Field completeness
When too few records have a value, reports and agents stop working well. The audit shows the fill rate for every field on every object.
A Salesforce data quality audit grades every object A to F for complete, accurate and unique records. It lists duplicates and empty critical fields, such as Lead.Rating.
Sample org · Data Quality Overview
5 critical objects · 23.9% avg completeness · 2,663 duplicates
Clientell AI, the AI agent for Salesforce admins, scans your org, drafts each fix and asks before it changes anything in production.
Four patterns this audit is built to find. The examples are not real data.
When too few records have a value, reports and agents stop working well. The audit shows the fill rate for every field on every object.
Accounts with no Industry value make territory rules and routing flows fail silently. The audit shows the blank rate for each key field.
Leads with no Rating leave routing agents to guess the value or refuse to act. The audit shows how many are empty.
Duplicate Contact pairs are a routing risk and a deliverability risk. The audit lists the pairs it finds with fuzzy matching.
Click an object to see its completeness heatmap and top field-level gaps. The data below is a sample, not a real org.
Validation rules + dedupe pass proposed per object. Sandbox-tested.
Get my auditSix areas. Each gap is tied to its effect on reports, automation or agents.
Every standard and custom object gets a grade: Account, Lead, Opportunity, Contact, Case and your custom objects. The grade comes from average field completeness, duplicate density and required-field gaps.
A visual map of every field on every object, with color-coded fill rates. Spot the cluster of empty fields in seconds. Drill into any cell for the field name, the fill rate and the proposed fix.
The audit finds duplicate Accounts and Contacts with fuzzy matching on email, phone, name and address. Each pair comes with a merge recommendation.
It finds fields marked required on the page layout but not enforced by a Validation Rule, because the API can bypass a layout. It also finds fields that should be required but are not.
The audit checks the fields that agents and reports depend on: Lead.Rating, Account.Industry, Opportunity.NextStep and Account.AnnualRevenue. Each one is tied to the report or agent that breaks when it is empty.
For each high-impact gap, you get a proposed Validation Rule with its formula. For each duplicate pair, you get a merge recommendation and a winning record. Fixes are tested in a sandbox before you apply them.
| Fill rate | Grade | Operational impact | Action |
|---|---|---|---|
| Nearly all filled | A | Reports are accurate and automation behaves the same each time. | Keep it that way with Validation Rules. |
| Most filled | B | Reports are usable. Automation hits a few edge cases. | Tighten validation. Backfill gaps each quarter. |
| About half filled | C | Reports show gaps. Routing fails on edge cases. | Backfill soon. Add Validation Rules. |
| Few filled | D | Reports are unreliable. Automation can give wrong outcomes. | Stop building new automation on it. Clean up first. |
| Almost none filled | F | The field is effectively dead. Anything that depends on it is broken. | Fill it through enrichment, or remove it from what depends on it. |
If any of these sound familiar, the data is the cause.
Both use the same data source but aggregate it differently. The cause is often empty fields skewing the rollups, such as Opportunity.Amount or CloseDate. The audit shows which fields cause the gap.
Lead.Rating is empty on many records, Account.Industry is blank and a custom score field is half filled. Marketing routes on this data and sales scores on it. Both are flying blind.
Third-party enrichment fills gaps that should have been blocked when the record was created. The audit shows which Validation Rules would stop those gaps from forming.
Duplicates are the usual cause. One customer has two Account records and two Opportunity histories. The audit lists every pair.
Agentforce is Salesforce's AI agent product. Empty fields can push an agent to guess the missing value. The audit shows which fields the agent reads are mostly empty.
Five deliverables, made for RevOps action and executive review.
Every standard and custom object is graded. You get the aggregate score, the fill rate, the top gaps and the duplicate count. It exports as a PDF for executive review.
Every field on every object with its fill rate. Filter by low fill rate. Sort by gap impact.
Every duplicate Account and Contact pair. Each has a recommended winning record, chosen by completeness, recent activity or revenue. It exports to XLSX for review.
For each high-impact gap, you get the Validation Rule formula with a helpful error message. Copy it into Setup (Object Manager > Validation Rules) or review it in a Clientell AI sandbox dry run.
Every custom field with its usage: reports that reference it, layouts that contain it and its fill rate. Dead fields are flagged for safe removal.
What RevOps and data teams ask before they start.
It measures how complete, accurate and unique the data in your org is. It finds empty fields that should be filled, duplicate records, fields marked required but not enforced, and the reports that break when data is missing. Each object gets a grade from A to F.
Salesforce sets no standard. Set the line per field. A field that drives routing, like Lead.Rating, needs a much higher fill rate than an optional notes field.
Salesforce publishes no fixed limit. Even a small share of duplicates hurts on objects that drive routing, such as Lead, Contact and Account. Track the share after each cleanup.
It matches on email (normalized), phone (digits only), name (Levenshtein distance, a count of character edits) and address. Each candidate pair gets a confidence score and a merge suggestion. It never merges on its own, so a person reviews each pair.
Lead routing flows use Rating to pick the right team. Lead scoring models use it as an input. Agentforce lead agents use it to decide which leads to engage. When Rating is blank, all three degrade without any error.
Yes. Every custom field is scored on completeness and on use: is it on a page layout, in a report, in a Validation Rule or read by Apex? Dead custom fields are flagged for safe removal.
Cloudingo and DemandTools clean up data after the problem exists. This audit finds the cause: missing Validation Rules, layout-only required fields and low-use custom fields. Use the audit to map the problem, a cleanup tool to fix it at scale, then Validation Rules to stop it coming back.
Partly. It finds format errors (an email without @), out-of-range values (Account.NumberOfEmployees set to 0 or below), stale data (an old LastActivityDate) and picklist drift ('Closed Won' versus 'ClosedWon'). It cannot judge meaning, such as whether a prospect really became a customer.
Yes. An AI agent is built to give an answer. When the field it needs is empty, it may guess a plausible value instead of asking. The audit shows which fields your agents read are mostly empty.
Data quality is one of the dimensions in the Agentforce Readiness Audit. This audit goes deeper, with per-object grades and field heatmaps. Use Readiness for the launch decision and this audit for the cleanup.
Audits, services and tools for teams that clean up CRM data before reports, automation or AI.
AI-powered data cleanup: duplicate detection, missing field backfill and format standardization. Reversible.
Read moreAuditData quality is one of the AI readiness dimensions. Score your org for agents.
Read moreAuditPer-user risk scoring. Pairs with data quality for full org review.
Read moreAuditRace condition detection on the automation side.
Read moreServiceA broader audit that covers data quality and other areas of your org.
Read moreProductAn AI alternative to Data Loader. Bulk operations driven by plain language.
Read moreProductThe agent that applies the validation rules and dedupe fixes the audit surfaces.
Read moreServiceContinuous data quality monitoring instead of point-in-time audits.
Read moreTrust + security posture
Agents guess when fields are empty. Fix the data before agents go live.
Validation rule gaps let dirty data through. Audit the gatekeeper.
Field-level security (FLS, which fields a user can see or edit) gaps can leave fields empty for some users.
Track when picklist values changed and which records still hold the old labels.
Last reviewed October 2026
A to F grades per object. Field-level heatmap. Duplicate detection. Validation rule prescriptions. Free for 14 days.
14-day free trial · No credit card required