Custom CRM vs subscription: how to do the 5-year math yourself

I build custom CRMs, so you'd expect me to tell you custom is cheaper. Sometimes it is. Often it isn't. The only way to know is to do the sum with your own numbers, and it takes about twenty minutes with a spreadsheet. Here's how.

Why five years

A subscription spreads its cost out. A custom build front-loads it. Compare them over one year and the subscription always wins; compare them over ten and you're guessing about a decade of software you can't predict. Five years is long enough for the up-front cost to matter less and short enough to estimate with a straight face. Use three if you're unsure the business will look the same by then.

What to count on the subscription side

Seats, now and later. Count the people who'd actually log in today. Then make a sensible guess at year three. Per-seat pricing is the line that grows fastest when you hire.

The tier you'd really need. Not the tier on the pricing page you'd like to be on. The one that includes the features you'd actually use: custom fields, automation, the reporting your manager will ask for in month two. Check the vendor's current pricing yourself; I'm deliberately not quoting anyone's.

Add-ons. Extra modules, marketplace apps, a paid integration to your accounting tool, extra storage, a phone or SMS add-on.

Admin time. Somebody configures the thing, fixes the views, cleans duplicates, and keeps the spreadsheet that still exists alongside it. Put a rough number of hours a month on that, and an hourly cost.

Price increases. You can't know them. Treat them as a variable: run the sum once flat and once with a yearly increase you pick.

What to count on the custom side

The build. Use the top of your quoted range, not the bottom. My CRM builds start at $16,000, and for most scopes you should expect $16,000–$35,000, depending on integrations and user roles.

Hosting. A server or managed platform. For a small internal tool this is usually a modest monthly bill. Ask for the real figure.

Looking after it. Either a care plan (my App Care is $990/month for up to 10 hours) or changes paid hourly when you need them ($110/hour). Be realistic about how much you'll ask for. New software tends to generate a list of "one more thing".

Later changes. New features you'll want in year two or three. Budget something, even if it's a guess.

A worked example

When the subscription wins

• Small team, stable headcount. Seat costs stay low.
• Your process fits the software without much bending.
• You want features the vendor will keep adding: email marketing, a mobile app, a big integration marketplace.
• Nobody has time to be the "owner" of a custom system, even for an hour a month.

If that's you, keep the subscription. Spend the money on someone configuring it well.

When custom wins

• Seat count is high or climbing, and most of those people only need two screens.
• Your work doesn't look like a sales pipeline: jobs with technicians, parts and site visits, or cases with documents and deadlines.
• You're paying for a tier to unlock one feature.
• The real data still lives in a spreadsheet next to the CRM, and someone reconciles them by hand.
• You want the data, the code and the hosting in your name, not rented.

The things the sum leaves out

Some costs don't fit neatly in a spreadsheet. Switching later is expensive in either direction. A subscription can change its pricing or retire a feature you rely on. A custom system depends on there being someone to maintain it, which is why it should be written in a common stack with docs, so any competent developer can take over. Neither option is risk-free; they're different risks.

Try it

The calculator on the custom CRM page does the basic version of this sum with your numbers. It doesn't pre-fill any vendor's prices, because I don't want to guess at them for you.

If you'd like to go through it with someone, that's a reasonable use of the free 20-minute intro call. I'll tell you if the subscription looks like the better deal.