Do you need a CRM? An honest checklist

Ask a CRM company whether you need a CRM and the answer is yes. Ask them when, and the answer is now. That answer comes from a pricing page, so here is the version nobody selling software will give you: the actual signs you’ve outgrown your inbox, the signs you haven’t, and what to use at each stage, including the stages where the right answer costs nothing.

First, the term without the mystique. A CRM (customer relationship management system) is a shared record of every person who might pay you or has paid you, what’s been said, where each conversation stands, and what’s supposed to happen next. That’s the whole machine. Everything else is features.

Which means the question “do I need a CRM” is really the question “is my current way of remembering all that starting to drop things?” For a lot of businesses the honest answer is yes and they can name the dropped thing. For plenty of others the answer is not yet, and buying software anyway just adds a login to ignore.

Signs you need one now

Go through these and count.

A lead went cold because nobody followed up. Lost to silence rather than to a competitor’s better offer. Someone asked for a quote, you meant to get back to them, and by the time you did (or didn’t), they’d hired someone else. If you can name one of these from the last few months, there are probably others you never noticed. This is the single clearest sign, because follow-up is exactly the thing a CRM exists to make impossible to forget.

You can’t answer “how many open deals do we have?” without reconstructing it. If the answer lives across an inbox, a notes app, two phones, and someone’s memory, then nobody actually knows what’s in the pipeline, which means nobody knows next month’s revenue, which means every business decision is being made on vibes.

More than one person talks to customers. The moment a second person answers inquiries, “it’s all in my head” stops being a system. Customers repeat themselves, two people answer the same email differently, and handoffs happen by shouted summary. A shared record is the only way two people can act like one company.

Follow-up runs on inbox flags and mental notes. Starred emails, “I’ll remember to check in Thursday,” a sticky note on a monitor. Each one works until the week gets busy, and the weeks when you’re busy are precisely the weeks leads pour in. The system fails exactly when it matters.

Past customers never hear from you again. If repeat business or referrals matter in your trade, and your old customers exist only as buried email threads, there’s revenue sitting in a list you technically own and practically can’t use. Before you start emailing that list, check that your domain meets Gmail’s and Yahoo’s sender rules.

Losing your phone would mean losing your pipeline. If your customer knowledge lives in one device or one skull, your business records are a single point of failure with a data plan.

Three or more of these, and you’re past due. One or two, keep reading, because the fix might not be software yet.

Signs you don’t need one yet

The vendors skip this section.

You get ten or fewer leads a month, and you’re one person. That volume is not a data-management problem. Most CRMs at this volume cost more attention than they save, and an untended CRM is worse than none, because a pipeline full of stale entries lies to you.

Your sales close in one interaction. Walk-in retail, single-visit services, jobs quoted and won on the spot. There’s no follow-up to drop, so a follow-up machine solves nothing. Your leverage is elsewhere (being found in the first place, which is a different problem).

You already run a sheet that works. A spreadsheet with a row per lead, a status column, and a next-action date, actually kept current, is a CRM. It’s a manual one, and at low volume, manual is fine. Don’t let anyone make you embarrassed of a system that isn’t dropping anything.

You’re hoping software will impose discipline you don’t have. It won’t. If leads go unanswered because nobody’s assigned to answer them, a CRM gives you a beautifully organized list of unanswered leads. Software records a process; it can’t invent one.

What to use at each stage

Ten or fewer leads a month, one person: the spreadsheet. One row per lead, columns for name, contact, what they want, status, and next action with a date. Look at it every morning. Total cost: zero, plus the discipline of the morning look. When you want the totals on one screen and the sheet lives in Google Sheets, a free Data Studio report can connect to it. Our Looker Studio vs Power BI comparison covers the options.

Growing volume, or a second person: an off-the-shelf CRM, and start with a cheap or free tier. Before you pick one, do the exercise that matters more than the pick: write down the stages a customer actually moves through in your business, from first contact to paid, in your own words. That list of stages is your pipeline. A tool configured around your real stages beats a fancier tool running its defaults. Expect to spend an evening setting it up and a few weeks making its use a habit, and expect the habit to be the hard part.

Established process, real volume, tools starting to sprawl: this is where it gets interesting, because the payoff stops being memory and starts being motion. New inquiry creates a record on its own, the intake response goes out in minutes, the quiet lead gets surfaced automatically, the invoice follows the closed job without anyone remembering to send it. We covered what’s realistically automatable in what AI can actually automate in a business; the boring rule from that piece applies doubly here: automate the repetitive, keep humans on judgment.

One honest warning that applies at every stage: plenty of CRM rollouts die quietly. The standard failure isn’t picking the wrong tool; it’s the third week, when entering leads starts feeling like homework and the team quietly goes back to the inbox. From then on the CRM shows a partial picture, a partial picture can’t be trusted, and an untrusted system gets abandoned by everyone else too. The defense is unglamorous: make the CRM the only place a lead can live, and have one person own keeping it true. If nobody will own that, save the subscription fee, because you already know how the third week goes.

A word on custom-built CRMs, since we build them: most businesses don’t need one, and we’ll say so when asked. Off-the-shelf covers standard sales pipelines well. Custom earns its cost when your process genuinely doesn’t fit the templates, or when the CRM needs to talk to systems the off-the-shelf tools won’t, and it should be a decision you arrive at after outgrowing something, never a starting point.

The checklist, compressed

The six signs: leads dying of silence, a pipeline nobody can count, multiple people flying blind, follow-up on sticky notes, past customers unmined, everything on one phone. Count your yeses. Zero to one: run the sheet and revisit in six months. Two: run the sheet, and let a second person answering inquiries be the trigger that moves you to software. Three or more: pick a tool this month, configure your real stages, and go get the leads that have been dying quietly.

And if you’re at the stage where the records exist and the motion is what’s missing, that’s the work we do: the pipeline, the follow-up, and the plumbing between the tools, built around how your business actually runs.