CRM for cabinet makers and millwork shops
A cabinet shop's problem with CRM is that the sale is the shortest part of the job. A kitchen or a commercial millwork package lives for months after the contract — deposit, drawings, revisions, production, delivery, install, deficiencies, final payment — and a sales CRM closes the record at exactly the moment the real correspondence begins. Meanwhile the shop's actual sales engine, the designers, architects and contractors who specify it into their projects, has no model at all in a tool built to chase individual buyers.
Where the generic CRM breaks:
Closed-won is month one of six. The deposit starts the project; the CRM thinks it ended. Every question after — where is the balance, what was agreed in revision three, when is install — gets answered from an email archive instead of a record.
Money arrives in stages the CRM cannot see. Deposit, progress payment at production, balance at install, sometimes a holdback after deficiencies. One amount and one close date cannot represent it, so receivables are tracked in the owner's head against jobs the CRM says are finished.
Specifiers are not customers, so they are nobody. The designer who has placed four kitchens with you is worth more than any single homeowner, but she never buys anything herself — so the CRM files her as a contact on other people's deals and her next referral is left to luck.
The correspondence trail has no spine. Months of revisions, selections and change discussions accumulate in email threads. When a client disputes a finish or a dimension at install, the shop rebuilds the history from inboxes — usually while the crew waits on site.
The data model that actually fits:
Project ledger. The job from deposit to deficiency close-out: payment milestones with amounts, due triggers and status, so the answer to what is owed on which jobs is a report, not an investigation.
Specifier account. Designers, architects, contractors and builders as scored referral accounts — projects placed, revenue produced, last contact — worked as the sales territory they actually are.
Decision record. Selections, revisions and change agreements captured against the project with dates and acknowledgements, so the install-day dispute is settled by the record instead of by whoever kept better email.
Install and deficiency list. Install dates, punch items with owners, and the sign-off that releases final payment — because unresolved deficiencies are where the last ten percent of every job goes to die.
Quote pipeline with ageing. Outstanding quotes as live objects with follow-up, since a cabinet quote is often shopped for months and the shop that follows up in week six wins jobs it had already written off.
Our verdict: If the shop's pain is on the production floor — cut lists, scheduling, materials — that is manufacturing software, not CRM, and it should be bought, not built. The CRM gap is in front of and after the shop: the project ledger, the specifier accounts and the decision record. A generic CRM configured hard can hold maybe half of it; the payment milestones and decision trail are the halves it cannot, and they are a compact build that fits beside whatever runs production.
Why does the last payment on a job take so long to collect?
Because it is usually gated on deficiencies nobody is tracking. The client withholds the balance over three punch items, the items live in a text thread, and weeks pass. A deficiency list with owners and a sign-off that triggers the final invoice turns that drift into a short, dated process.
How should a shop manage designers and architects who refer work?
As accounts with numbers attached: projects placed, revenue, time since last contact. The shops that grow steadily treat their top ten specifiers the way a sales team treats key accounts — regular contact, first look at new capabilities, flawless follow-through — because five reliable specifiers are a full order book.
What belongs in the decision record?
Anything the client chose or changed: finishes, hardware, dimensions, layout revisions, and the price consequence of each, with the client's acknowledgement. It is thirty seconds of capture per decision against hours of email archaeology per dispute — and disputes on months-long custom work are a matter of when, not if.
Can one system cover quoting, the shop floor and the client side?
In practice, no single good one does. Production software is its own discipline and should stay that way. The realistic architecture is a commercial layer — quotes, ledger, specifiers, decisions — that hands a released job to production and receives dates back. Shops that try to force one tool to do everything end up with one that does nothing well.
Last reviewed 27 August 2026