No number to start from
"We have a lot of duplicates" is not a baseline. Without dated counts per object, nobody can say later whether the cleanup, the new import rules or the extra licence actually changed anything.
Salesforce · Cloudingo Data Quality Audit
Salesforce data quality audit measured before anything merges
A Salesforce data quality audit is a fixed-scope diagnostic: it measures duplicates, completeness, consistency, accuracy and validity across your key objects, records a dated baseline, and ranks what to fix first. Twopir Consulting runs it with Cloudingo as the measuring instrument, in a read-only pass that merges nothing and changes no records, then scores every finding with our own method. You leave with a baseline, a ranked findings register and a roadmap.
12+
500+
250+
40+
Trusted by 500+ organizations — including growing and mid-market companies that needed to know what was in their Salesforce data before deciding how to fix it.
What We Audit
- Salesforce Partner
- Leads & Contacts
- Accounts
- Person Accounts
- Opportunities & Cases
- Custom Objects
- Lead-to-Contact Overlap
- Record Entry Points
Cleanup without a baseline cannot prove it worked
Most teams know their Salesforce data is wrong. Few can say how wrong, on which objects, or which problem costs them most. Starting the fix before answering that is how cleanup budgets get spent on the loudest complaint instead of the most expensive one. These are the gaps an audit closes.
Duplicates counted with one loose filter
A single fuzzy filter run across every lead produces a big, frightening number that mixes true duplicates with lookalikes. It cannot tell you which matches are safe to merge or where they come from.
Fields that are full but wrong
A fill rate looks healthy until you read the values: placeholder phone numbers, "N/A" in the industry field, a country written five ways. Completeness and accuracy are different questions.
The source keeps producing the problem
A web form, an integration or a monthly import creates new bad records every day. Clean the backlog without finding the entry point and the numbers drift straight back.
Fixes run in the wrong order
Merging accounts before company names are standardized, or deduplicating contacts before leads, means matching the same records twice and arguing about survivors that a different order would have settled.
No agreement on what matters
Sales cares about account hierarchy, marketing about email, finance about billing contacts. Without a shared ranking, each team argues for its own problem and the backlog never moves.
Definition
Data quality audit, defined
A Salesforce data quality audit is a fixed-scope assessment that measures the condition of the records in your org, at a point in time, and ranks what to fix. It covers duplicates within and across objects, the completeness of the fields your processes depend on, the consistency of values and formats, the accuracy of a verified sample, and the validity of values against the rules they should follow. The result is a dated baseline, a scored findings register and a prioritised roadmap. It is diagnostic only: nothing is merged, updated or deleted during the audit.
Three things are involved, and it helps to keep them apart. Cloudingo, a Salesforce deduplication and data-management application from Symphonic Source, is the measuring instrument: its dashboard counts duplicates, its filters work like reports and change nothing until a merge is submitted, and its field analysis shows how fields are used. Salesforce supplies the reports, field metadata and record history that complete the picture. Twopir Consulting supplies the method: what to measure, how to score each finding, and in what order to fix it. Cloudingo does not produce a scored audit report, so the scoring and prioritisation on this page are ours. Product behaviour is documented in the Cloudingo help documentation on filters.
Do you need an audit, or can you go straight to cleanup? If the problem is one known duplicate pattern on one object, skip the audit and fix it. An audit earns its place when several teams disagree about what is wrong, when you are about to buy or re-scope a tool, after a migration or an org merge, or before an AI, routing or reporting project that will act on the records as they are.
New to the product itself? Start with our getting-started guide to Cloudingo. Running a different duplicate tool? We also work with Duplicate Check for Salesforce.
Five dimensions, measured object by object
Each dimension answers a different question, and each needs a different instrument. We measure all five on the objects you agree are in scope, usually Leads, Contacts and Accounts first, then Opportunities, Cases or custom objects where they carry the business process.
