The CRM fields a service business actually needs (and the forty it does not)

A service business needs about eight fields on a customer record: name, mobile number, email, home branch, source, consent status per purpose, a status (lead, customer, inactive), and an owner. Plus a handful specific to the industry: the insurance plan for a dental office, the vehicle for a mechanic, the pet for a vet. Everything else is a note or a tag. A record with forty fields is a record with thirty-two empty ones, and empty fields teach the team that fields do not matter.

The eight

1. Name. As they gave it. First name is what the AI uses.
2. Mobile number. The key identifier. More than one is allowed.
3. Email. For confirmations and anything to keep.
4. Home branch. For multi-location businesses; where they usually go.
5. Source. How they first arrived: missed call, web form, Instagram, referral, walk-in. Set once, automatically where possible.
6. Consent. Per purpose (reminders, marketing) and per channel, with the record behind each.
7. Status. Lead, customer, inactive. Three values. Not a nine-stage pipeline.
8. Owner. The person or queue responsible for them, if anyone is.

The industry handful

Dental and medical: insurance plan and member ID (for the office to verify; not what is covered). Preferred provider.

Med spa and salon: preferred provider. Deposit on file. Notes on preferences.

Home services: service address (may differ from billing). Equipment or system details. Maintenance plan.

Veterinary and grooming: pets, each with species, breed, name, vaccination status.

Automotive: vehicles, each with year, make, model, plate.

Legal and accounting: matter type or entity type. Conflict-check names.

Real estate: buying or selling, area, timeline, representation.

Each of these changes what the AI asks or what a person does. That is the test for a field.

Why not more

Every field is a promise that someone will fill it in and keep it current. Fields that are not needed for a reply or a report get filled in once, badly, and then rot. The team learns to ignore them, and then ignores the ones that matter. Fewer fields, all of them used, is the only setup that stays accurate.

Notes and tags for the rest

"Prefers mornings." "Allergic to latex." "Husband is also a patient." "Sensitive about pricing." These are notes on the timeline, in the person's words, dated. Tags are for filtering groups: "maintenance plan", "VIP", "do not book with Sam". A tag you cannot name a use for is a field in disguise.

What the AI fills in

Source, on first contact. Consent, when given in conversation. Home branch, on first booking. Status, on first booking (lead to customer). Industry details the customer mentions (the pet's name, the vehicle), proposed for confirmation rather than written silently. Everything the AI writes is marked and reversible.

Reviewing the fields

Once a quarter, look at the share of records with each field filled. Below half, ask why: is it not needed, or not captured? Either fix the capture or drop the field. Fields nobody fills are the ones to delete.

Frequently asked questions

Should I keep a "birthday" field?

Only if you will send something on it and have marketing consent to do so. Otherwise it is a field that is never used.

What about custom fields per branch?

Rarely needed. Branches differ in services and staff, not in what a customer record holds.

Where does the pipeline go?

Leads have a status and an owner and a next action; that is the pipeline for most service businesses. If you sell long, multi-stage projects, a small pipeline on the lead makes sense. Not on every customer.

What Frontiva does here

Frontiva's CRM creates the contact from the first message and keeps notes, tags and a lifecycle stage (lead, prospect, customer, lapsed), and you define custom fields (text, number, date, select, multi-select or yes/no) once for every contact. Your industry template seeds services and knowledge at signup. See what an AI CRM is for the wider picture.