The short answer
Between 4,900 and 19,900 US dollars for the large majority of businesses under a hundred people, as a fixed price agreed before any code is written. Hosting and support afterwards is optional at 290 a month.
| Band | Price | What it is | Build window |
|---|---|---|---|
| Starter | 4,900 | One pipeline, one team, replacing a spreadsheet | 3 to 4 weeks |
| Business | 9,900 | Several record types, quoting, accounting sync, a migration | 6 to 8 weeks |
| Custom system | From 19,900 | Portals, scheduling or inventory modules, multi location permissions | 8 to 14 weeks |
| Aftercare | 290 a month | Hosting, backups, monitoring, up to 4 hours of changes | Cancellable |
Two things about those numbers. They are fixed, not estimates, which means the risk of a scope surprise sits with us rather than with the client, and that is priced in. And they are one time: there is no per seat fee, so a build that serves eight people and one that serves eighty cost the same. Full detail is on the pricing page.
If you have been quoted under two thousand for a custom CRM, you are being quoted a configuration of something, not a build. That can still be the right purchase. It is just not the same product.
The seven things that move the price
Quotes in this market look arbitrary because the drivers are never explained. There are seven, in order of impact.
- Number of record types. Not the biggest cost individually, but the multiplier on everything else. Each record type needs screens, permissions, relationships, search and reports. Three record types is a Starter. Eight is a Custom system.
- Integrations. Roughly 600 to 2,500 each depending on the far side. Accounting is well trodden. A twenty year old industry platform with a file based interface is the expensive end, and the price is mostly in handling its failures.
- Data migration. Between nothing and several thousand, decided entirely by the state of the source. A clean export from a known CRM is routine. Fourteen spreadsheets with inconsistent names and attachments in a shared drive is a project.
- Portals. A logged in area for customers, partners or dealers adds authentication, permissions, an audit trail and a second interface to design. Two to four weeks, and it is the most common reason a quote crosses into the third band.
- Document generation. Quotes, contracts, reports and invoices that look right on a customer's screen. Cheap for a simple template, expensive when the layout has rules or an industry standard to satisfy.
- Scheduling. Calendars, routing, availability and conflict rules are genuinely difficult, and any system with dispatch in it is a larger build than its record count suggests.
- Compliance requirements. Audit trails, data residency, retention rules, consent evidence and access logging. Not hard individually, but they touch every screen, so they raise the floor of the whole build.
Notice what is not on the list: number of users. It does not affect a build price, which is the structural difference between owning and renting.
Where the hours actually go
A Business build at 9,900, broken down the way we estimate internally. Useful whoever you hire, because it tells you which conversations save money.
| Phase | Share | What happens | What you do |
|---|---|---|---|
| Discovery and design | 15% | Record types, stages, rules and screens agreed in writing | Answer questions, in writing, quickly |
| Core build | 35% | Records, relationships, permissions, search, the daily screens | Nothing yet |
| Automations and documents | 15% | The when X happens rules, quote and report output | Provide real templates |
| Integrations | 15% | Accounting, email, calendar, phone, web forms | Provide credentials and a test account |
| Migration | 10% | Export, map, clean, import, reconcile the counts | Decide the identity and merge rules |
| Testing and handover | 10% | Real data, real users, fixes, training clips | Put a real week of work through it |
The two rows where a client changes the price are the first and the fifth. A business that answers discovery questions in two days rather than three weeks compresses the whole schedule, and a business that decides its own deduplication rule removes the most common migration overrun. Neither costs anything except attention.
The row that surprises people is core build at thirty five percent. Everyone expects the features to be the cost. In practice permissions, search and the relationships between records are most of the engineering, and they are invisible in a demo, which is exactly why configured platforms look cheap until you need them to behave.
What it costs after launch
The honest version, including the parts that are not ours.
- Hosting: 10 to 60 a month for most businesses, paid to your own hosting provider, in your own account. A CRM for forty people is not a heavy application.
- Domain and certificates: 15 to 40 a year. Certificates are free.