| Dimension | The question it answers | How we measure it |
|---|---|---|
| Duplicates | How many records describe the same person or company, within one object and across leads and contacts? | Cloudingo filters run as reports, from tight exact matches to looser rules, with the duplicate groups counted per filter and per object. Nothing is merged. |
| Completeness | Are the fields your routing, reporting and outreach depend on actually filled, at the stage where they should be? | Cloudingo field analysis for field usage, Salesforce reports for fill rate by record type, stage and source. |
| Consistency | Is the same thing written the same way: company names, countries and states, picklist values, phone formats? | Value distributions per field, variant counts for the values that should be standard, and picklist values no longer in use. |
| Accuracy | When a field is filled, is it true? | A sample of records checked against a source of truth your owners name, such as billing, the website or the person themselves. Tools cannot prove truth, so this is sampled and stated as such. |
| Validity | Does each value fit the rules for its field: format, allowed range, required pairing? | Rule checks against your field standards, such as email shape, postal code by country, dates in range and dependent fields left blank. |
Alongside the five, every finding records where the bad records come from: the form, integration, import or user that created them. A finding without a source is only half a finding, because the fix has to close the entry point as well as the backlog.
How Cloudingo becomes the measuring instrument
Cloudingo is built to find and merge duplicates. For an audit we use only the finding half: the parts of the product that report on your data without changing it. Each one answers part of the question, and each has a limit we account for.
| Cloudingo instrument | What it tells the audit |
|---|---|
| Dashboard | Duplicate counts in the org, recent job runs and activity. Every Cloudingo plan includes an analysis of the current health of your Salesforce data. The counts depend on the filters behind them, which is why we design those first. |
| Filters, run as reports | Duplicate groups for each matching rule. Cloudingo's own advice is to start with tight exact-match filters and loosen gradually, and we record the count at each step, so you can see how many groups are certain and how many need review. |
| Cross-object filters | Leads that already exist as contacts, and leads that match an existing account: the overlap marketing and sales usually argue about. |
| Matching algorithms | Exact, fuzzy and synonym matching, plus phone, company-name and domain cleaning, show how much of the duplication only appears once formatting differences are ignored. That gap is itself a consistency finding. |
| Field analysis | The number, type and usage of fields on any object: the starting inventory for the completeness and consistency checks. |
| Pre-merge reports | Which records would merge into which master, with field values and IDs, without submitting the merge. We use them to test whether a proposed rule is safe, and hand them to the deduplication work as evidence. |
How we score and rank every finding
This is Twopir's method, not a Cloudingo feature. Cloudingo tells us how many records are affected; the score adds what that costs the business and what fixing it takes. Every finding is scored on the same four factors, so a duplicate problem and a completeness problem can be ranked against each other.
| Factor | What it weighs | How it is scored |
|---|---|---|
| Business impact | Which process the problem breaks: routing, forecasting, billing, campaigns, service or compliance, and who feels it. | High, medium or low, agreed with the owner of the process it affects. |
| Reach | The share of in-scope records affected, and whether those are active records or long-dormant ones. | Banded from the measured count, so the score traces back to a number in the baseline. |
| Trend | Whether an entry point is still creating the problem today, or the backlog is historical and static. | Active source or closed source, from record creation dates and sources. |
| Effort and risk | Whether the fix can be automated safely or needs a person to review each case, and what else depends on it. | Automatable, rule plus review, or manual, with any fix that must come first named. |
Fix First
High impact, an active source, or a fix that unblocks others. These open the roadmap.
- Typical
- An integration creating a duplicate account on every transaction
- Hand-off
- Close the source, then remediate the backlog
Plan Next
Real cost, but a static backlog or a dependency on a first-tier fix. Sequenced, not urgent.
- Typical
- Company-name variants that block safe account matching
- Hand-off
- Standardize before the related merge work
Monitor
Low reach or low impact. Recorded with its count and re-measured, not worked yet.
- Typical
- Gaps in an optional field on dormant records
- Hand-off
- Re-measure against the baseline next quarter
Six steps from access to a ranked roadmap
The scope is fixed before the audit starts: which objects, which questions, which owners. The duration follows from that scope and from record volumes, and we agree it with you before any work begins.
Scope & Questions
We agree the objects in scope, the business questions the audit must answer, and the owner for each process, so every finding later has someone to agree its impact.
DeliverableAudit scope and owner list
Access & Inventory
Cloudingo connected over OAuth with the right field permissions, then a field inventory per object from field analysis and Salesforce metadata, including record types and entry points.
DeliverableField and source inventory
Baseline
Record volumes and duplicate groups per object, captured with named, dated filters from exact to loose, plus lead-to-contact and lead-to-account overlap.
DeliverableDated baseline
Measure
Completeness, consistency and validity checks on the fields that matter, an accuracy sample verified with your owners, and the source of each problem traced.
