Articles
What is a CRM database? Records, fields and how to keep them clean
A CRM database is the shared record of who your customers are and what happened with each of them. Here is what goes into it, how it differs from a spreadsheet, and how to keep it usable.
What is a CRM database?
CRM stands for customer relationship management. The CRM database is the part that stores the facts: who the person is, which company they work for, what you discussed, which deals are open and what has to happen next.
The word "database" matters because the records are linked. A phone call is not a loose note; it belongs to a contact, and that contact belongs to a company and to one or more deals. Open any one of them and you see the rest.
A good CRM database answers everyday questions without a hunt through inboxes: who spoke to this customer last, which deals have gone quiet, and where a lead came from.
What does a CRM database example look like?
Most systems keep a few record types, and each one has its own fields. This is the shape you will meet almost everywhere.
| Record | What it describes | Typical fields |
|---|---|---|
| Contact | One person | Name, email, phone, job title, company, city, source, owner, marketing consent |
| Company | An organization | Name, industry, website, owner |
| Deal | A sale in progress | Title, amount, pipeline stage, linked contact, line items, invoices, tasks |
| Activity | Something that happened | Email, call, chat message, note, meeting |
| Task | Something to do | Description, due date, assignee, repeat rule |
| Custom field | Anything specific to your business | Policy renewal date, property type, preferred language |
A concrete example: Dana Ortiz, office manager at a dental supply distributor, came through a web form, has a deal called "Spring order" worth $4,800 at the "Quote sent" stage, and a task to call on Thursday. That one sentence touches four record types. In a CRM database each fact lives in its own place, linked to the others, so nobody retypes it.
Is Excel a CRM database?
Partly. A spreadsheet is a fine place to start and a poor place to stay.
A spreadsheet can store the list of customers. What it does not do on its own:
- Keep history per customer. An edited cell replaces the old value, and file version history is not a timeline of one client. A CRM keeps the conversation and the stage changes on the record.
- Link things. One row per person cannot show that three people work at one company and share one deal.
- Control access. A shared file is visible in full to everyone who opens it.
- Act. It does not create a task when a deal stalls.
- Stay consistent. "Acme Inc", "ACME" and "Acme, Inc." become three companies.
Once more than one person talks to customers, move the working data into a CRM.
What are some popular CRM databases?
Many well-known systems fill this role, and the right one depends on your team size, channels and budget. Names you will see in most comparisons include Salesforce, HubSpot, Zoho CRM, Pipedrive, Microsoft Dynamics 365 and Odoo, along with NeuroCRM. The comparison pages put NeuroCRM next to several of them, and the guide to simple CRM helps a smaller team avoid paying for features it will not use.
Which fields should a CRM database have?
Start with fewer fields than you think you need. Every field is something a person has to fill in.
- Identity: name, email, phone, company. These are what you match on when you look for duplicates.
- Origin: source, such as web form, referral or ad. Without it you cannot tell which channel pays back.
- Ownership: the person responsible. A record without an owner is a record nobody follows up.
- Stage and value: where the deal is and what it is worth.
- Two or three fields specific to your trade: a renewal date, a site address, a preferred language.
Choose the field type with care. A date typed as free text cannot be sorted or used for a reminder; a dropdown keeps "Referral" from turning into "referral", "Referal" and "word of mouth".
How do you keep CRM data clean?
Clean data is mostly a habit, supported by a few settings.
Decide what is required. Pick the fields without which a record is not worth saving, usually name plus one way to reach the person. Leave the rest optional.
Merge duplicates early. The same customer arrives by phone, then by web form, then through an import. Scan on a schedule, such as monthly, and keep the more complete record as the primary one.
Map columns on purpose when you import. Import damage happens at the mapping step: a "Mobile" column in the wrong field, or notes overwriting names. Review the preview before you confirm.
Record consent as a field. If you send marketing messages, store whether the person agreed. What the law requires depends on your country and industry, so check with your own adviser.
Limit who can change what. Everyone can read a customer's history, but only a few people need to delete or merge records.
How do you move a spreadsheet into a CRM?
- Clean the sheet first. Delete empty rows and put one phone number per cell.
- Give every column a destination. Name, email, phone and company usually exist as standard fields. Anything else needs a custom field.
- Keep the original file untouched as a backup.
- Import ten rows first. They show whether dates, phone formats and special characters survive.
- Run the full import, then check the duplicates and spot-check a few records against the source.
- Retire the spreadsheet. Two live copies quickly drift apart.
How NeuroCRM stores and cleans customer data
Records and linked history
Contacts hold name, email, phone, job title, company, address, source, first and last touch and a marketing consent checkbox. Companies carry name, industry, website and owner. Deals hold stage history, line items, invoices, tasks and linked conversations. Messages from connected channels land in one inbox and attach to the contact.