- Aftercare with us: 290 a month, covering hosting management, backups, monitoring, security updates, integration maintenance when an API changes, and up to four hours of changes. Cancellable, and about a third of clients eventually stop it once the system is stable.
- Changes beyond that: 60 an hour, quoted before the work.
- Nothing per seat, ever. Adding your twentieth or hundredth user costs nothing.
Two costs people forget to count, and they are real. Somebody internally has to own the system, which is a few hours a month rather than a hire. And in year three or four you will want a meaningful change, because your business will have moved, which is typically a few thousand rather than a replacement. Budget for evolution instead of pretending software is furniture.
Five years, both routes
A twenty person business, all needing access, comparing a mid tier subscription against a Business build with aftercare. US dollars, rounded, using list prices common at the time of writing.
| Subscription, 20 seats | Custom build | |
|---|---|---|
| Up front | 2,000 implementation | 9,900 fixed |
| Year 1 | 14,400 | 3,480 |
| Year 2 | 14,400 | 3,480 |
| Year 3 | 15,800 after a price rise | 3,480 |
| Year 4 | 15,800 | 3,480 |
| Year 5 | 17,400 | 3,480 |
| Five year total | 79,800 | 27,300 |
| If headcount doubles | About 155,000 | Unchanged |
The break even lands late in year one for a team this size and earlier for larger ones. Below about eight seats the arithmetic favours subscribing and we say so, which is why our smallest band exists for businesses replacing a spreadsheet rather than a platform.
What CRM pricing looks like by industry
People often search for CRM pricing for their own industry, and the honest answer is that industry matters less than shape. Still, the same industries tend to land in the same band, for reasons that are easy to name.
| Industry | Usual band | What drives the price |
|---|---|---|
| Car wash and auto detailing | Starter, 4,900 | Bookings, memberships and repeat reminders; one location |
| Salons and spas | Starter or Business | Online booking and staff rotas push it to Business |
| Nonprofits and charities | Business, 9,900 | Donations, Gift Aid or tax receipts, consent history, migration |
| Staffing and recruitment | Business or Custom | Two pipelines, candidates and clients, plus job matching |
| Public sector bodies | Custom, from 19,900 | Permissions, accessibility, audit trails and procurement paperwork |
The per seat comparison changes the picture by industry too. A detailing business with three staff gains little from avoiding seat fees, so its case rests on fit. A staffing agency with thirty consultants pays per seat on every one of them, which is usually what pushes it toward owning the system. Our recruitment, nonprofit and detailing pages show what each build includes.
For a budget CRM, the cheapest honest answer is often a free tier or a spreadsheet until the business has a second person following up. We say so in writing when that is true.
The free CRM question
Free tiers are a serious option and we recommend them regularly, so it is worth being precise about where they end.
A free tier is genuinely enough for a small team doing simple follow up, and paying anybody, including us, before that point is waste. What free actually buys is a contact list, a pipeline and basic tasks for a capped number of users or records.
The ceiling arrives in one of four ways, usually suddenly. A record or contact limit is reached, which for marketing suites can happen in a single import. A feature you now depend on turns out to be two tiers up. The number of users crosses the free cap. Or the reporting you need to run the business does not exist at the free level, and the paid level that has it is priced for a different size of company.
The genuine cost of a free CRM is not money, it is the migration you will run later plus the habits formed inside somebody else's data model. That cost is small if you keep the data clean and export it monthly, and large if you let three years of history accumulate in a tool you cannot export properly. Export monthly from day one, whatever you use, and keep the files.
Free, freemium and freeware CRM programs compared
Free CRM programs come in three kinds, and they differ more than the word free suggests.
| Kind | What you get | Where the cost appears |
|---|---|---|
| Freemium | A hosted CRM with a free tier, often no credit card required | Seat, contact or feature limits; the paid tier you need is priced per user |
| Freeware | Desktop or self hosted software, free to use | Your time: setup, backups, updates and support |
| Open source | Free code you host and can change | A developer to install, customise and keep it secure |
Free tiers with unlimited contacts usually limit users, pipelines or automated workflows instead. Freeware contact management software suits one person on one computer and struggles the day a second person needs the same list. For an inexpensive CRM, the honest comparison is three years of the paid tier you will actually need against a one time build, which is the table earlier in this guide.