DeliverableMeasured findings by dimension
Score & Prioritise
Each finding scored on impact, reach, trend and effort, reviewed with the owners, and placed in Fix First, Plan Next or Monitor.
DeliverableScored findings register
Readout & Roadmap
A working session with your owners and admins: what we found, why it ranks where it does, and the sequenced roadmap, with each item assigned to the work that fixes it.
DeliverableReadout and prioritised roadmap
What you receive at the readout
Six deliverables, written for two audiences: the leaders deciding what to fund, and the admins and specialists who will do the work. Each one is yours to keep, re-run and share with whoever does the fix.
Data Quality Baseline
The dated starting point every later improvement is measured against.
- Record volumes per object in scope
- Results for all five dimensions
- The method behind each number
- Repeatable on the same filters
Duplicate Analysis
Duplicate groups by object and by matching rule, separated into certain and needs-review.
- Groups per filter, exact to loose
- Lead-to-contact and lead-to-account overlap
- Sample pre-merge reports
- The entry points creating new duplicates
Field Health Map
The fields your processes rely on, and how well each one holds up.
- Fill rate by record type and stage
- Variant counts for standard values
- Validity rule failures per field
- Unused fields and picklist values
Accuracy Sample
A verified sample, stated as a sample, with the source of truth named for each field.
- Sample size and selection method
- Fields checked and their source
- Error patterns found in the sample
- Which errors need people, not rules
Scored Findings Register
Every finding, its evidence and its score, in one list your owners have reviewed.
- Impact, reach, trend and effort
- Fix First, Plan Next or Monitor
- Source and owner for each
- Dependencies between findings
Prioritised Roadmap
The order of work, and which kind of work each step is.
- Sequenced remediation steps
- Source fixes before backlog fixes
- Cloudingo plan fit for the work ahead
- Re-measurement points against the baseline
A roadmap that hands off to the right next step
The audit stops at the roadmap. Each roadmap item names the kind of work that fixes it, and your team, Twopir or another partner can pick it up with the evidence already attached.
Set up or repair Cloudingo
When the audit shows the plan, permissions or filter set cannot support the work ahead, the first step is Cloudingo implementation and configuration.
Merge the duplicates
Duplicate findings move to Cloudingo deduplication services with the filters, counts and pre-merge reports from the audit as the starting evidence.
Standardize and complete
Consistency, completeness and stale-record findings become a Cloudingo data cleansing plan, often run before the related merges.
Keep it from coming back
Where the audit shows nobody owns the data, ownership, standards and a review routine belong to Cloudingo data quality consulting.
Measurement, scoring and a plan
Scope, access, the baseline, the five dimensions, the accuracy sample, source tracing, the scored findings register, the readout and the roadmap.
The org is left exactly as we found it, so the baseline describes your data, not our changes.
Any change to your records
Merges, mass updates and deletes, imports, new filters running on a schedule, validation or Flow changes, and fixes to forms and integrations.
Those are remediation. They are scoped from the roadmap as separate work, so the scope of the audit never grows with what it finds.
Two cleanups that started with a diagnosis
Both customer stories below are published by Cloudingo, and the figures are Cloudingo's, not Twopir results. They are here because each turns on a question an audit answers before the fix: what order to work in, and where the bad data comes from. Beside each, what we take from it into our own audits.
Audit Lesson
The story begins with an audit of systems and processes, and the cleanup began with a question about the hierarchy of cleansing. The worst duplication came from one application creating an account on every card transaction. That is a Fix First finding twice over: an active source, and a problem that would undo any merge until it was closed.
Twopir Consulting Our reading of a customer story published by Cloudingo
Docker, Inc.
A credit-card processing app created a new account for each transaction, and the sales team stopped relying on Salesforce.
60 Accounts possible for one company
$50K–$1M Yearly cost of bad data, as reported
4 Stages in the cleanup order
Audit Lesson
Here the problems arrived with a migration that moved the data without cleaning it, into an org where new records land every day. An audit after a migration measures the inherited backlog and the daily inflow separately, because they need different fixes and a different cadence.
Twopir Consulting Our reading of a customer story published by Cloudingo
Hotze Health & Wellness Center
Migrated from Microsoft Dynamics with no attention to data quality, so the new Salesforce org inherited the old problems.
