a meeting. Write a next step in one sentence.
4. Agree a kill rule: 14 days with no next step goes to Lost or On hold. Otherwise the board dies of politeness.
5. Start the morning on the board twice a week, even if the rest of the days start in the inbox.

Do not import CSV from a previous tool until this board has been true for two weeks. Migration is how small teams postpone the habit. Ten honest cards beat 400 stale contacts.

When you should not start in the inbox

If a parent company already standardized on Salesforce or HubSpot, use their Gmail plugin and stop fighting. The system of record is political. A shadow board in the founder’s Chrome will not survive the first forecast meeting.

If your “sales” is inbound forms into a marketing automation tool, and nobody lives in Gmail for deals, start where the leads already are. This article is for the common case: quotes, introductions, and follow-ups in mail, with a website that is not a lead machine.

If you need mail merge, sequences, and shared boxes for a small sales pod, look at a heavier Gmail CRM (Streak is the usual one). A personal Kanban will feel too small. Be honest about the job before you pick a lighter tool because it is free.

Fields that earn their keep in a small firm

You do not need forty custom properties. You need:

• Stage — a column you drag, not a status someone forgets to change in a sidebar.
• Company — so two emails from the same account do not become two forecasts.
• Value — even approximate. Column totals are the only dashboard most founders will look at.
• Close date — a date that can be wrong. That is useful. A date that is never updated is not.
• Next step — the field that turns the board into work. See why it outranks the rest.
• Notes — procurement names, legal blockers, “invoice under the group.” Not a copy of the thread.

Phone numbers and job titles are fine if you type them when you have them. They are a bad reason to pick a CRM that demands a complete contact record before you are allowed to save. Incomplete cards that you use beat complete cards that you don’t.

Privacy, for a company that is still a mailbox

Small teams are rightly nervous about tools that want to read every email “to auto-create opportunities.” That is a lot of trust for a pipeline you could have built by choosing threads. Prefer a model where you pick the conversation and the vendor stores the metadata you typed — not a robot deciding what a deal is.

KanbanMail is built that way: you send the email to the board. We do not scan the mailbox. We do not request gmail.modify, and we do not read mail from our servers. Sign-in is Google profile so we can attach the board to you.

How KanbanMail does the inbox-first version

KanbanMail is a free Chrome extension that puts a Kanban board inside Gmail, with a Gmail add-on for the open message. You choose the threads. Cards hold company, value, close date, next step, and notes, plus a link back to the mail.

It will not replace an ERP, a helpdesk, or a marketing database. It will show a small company what is in play without a second empty tab. If you later outgrow it, you will outgrow it with a pipeline you actually used — which is the only migration that works.

For the Gmail-native method, read how to use Gmail as a CRM. If you are one person rather than a small team, the freelancer version is the same idea with fewer meetings. For stages, see a sales pipeline inside Gmail.