Why Your Salesforce Reports Are Lying to You (And How to Fix It)

There is a particular kind of organizational dysfunction that is almost impossible to diagnose from the outside. The dashboards look polished. The pipeline numbers get presented in leadership meetings without question. The reports are color-coded and formatted and exported to PowerPoint. And underneath all of it, the data is wrong.

Not obviously wrong. Not wrong in a way that triggers an alert or causes a system error. Wrong in the quiet, compounding way that bad data always is — duplicates that inflate contact counts, closed-won opportunities that were never actually closed, activity logs that nobody fills out, fields that mean different things to different reps.

I have worked in enough Salesforce orgs to know that this is not an edge case. It is the default state of most orgs that have been running for more than two or three years without deliberate data governance. Here is how to find the problems and fix them.

The Most Common Ways Salesforce Data Goes Wrong

Before you can fix bad data, you need to understand how it gets created. In my experience, the sources are almost always the same.

Duplicate records

Duplicates are the most visible data quality problem and often the most politically charged. When two reps both create a contact for the same person, or when a lead gets converted without checking for an existing account, you end up with split history, conflicting data, and reports that count the same entity twice. Salesforce has native duplicate management tools, but they only work if someone has configured them and enforced them consistently.

Stale pipeline data

Opportunities that have not been touched in 90 days but are still sitting at Stage 3. Close dates that have been pushed forward six times. Deals marked as "Negotiation" that the rep has mentally written off but never updated. This is not laziness — it is a process problem. If updating Salesforce feels like extra work with no personal benefit to the rep, it will not get done.

Inconsistent field usage

A picklist field with 14 values, half of which mean roughly the same thing. A "Lead Source" field where some reps type free text instead of selecting from the list. A "Region" field that some people use for geography and others use for business unit. When the same field means different things to different users, any report built on that field is unreliable.

Missing required data

Fields that are technically required in the UI but have been worked around — either by entering placeholder values like "TBD" or "N/A," or by using a different record type that does not enforce the same rules. Required fields that are not actually required in practice are worse than optional fields, because they create the illusion of completeness.

How to Audit Your Salesforce Data

A data quality audit does not require a third-party tool or a multi-week project. You can learn a lot with native Salesforce reporting and a few hours of focused attention.

Start with these reports:

• Contacts or Leads with no Activity in the last 180 days — these are likely stale or duplicated
• Opportunities with a Close Date in the past that are still Open — your pipeline is overstated
• Accounts with no associated Contacts — orphaned records that clutter searches and inflate counts
• Records where key fields (Industry, Lead Source, Account Type) are blank — gaps in your segmentation data
• Duplicate Leads or Contacts report using the native Duplicate Management tool

Run each of these reports and look at the record counts. If you have hundreds or thousands of records in any of these categories, you have a data quality problem that is actively distorting your reporting.

Fixing the Data vs. Fixing the Process

Here is the mistake most organizations make: they run a data cleanup project, deduplicate their records, fill in the missing fields, archive the stale opportunities — and then six months later, the data is just as bad as it was before.

Cleaning the data without fixing the process that created the bad data is like bailing out a boat without patching the hole. The cleanup is necessary, but it is not sufficient.

After any data cleanup effort, you need to address the upstream causes. That usually means some combination of:

• Configuring duplicate rules and matching rules so new duplicates are caught at entry
• Simplifying picklist values so there is one clear option for each concept
• Adding validation rules that enforce data completeness at the point of entry, not after the fact
• Building reports that make data quality visible to managers so it becomes a coaching conversation
• Reviewing your page layouts to remove fields nobody uses and surface the ones that matter

When to Bring in Outside Help

Some data quality problems are straightforward enough to handle internally. Others have gotten complex enough — or politically sensitive enough — that an outside perspective is genuinely useful.

If your org has been running for several years without a formal data governance process, if you have had multiple admins with different approaches, or if leadership has lost confidence in the numbers coming out of Salesforce, a structured data quality assessment is worth the investment. The goal is not just clean data — it is restoring trust in the platform so people actually use it.