CRMSmall BusinessAI

Do You Actually Need a CRM?

JUL 16, 2026·10 min read·Alex Boquist

The honest answer to 'do I really need a CRM for my small business,' written by someone with no CRM to sell you. What the CRM promised, why the data entry betrays it, the cases where a classic CRM is still the right buy, and what to demand instead.

Here's the answer up front, because you deserve one. You need what the CRM promised: a current record of every relationship, and follow-through that does not depend on anyone's memory. You do not need the data entry. And with most CRMs, the data entry is the product.

That distinction is the whole article. The promise and the mechanism are two different things, and almost every "do you need a CRM" guide on the internet quietly fuses them, so that wanting the record means accepting the typing. You can want one without the other. I'd argue you should.

First, notice who wrote all the other answers

Search this exact question and look at the bylines on the top results. HubSpot. Salesforce. Zoho. Pipedrive. Every one of them concludes, after a thoughtful weighing of the evidence, that yes, you probably do need a CRM, and wouldn't you know it, there's a free trial right here.

Asking a CRM vendor whether you need a CRM is asking a barber whether you need a haircut. The conclusion was written before the question was.

So let me put my own bias on the table. I'm a founder building software for service businesses, and I think the CRM category is ending. Not "needs a fresh coat of AI," ending. That's a bias too. But at least it's a different one, and it means I have no reason to talk you into a database you'll resent. What follows is the case as honestly as I can make it, including the situations where you should ignore me and buy the classic thing.

What the CRM was actually for

Strip away thirty years of feature lists and the CRM promised three things:

Never lose the thread. Every lead followed up, every client remembered, nothing slipping because someone got busy.
Know your pipeline. How many people are in each stage, where they stall, what's likely to close.
Institutional memory. The knowledge of your clients living in the business, not in one person's head, so it survives a vacation or a resignation.

I want to be clear: this promise was right. It's still right. If your business runs on relationships, you genuinely need all three of those things, and the businesses that have them beat the businesses that don't. The people who tell you "you don't need a CRM, just hustle" are selling you a memory you don't have.

The problem was never the promise. The problem is the mechanism every CRM used to deliver it.

Testimony versus evidence

Every classic CRM is built on the same mechanism: it asks humans to testify about the work. Log the call. Update the stage. Fill in the field. Set the next task. The record is a sworn statement, written after the fact, by the busiest person in the building.

And testimony has three well-known problems. It forgets: the call that never got logged on a slammed Tuesday. It flatters: the deal marked "negotiation" because the rep is optimistic, not because anything happened. And it rots: it was true the day it was typed and the world moved on without updating the field.

This is not a discipline problem you can fix with a better Monday meeting. The numbers on it are grim and remarkably stable. Salesforce's own State of Sales research finds that sales reps spend less than 30 percent of their time actually selling, with the rest going to administration, meetings, and manual updates (Landbase, summarizing Salesforce's State of Sales research). And even when the typing gets done, the record decays underneath it: MarketingSherpa's research puts B2B contact data decay at 2.1 percent per month, about 22.5 percent a year (HubSpot's database decay tool, citing MarketingSherpa). A survey of CRM administrators puts it higher, around 34 percent annually, with 31 percent of admins saying bad data costs their organization more than a fifth of annual revenue (ZoomInfo, citing Validity's State of CRM Data Management).

Read those together and the picture is: you pay a real slice of your team's week to produce a record that is partly fiction on the day it's written and a quarter wrong within a year.

Meanwhile, the actual truth about your business already exists. The call happened. The email thread is right there. The booking is on a calendar. The payment cleared. Evidence, not testimony. Every fact the CRM begs your team to type is a restatement of something a system already witnessed. The CRM's original sin is that it ignored the evidence and built the record on the sworn statements instead.

The honest cases where a classic CRM is the right buy

Here's the part the vendor articles skip and I won't: there are businesses for which the traditional CRM is genuinely the correct purchase.