How to spend less without getting less
Five things that reduce a quote, all of which we suggest before being asked.
- Cut the first release, not the system. Build the daily workflow and the one integration you cannot live without. Add the portal and the reporting in phase two, once the system is in use and you know what you actually want. A phased build is regularly thirty percent cheaper across both phases, because phase two is specified from evidence rather than imagination.
- Clean your own data. Migration is priced by the mess. A weekend spent deduplicating a spreadsheet and standardising a column can remove a thousand from a quote, and nobody knows your records better than you do.
- Drop the integration nobody uses. Every integration list we receive has one system on it that two people log into twice a year. Ask when each connection was last needed and by whom.
- Use one document template, not six. Layout variations are cheap individually and expensive collectively. Standardise first, then automate.
- Answer discovery quickly. The most expensive client is not the demanding one, it is the slow one, because a stalled project is repriced. Two days of focused answers early is worth more than any negotiation.
Two things we will not cut, and it is fair to know why. Backups and an export path, because a system you cannot leave is not owned. And permissions, because retrofitting them into a live system with real data is several times the cost of building them in.
What we will not quote
Turning work down is part of pricing honestly, so here is the list.
A CRM for a process that does not exist yet. A business six weeks from launch will change its process five times before Christmas. Rent something for a year and build on evidence.
A replacement for a system that works. If the complaint is the interface is dated but the data is sound and the team is fluent, the payback is not there. Integrate with it or report on it instead.
An ERP wearing a CRM label. If the actual problem is stock, purchasing or production, a CRM will pull the same wrong numbers from the same place. The split is explained in ERP versus CRM.
A build with no internal owner. If nobody is accountable for the system after launch, it will decay, and we would rather not take the money.
A fixed price on a scope nobody can describe. If discovery cannot produce a two page scope, the honest product is a paid discovery engagement, priced separately, ending in a document you own whether or not you continue with us.
How the money is actually paid
The structure matters as much as the number, because it decides who carries the risk.
Forty percent on signature, forty at the first working release with your real data in it, twenty at handover. The middle payment is deliberately tied to something you can log into and use rather than to a date on a plan, so a slipping project is our problem financially rather than only yours.
The fixed price covers everything in the two page scope, including the integrations named in it and twelve months of fixing them when the far side changes. Work outside that scope is quoted separately before it starts, at 60 an hour, and we would rather have that conversation than quietly absorb changes and rush the rest.
At handover you receive the code in your own repository, the database in your own hosting account, the deployment instructions and the credentials. Nothing is held back as leverage. That is a deliberate commercial position: the aftercare is cancellable and everything is transferable, which means we have to keep earning it. Tell us how your business works and a fixed price with a two page scope comes back inside two business days, in writing, from the contact page.
What the two page scope actually contains
A fixed price is only meaningful if both sides agree what it covers, and the document that does that is short on purpose. Two pages, written in your vocabulary rather than ours, so anybody in your business can read it and object.
- Record types, listed, with the fields that matter on each. This is the line that decides the price, so it is first.
- Pipeline stages, in your words, with the event that moves a record out of each one.
- Automations, numbered. Each one written as when X happens, do Y. Counted, because the count is priced.
- Integrations, named, with direction and what happens when the far side is unavailable.
- Documents and reports, listed by name, with a sample attached where one exists.
- Migration source and row counts, plus the identity rule for duplicates, agreed before anything is imported.
- What is explicitly not included, which is usually the longest section and the most useful one.
That last section does more work than the rest combined. Most disputes in software projects are not about quality, they are about an assumption nobody wrote down, so we write down the assumptions that would be expensive to be wrong about, in the negative, and ask you to challenge them.
The scope is yours whether or not you continue. If you take it to another developer and get a cheaper quote on the same document, that is a good outcome for you and a fair test of us, which is the point of writing it down rather than holding it as leverage.