~30,000 Patients on the roster
Daily New records and Cloudingo runs
Monthly Review of what did not merge
An audit read by people who build Salesforce orgs
A tool can count duplicates. Explaining why they exist, which process they break and what to fix first takes people who know how leads, accounts, integrations and automation fit together in a real org.
We trace problems to their source
Forms, imports, integrations and automation are where bad records start. We read the org the way its builders would, so a finding names the entry point as well as the count.
Our scoring is written down
Every finding carries its count, its method and its score on the same four factors. You can challenge a ranking, re-run a measurement, or hand the register to anyone.
We keep the audit read-only
No merges, no scheduled jobs, no mass updates during the audit. The baseline describes your data as it is, which is the only way a later improvement can be measured honestly.
Your owners agree the impact
Impact scores are set with the people who run the process, not by us alone. That is why the roadmap survives the first budget meeting.
The roadmap is not a sales funnel
Some findings belong in the Monitor tier, some fixes are a single validation rule your admin can add. We say so. When you want help with the rest, ongoing Salesforce support is a standing offer.
Data quality audit questions, answered plainly
A Salesforce data quality audit is a fixed-scope assessment that measures the condition of the records in your org at a point in time and ranks what to fix. Twopir Consulting measures duplicates, completeness, consistency, accuracy and validity on the objects in scope, records a dated baseline, scores every finding, and delivers a prioritised roadmap. It is diagnostic only: nothing is merged, updated or deleted.
No. Cloudingo filters behave like reports and change nothing until a merge is submitted, and during the audit we submit no merges, schedule no automation jobs and run no mass updates, deletes or imports. The org is left as we found it, so the baseline describes your data rather than our changes. Any fix is scoped afterwards as separate work.
No. Cloudingo provides a dashboard with duplicate counts, filters that report duplicate groups, field analysis, pre-merge reports and, on every plan, an analysis of the current health of your Salesforce data. It does not produce a scored audit report. The scoring and prioritisation in a Twopir audit are our own method, which weighs business impact, reach, trend and effort for every finding.
If you already run Cloudingo, we use your connection. If you do not, we agree how it is licensed for the audit window during scoping. Cloudingo offers a 10-day free trial and says it shows how many duplicates you have within minutes. Cloudingo's custom-object merging, which covers Opportunities, Cases, Campaign Members and true custom objects, needs the Professional or Enterprise plan and is enabled by Cloudingo on request; where those objects are in scope without it, we measure them with Salesforce reports instead.
It depends on the number of objects in scope, record volumes and how many process owners need to agree the impact scores, so Twopir fixes the duration when the scope is agreed, before the audit starts. An audit of Leads, Contacts and Accounts with one owner per process is the shortest. Custom objects, several business units and an accuracy sample across many fields add time.
Cloudingo runs in the cloud and connects to Salesforce through an OAuth connection over the API, and it states that customer data is neither stored nor cached. Cloudingo also lists SOC 2 Type II compliance and 256-bit SSL encryption. We check the connected user's permissions before measuring, and filter exports and accuracy samples are kept in storage your team controls.
You receive the baseline, the duplicate analysis, the field health map, the accuracy sample, the scored findings register and the prioritised roadmap at a readout with your owners and admins. Each roadmap item names the work that fixes it: Cloudingo setup, deduplication, cleansing, a source fix or an ongoing data quality program. Your team, Twopir or another partner can take it from there, and the baseline lets anyone measure the result.
Next Step
Measure it first, then decide the fix
A first conversation covers the objects that matter, the teams that disagree about the data, how records arrive, and whether Cloudingo is already in place: enough to scope the audit and say what it will answer. No obligation to proceed.
Talk to a Salesforce architect who audits with Cloudingo
Salesforce · Cloudingo Data Quality Audit
Salesforce data quality audit measured before anything merges
A Salesforce data quality audit is a fixed-scope diagnostic: it measures duplicates, completeness, consistency, accuracy and validity across your key objects, records a dated baseline, and ranks what to fix first. Twopir Consulting runs it with Cloudingo as the measuring instrument, in a read-only pass that merges nothing and changes no records, then scores every finding with our own method. You leave with a baseline, a ranked findings register and a roadmap.
12+
500+
250+
40+
Trusted by 500+ organizations — including growing and mid-market companies that needed to know what was in their Salesforce data before deciding how to fix it.
