Build vs buy a CRM for your service business: a test you can run this week

Build vs buy CRM for a small business: what's the short answer?

Buy first. A packaged field service app gives a new or small company a working system in an afternoon, a sensible default workflow, and a vendor that handles hosting, security patches and phone support. Jobber's pricing page showed single-user Core at $49 with no commitment, or $29 a month when paid yearly, when we read it on October 31, 2026. Housecall Pro's pricing page listed Basic at $59 a month on annual billing. A custom build can't match that for speed or year-one cost, and it shouldn't try.

Building becomes the better answer later, and only for some companies. The generic build-vs-buy articles you will find are written for sales teams buying Salesforce seats. A trades business has different pressure points: crews in trucks, recurring routes, deposits, card fees, texting from a shared number, and payroll rules that vary by state. This guide ties the decision to those, using limits the vendors publish themselves.

When is buying clearly the better choice?

Say yes to buying if most of these describe you:

• You are under two years old. You don't yet know how you'll quote at volume or what your recurring work will look like. Software with opinions teaches you a workflow, and you can rebuild later with real knowledge.
• Your work follows the standard loop. A customer asks, you quote, they approve, a crew does the job, you invoice, they pay. Every major product is built around exactly that.
• You have one brand and one price list. Multi-brand setups are where packaged tools strain. One brand rarely does.
• Your team is small and stable. Per-user pricing hurts most when headcount grows fast. A crew of three that stays three pays a modest, predictable bill.
• You need it now. Housecall Pro and Jobber both offered 14-day trials with no card in October 2026. A build takes weeks after discovery.
• You don't want to manage a vendor relationship around code. With packaged software, you never have to think about who maintains it.

If that is you, pick a product, use it hard for a year, and keep a running list of what it can't do. That list is the best input you'll ever have for a future build, and it may stay short. For product choices by trade, start with our HVAC software guide or the field service software pricing roundup.

Is it worth building a custom CRM?

It is worth it when the cost of not fitting is large and permanent. The signs we see most often:

• Shadow systems. A spreadsheet for deposits, a notebook for door-to-door leads, a group text for route changes. Each one is a feature the CRM is missing, and each one leaks money in missed follow-ups and double entry.
• Several businesses under one owner. Separate brands, numbering, price lists and payment accounts, with one owner who wants to see all of them at once.
• Billing the product treats as an edge case. Subscriptions priced from your website, payment schedules on large jobs, card surcharges on some payment types, or recurring plans that re-price when the customer adds a dog or a second property.
• Sales models off the standard path. Canvassing, yard signs, outbound call lists, or a pipeline with stages your quote screen doesn't have.
• A hard requirement to own the data. Some owners want the database in their own name, with direct access for reporting, and that is a legitimate business requirement on its own.

The outgrown-your-CRM checklist goes deeper on these patterns.

Which vendor limits should shape the decision?

These are the published limits that most often push an owner toward building, read on the vendors' own pages on October 19, 2026. None of them makes the products bad. They tell you where the ceiling is.

Sources: Jobber pricing, Housecall Pro pricing, Housecall Pro API overview, Workiz pricing, Jobber job import, Jobber client export and ServiceTitan pricing, all read October 2026. More detail by plan is in our API access by plan guide.

When should you build your own field service software?

Run this test. Score one point for each statement that is true today, not one you expect to be true someday.

1. Your office keeps at least two spreadsheets or side apps to cover gaps in the CRM.
2. You run more than one brand, or more than one service line with its own pricing and crews.
3. You have asked your vendor for the same missing feature for more than a year.
4. You would need the top plan mainly for API access or user count, not for its features.
5. Your expected team size in three years puts you in a much higher price tier.
6. Your process is settled: you could describe how a job moves from first call to paid in detail.
7. You can give a builder several hours of your time over the first month.

Zero to two points: buy, or stay where you are. Three or four: price both sides carefully. Five or more: a build is worth a serious quote. Points six and seven matter most. A build without a settled process or the owner's time tends to go badly, no matter who builds it.

How do you compare five-year cost honestly?

One year is the wrong window, because a build front-loads its cost and a subscription spreads it. Use five years, and fill in four lines for each side.

• Software. For buying: the plan at your expected team size at the regular price, ignoring any first-year discount, plus every add-on you would use, times 60 months. Jobber's page on October 31, 2026, for example, showed Grow for up to ten users at $299 a month on annual billing, or $17,940 over 60 months by our arithmetic. For building: the quoted build, plus hosting and the maintenance arrangement for 60 months.
• Payments. Card fees exist either way. Housecall Pro listed card fees starting at 2.59% and 1% on bank payments. Compare the rates you would actually pay through each setup, including your own processor account on a custom system.
• Switching. Data import, training and a month of running two systems side by side. This cost applies to any move, including between two packaged products.
• Workaround hours. Office hours per week spent on double entry and side spreadsheets, times a loaded wage. This line is usually where the answer flips, so measure it rather than guessing.

Our software rent calculator handles the subscription side, and the cost by crew size guide shows per-user math. If buying still wins, buy. That is a good outcome, and you saved yourself a project.

If you build, what should the agreement say?

• Ownership. The code repository, hosting account, database and domain are in your company's name from day one.
• Payments in your account. Card and bank payments run through a processor account you own, so saved cards don't belong to the software vendor.
• Backups. How often, how long they are kept, and how a restore works. Ask to see one.
• Maintenance. Who fixes what, how fast, and what it costs. Browsers, payment APIs and phone carriers change every year.
• Handover. Documentation good enough that another developer could take over.
• Migration. Which records come over from your current software, and how you check them.

Our own service companies were the first customers for a CRM we built. The case study describes one app running three businesses with quotes, routes, payments, texting and payroll exports, and nightly database snapshots kept for 30 days. Through our custom CRM service, the database, code and domain belong to the client from day one, every engagement starts with a free written quote, there is no public price list, and no per-user fees apply. Software alone can't promise revenue, and neither do we. When the test above says buy, we will say so too.

For the head-to-head pages, see custom CRM vs Jobber and custom CRM vs Housecall Pro vs ServiceTitan. Whichever way you go, the switching guide covers moving your history safely.

Frequently asked questions