Why most CRM projects disappoint

Almost every growing business reaches the same point. Leads arrive by email, phone, web form and word of mouth. Someone keeps a spreadsheet, someone else keeps it in their head, and follow-ups depend on memory. A CRM is the obvious fix, so the business picks one, imports a contact list, and six months later the real pipeline is back in a spreadsheet.

The software is rarely the problem. The problem is that the CRM was chosen by feature list rather than by how the business actually sells and delivers. A CRM only works if it matches the way your team already works closely enough that using it is easier than not using it.

This guide is how we approach the decision for clients, and for our own businesses.

Start with the process, not the product

Before you look at a single product, write down how a customer moves through your business today. Keep it plain:

Where do enquiries come from? Website forms, phone, email, referrals, marketplaces, walk-ins.
Who picks them up, and how quickly? Is there an owner for every new lead, or does it depend on who notices first?
What has to happen before someone becomes a customer? A call, a quote, a site visit, a proposal, a deposit.
What happens after they say yes? Onboarding, scheduling the work, invoicing, follow-up.
What do you need to see at a glance? What is in the pipeline, what has been promised, what is overdue.

That list is your requirements document. It will tell you more about the right CRM than any comparison table, and it often shows that the real gap is not software at all but unclear ownership of each stage. If nobody owns a lead, no tool will follow it up.

The five questions that decide the choice

1. How complex is your sales process?

A business that sells one service through a short conversation needs a simple pipeline: new, contacted, quoted, won, lost. A business with long sales cycles, several decision-makers or multiple product lines needs more structure. Buy for the process you have, not the one you imagine in three years. Most CRMs let you add complexity later; very few make it easy to remove.

2. Where does the rest of your business live?

A CRM that does not talk to your email, calendar, forms and invoicing becomes another place to type things. List the tools it must connect to and check each integration properly, not just whether a logo appears on the vendor's website. Where there is no native integration, a workflow automation tool such as n8n can usually fill the gap.

3. Who will actually use it every day?

The people updating the CRM decide whether it works. If your team is on the road or in a kitchen, mobile use matters more than reporting depth. If only one person will use it, a lighter tool will serve you better than a platform built for a sales floor.

4. Who owns the data?

Check how easily you can export everything, including history and notes, in a format another system can read. A CRM you cannot leave is a CRM that can raise its prices on you.

5. What will it really cost?

Look beyond the headline per-user price. Many CRMs put the features you actually need, such as automation, custom fields or reporting, in higher tiers. Add the time to set it up properly and to migrate your data. Our guide to budgeting for technology projects covers the hidden costs that apply to any system, not just AI.

Buy or build?

For most small businesses the honest answer is buy. A well set-up off-the-shelf CRM is cheaper, faster and better supported than anything built from scratch, and there is no competitive advantage hiding in how you store a contact record.

Building makes sense in a narrower set of cases:

Your process is genuinely your own. Bookings, jobs, allocations or compliance steps that an off-the-shelf tool would have you working around every day.
The CRM is really part of a bigger system. Customer records that need to sit alongside scheduling, stock, or a client portal in one place.
You are paying for a large platform and using a fraction of it. Sometimes a focused internal tool does the job better at a lower running cost.

We go through the full decision in build vs buy, and the same logic applies whether the tool involves AI or not.

Setting it up so people actually use it

Choosing the tool is half the job. The other half is set-up, and this is where most CRM projects quietly fail.

Mirror your real stages. Use the words your team already uses. If everyone says "quoted", do not call the stage "proposal sent".

Keep required fields to a minimum. Every mandatory field is a reason not to log a lead. Capture what you need to follow up and report; nothing else.

Automate the typing. Website forms, booking tools and email should create and update records automatically. The less anyone has to enter by hand, the more accurate the CRM stays.

Migrate carefully. Clean and de-duplicate before you import, and check the counts match before anything old is switched off.

Make it the only place. If the spreadsheet survives alongside the CRM, the spreadsheet will win. Retire it on a set date.

Signs your current CRM is the wrong fit

The team keeps a separate spreadsheet "just for now".
Nobody trusts the pipeline figures in the weekly meeting.
Leads go quiet and nobody notices until the customer complains.
You pay for features nobody has switched on.
Getting a simple answer, such as what is overdue, means asking three people.

If several of these sound familiar, the fix is usually a mix of process and set-up rather than a new product. Sometimes it is both.

Where to start

Write down your process using the five steps above, list the tools the CRM must connect to, and be honest about who will use it. That alone will narrow the field to two or three candidates.

If you would like a second pair of eyes, we set up and build CRM and operations systems for businesses that have outgrown spreadsheets, and it usually starts with a free 30-minute call about how you sell today. For the wider picture of where time and money leak across the business, see where small businesses lose time and money.

How to Choose a CRM for a Small Business (and When to Build Your Own)