What We Audit
- Salesforce Partner
- Leads & Contacts
- Accounts
- Person Accounts
- Opportunities & Cases
- Custom Objects
- Lead-to-Contact Overlap
- Record Entry Points
Cleanup without a baseline cannot prove it worked
Most teams know their Salesforce data is wrong. Few can say how wrong, on which objects, or which problem costs them most. Starting the fix before answering that is how cleanup budgets get spent on the loudest complaint instead of the most expensive one. These are the gaps an audit closes.
No number to start from
"We have a lot of duplicates" is not a baseline. Without dated counts per object, nobody can say later whether the cleanup, the new import rules or the extra licence actually changed anything.
Duplicates counted with one loose filter
A single fuzzy filter run across every lead produces a big, frightening number that mixes true duplicates with lookalikes. It cannot tell you which matches are safe to merge or where they come from.
Fields that are full but wrong
A fill rate looks healthy until you read the values: placeholder phone numbers, "N/A" in the industry field, a country written five ways. Completeness and accuracy are different questions.
The source keeps producing the problem
A web form, an integration or a monthly import creates new bad records every day. Clean the backlog without finding the entry point and the numbers drift straight back.
Fixes run in the wrong order
Merging accounts before company names are standardized, or deduplicating contacts before leads, means matching the same records twice and arguing about survivors that a different order would have settled.
No agreement on what matters
Sales cares about account hierarchy, marketing about email, finance about billing contacts. Without a shared ranking, each team argues for its own problem and the backlog never moves.
Definition
Data quality audit, defined
A Salesforce data quality audit is a fixed-scope assessment that measures the condition of the records in your org, at a point in time, and ranks what to fix. It covers duplicates within and across objects, the completeness of the fields your processes depend on, the consistency of values and formats, the accuracy of a verified sample, and the validity of values against the rules they should follow. The result is a dated baseline, a scored findings register and a prioritised roadmap. It is diagnostic only: nothing is merged, updated or deleted during the audit.
Three things are involved, and it helps to keep them apart. Cloudingo, a Salesforce deduplication and data-management application from Symphonic Source, is the measuring instrument: its dashboard counts duplicates, its filters work like reports and change nothing until a merge is submitted, and its field analysis shows how fields are used. Salesforce supplies the reports, field metadata and record history that complete the picture. Twopir Consulting supplies the method: what to measure, how to score each finding, and in what order to fix it. Cloudingo does not produce a scored audit report, so the scoring and prioritisation on this page are ours. Product behaviour is documented in the Cloudingo help documentation on filters.
Do you need an audit, or can you go straight to cleanup? If the problem is one known duplicate pattern on one object, skip the audit and fix it. An audit earns its place when several teams disagree about what is wrong, when you are about to buy or re-scope a tool, after a migration or an org merge, or before an AI, routing or reporting project that will act on the records as they are.
New to the product itself? Start with our getting-started guide to Cloudingo. Running a different duplicate tool? We also work with Duplicate Check for Salesforce.
Five dimensions, measured object by object
Each dimension answers a different question, and each needs a different instrument. We measure all five on the objects you agree are in scope, usually Leads, Contacts and Accounts first, then Opportunities, Cases or custom objects where they carry the business process.
| Dimension | The question it answers | How we measure it |
|---|---|---|
| Duplicates | How many records describe the same person or company, within one object and across leads and contacts? | Cloudingo filters run as reports, from tight exact matches to looser rules, with the duplicate groups counted per filter and per object. Nothing is merged. |
| Completeness | Are the fields your routing, reporting and outreach depend on actually filled, at the stage where they should be? | Cloudingo field analysis for field usage, Salesforce reports for fill rate by record type, stage and source. |
| Consistency | Is the same thing written the same way: company names, countries and states, picklist values, phone formats? | Value distributions per field, variant counts for the values that should be standard, and picklist values no longer in use. |
| Accuracy | When a field is filled, is it true? | A sample of records checked against a source of truth your owners name, such as billing, the website or the person themselves. Tools cannot prove truth, so this is sampled and stated as such. |
| Validity | Does each value fit the rules for its field: format, allowed range, required pairing? | Rule checks against your field standards, such as email shape, postal code by country, dates in range and dependent fields left blank. |
Alongside the five, every finding records where the bad records come from: the form, integration, import or user that created them. A finding without a source is only half a finding, because the fix has to close the entry point as well as the backlog.