Custom fields
On a contact card, the Custom fields section has an "Add field" button. You name the field and choose a type: text, long text, number, money, date, single select, multi-select, yes/no, email, phone or link. A field can be marked required. Deals have the same kind of extra fields.
Under Settings → Card layouts you can add rules for contacts, companies, deals and tasks: make a field required, hide it, lock it as read-only, limit the options in a list or validate the value.
If you start from an industry setup, the fields it adds are plain text, so where you need dates, create your own fields with the date type under different names.
Finding and merging duplicates
An admin finds this under Settings → Duplicates & merge. The "Find duplicates" button compares contacts or companies on email, phone and name spelling and sorts each pair it finds into one of three confidence levels. A second pass can ask an AI model to look at borderline pairs, up to 25 per run. You can also tick two or more contacts in the Contacts list and press Merge.
Read the merge dialog before you confirm. The second record is deleted, its deals, tasks and messages move to the primary one, and for conflicting fields you choose which value to keep. A snapshot goes into a merge log, but rolling a merge back is manual. Only owners and admins can merge. A match drawn from an anonymous form or from something the contact said in a message is marked unverified.
Importing a spreadsheet
There are two places to start.
- Contacts → Import takes CSV, TSV, Excel or OpenDocument files. The dialog shows the first five rows, suggests a mapping (you can ask AI for one), and lets you download a CSV template. A column that matches no standard field can become a custom field. If the file has several sheets, the first one is read. Each contact needs a name.
- Settings → Import / Export (admins) loads contacts or deals from CSV or XLSX files up to 20 MB.
In the Contacts → Import dialog, a row with the same email or phone as an existing contact updates that contact instead of creating a duplicate, and the summary says how many were updated. Fields already filled in on the card are not overwritten; the rows that differ are listed so you can decide. An import history lists previous files.
Companies are not imported as their own file. Put the company name in a column of the contacts file and it is created or matched from there.
Access and export
The Team screen gives each person one of three roles under the owner: administrator, manager or operator; a read-only viewer role can also arrive through SSO or SCIM. An operator sees only their own conversations, deals and tasks; a manager sees all of them but cannot change settings; admins and the owner see everything. In Settings → Security an admin can add custom roles that narrow a base role, and set field and stage permissions to restrict a sensitive field further.
In the Contacts list, Export downloads the selected contacts, or the whole current view, as CSV when your role allows it. Settings → Import / Export downloads all contacts or deals as UTF-8 CSV, handy as a backup.
Where the data lives and what it costs
NeuroCRM is hosted by Hetzner in Oregon, USA, with no choice of region; backups are stored separately, outside the US.
Each user seat is $13 a month, or $10.40 a month on annual billing. A new workspace gets the full product for 14 days with as many people as you invite, then switches to view-only until a plan is bought. The current price is on the pricing section of the home page.
What NeuroCRM does not do here
- No separate import of companies. Only contacts and deals can be imported; companies come from the company column of a contacts file.
- No direct transfer from most other CRMs. Bitrix24 data can be pulled over directly through its API; for HubSpot, GoHighLevel and most others you export a file and load it. Kommo has no direct transfer.
- No email sending service of its own. Mass email campaigns run through your own services.
- No free plan. Apart from the 14-day trial, every seat is paid.
FAQ
What is a CRM database example?
A contact such as "Dana Ortiz, office manager", linked to her company, a $4,800 deal at the quote stage, the web form she came from and a task to call her on Thursday. The linking is what makes it a CRM database and not a list.
Is Excel a CRM database?
Not by itself. Excel stores rows well, but it does not keep conversation history, link people to companies and deals or trigger follow-ups. Use it as a starting list or an export; the CRM is the daily working copy.
What are some popular CRM databases?
Salesforce, HubSpot, Zoho CRM, Pipedrive, Microsoft Dynamics 365, Odoo and NeuroCRM are all used as CRM systems. Which fits depends on team size, channels and budget, so compare the two or three that match your needs rather than the longest list.
What does CRM database stand for?
CRM is customer relationship management. The database is the linked record of customers, companies, deals and activity it works from.
Can I export my data out of NeuroCRM?
Yes. The Contacts list exports to CSV, and Settings → Import / Export downloads all contacts or deals as CSV.
Can I record marketing consent in NeuroCRM?
Yes. Each contact card has a "Marketing consent" yes/no field in its marketing section. It is a stored value you set or import; what you are allowed to send is a question for your own legal adviser.
How this article was made: written by the NeuroCRM team with AI assistance; product claims were checked against NeuroCRM code and help docs; prices are from the published pricing as of the update date.