High-volume transactional sales teams. If you run twenty reps through hundreds of short-cycle deals a month, the CRM is your assembly line: queues, dialers, territories, forecasts rolled up for a sales manager. The relationship with any one buyer is shallow and brief, so the record doesn't need to be deep, it needs to be uniform. Buy the CRM. This is what it's actually good at.

Deep integration needs. If your business already runs on a big ecosystem, with billing, support, and marketing tools that all expect to plug into one established hub, the boring choice is often right. Gravity is a real feature.

Businesses that don't run on relationships. E-commerce, one-off transactions, anything where the customer arrives, buys, and leaves. You need analytics and maybe an email tool. A CRM would be an expensive address book.

And one more: if your team already keeps a CRM current and it works, keep it. Rare, but real. Discipline that already exists is not a problem to solve.

If your business runs on relationships, demand more

Now the other side. If you're a coaching practice, a law firm, an agency, a consultancy, a health practice, if your revenue comes from people you serve for months or years and a dropped ball costs a client, then the calculus is different. You need the promise badly, and you can least afford the mechanism. Your best people's hours are the product. Spending them on typing up what already happened is paying your highest rate for your lowest work.

So here's what I think you should demand instead, and none of it is science fiction. I know because we build it, and I watch it run inside real firms every day.

A record that keeps itself. The emails, the calls, the bookings, the payments already say what happened. Software should read that stream and derive the record from it: who this client is, what stage they're in, what happened last, what's overdue. In Funal, stages and records are computed from real events, and one pattern from real deployments still surprises people: when staff stop updating records by hand, the data gets more accurate, because it stops depending on anyone's discipline.

Follow-through that arrives drafted. A reminder that says "follow up with Sarah" still leaves you the hard part. Demand the draft: the actual follow-up, grounded in Sarah's real history, waiting for your approval. You stay in charge of everything that goes out. You just stop starting from a blank page.

Numbers nobody typed. When every number on your scoreboard is computed from events instead of testimony, you can trust it enough to act on it. That extends all the way to marketing: because the click, the form, the booking, and the payment live in one record, you can trace which ad produced clients, not which ad produced clicks. And when your funnel is a chain of real numbers, it can do the one thing a dashboard never could: point at the weakest link, the single place effort pays off this week.

That's not a better CRM. It's a different premise. The old category asks your team to maintain a picture of the business. The new premise is that the business paints its own picture, and your team just reads it.

So, do you actually need a CRM? You need the record. You always did. But the end of the CRM era is not the end of the record. It's the record that keeps itself. The filing cabinet was never the business. Being able to see the business is.

FAQ

Is a spreadsheet enough?

Sometimes, briefly. Under maybe twenty active relationships and one person doing the following up, a spreadsheet plus a calendar genuinely works, and it's free. But a spreadsheet is pure testimony, the most decay-prone record there is, and it does no follow-through at all. The moment a second person needs the picture, or a lead slips because nobody was watching, you've outgrown it. Treat the spreadsheet as a starter home, and notice that the day you outgrow it, "a CRM" is no longer the only next step.

When is a CRM worth it?

When your sales motion is high-volume and transactional, when you need to plug into a big existing ecosystem, or when your team has proven it will actually keep one current. In those cases the classic CRM is a fine machine doing the job it was built for. Buy it without guilt.

What replaces a CRM?

For relationship businesses, a system that derives the record from evidence instead of collecting testimony: it watches the real events (email, calls, bookings, payments), keeps every client record current on its own, drafts the follow-through for a human to approve, and computes the numbers so nobody has to trust their own typing. Funal is our answer to that description, and it's why I don't call it a CRM. The honest generic advice: whatever tool you evaluate, ask one question first. If my team never typed another update, would this record still be true tomorrow? If the answer is no, you're buying the old mechanism with a new coat of paint.

I run Funal, so read my conclusions the way you'd read anyone's: as an argument, not a verdict. The industry figures above come from the public sources cited below, and the claims about Funal describe what runs in production today, nothing more.

Sources

Back to Journal