How Cloudingo becomes the measuring instrument
Cloudingo is built to find and merge duplicates. For an audit we use only the finding half: the parts of the product that report on your data without changing it. Each one answers part of the question, and each has a limit we account for.
| Cloudingo instrument | What it tells the audit |
|---|---|
| Dashboard | Duplicate counts in the org, recent job runs and activity. Every Cloudingo plan includes an analysis of the current health of your Salesforce data. The counts depend on the filters behind them, which is why we design those first. |
| Filters, run as reports | Duplicate groups for each matching rule. Cloudingo's own advice is to start with tight exact-match filters and loosen gradually, and we record the count at each step, so you can see how many groups are certain and how many need review. |
| Cross-object filters | Leads that already exist as contacts, and leads that match an existing account: the overlap marketing and sales usually argue about. |
| Matching algorithms | Exact, fuzzy and synonym matching, plus phone, company-name and domain cleaning, show how much of the duplication only appears once formatting differences are ignored. That gap is itself a consistency finding. |
| Field analysis | The number, type and usage of fields on any object: the starting inventory for the completeness and consistency checks. |
| Pre-merge reports | Which records would merge into which master, with field values and IDs, without submitting the merge. We use them to test whether a proposed rule is safe, and hand them to the deduplication work as evidence. |
How we score and rank every finding
This is Twopir's method, not a Cloudingo feature. Cloudingo tells us how many records are affected; the score adds what that costs the business and what fixing it takes. Every finding is scored on the same four factors, so a duplicate problem and a completeness problem can be ranked against each other.
| Factor | What it weighs | How it is scored |
|---|---|---|
| Business impact | Which process the problem breaks: routing, forecasting, billing, campaigns, service or compliance, and who feels it. | High, medium or low, agreed with the owner of the process it affects. |
| Reach | The share of in-scope records affected, and whether those are active records or long-dormant ones. | Banded from the measured count, so the score traces back to a number in the baseline. |
| Trend | Whether an entry point is still creating the problem today, or the backlog is historical and static. | Active source or closed source, from record creation dates and sources. |
| Effort and risk | Whether the fix can be automated safely or needs a person to review each case, and what else depends on it. | Automatable, rule plus review, or manual, with any fix that must come first named. |
Fix First
High impact, an active source, or a fix that unblocks others. These open the roadmap.
- Typical
- An integration creating a duplicate account on every transaction
- Hand-off
- Close the source, then remediate the backlog
Plan Next
Real cost, but a static backlog or a dependency on a first-tier fix. Sequenced, not urgent.
- Typical
- Company-name variants that block safe account matching
- Hand-off
- Standardize before the related merge work
Monitor
Low reach or low impact. Recorded with its count and re-measured, not worked yet.
- Typical
- Gaps in an optional field on dormant records
- Hand-off
- Re-measure against the baseline next quarter
Six steps from access to a ranked roadmap
The scope is fixed before the audit starts: which objects, which questions, which owners. The duration follows from that scope and from record volumes, and we agree it with you before any work begins.
Scope & Questions
We agree the objects in scope, the business questions the audit must answer, and the owner for each process, so every finding later has someone to agree its impact.
DeliverableAudit scope and owner list
Access & Inventory
Cloudingo connected over OAuth with the right field permissions, then a field inventory per object from field analysis and Salesforce metadata, including record types and entry points.
DeliverableField and source inventory
Baseline
Record volumes and duplicate groups per object, captured with named, dated filters from exact to loose, plus lead-to-contact and lead-to-account overlap.
DeliverableDated baseline
Measure
Completeness, consistency and validity checks on the fields that matter, an accuracy sample verified with your owners, and the source of each problem traced.
DeliverableMeasured findings by dimension
Score & Prioritise
Each finding scored on impact, reach, trend and effort, reviewed with the owners, and placed in Fix First, Plan Next or Monitor.
DeliverableScored findings register
Readout & Roadmap
A working session with your owners and admins: what we found, why it ranks where it does, and the sequenced roadmap, with each item assigned to the work that fixes it.
DeliverableReadout and prioritised roadmap
What you receive at the readout
Six deliverables, written for two audiences: the leaders deciding what to fund, and the admins and specialists who will do the work. Each one is yours to keep, re-run and share with whoever does the fix.
Data Quality Baseline
The dated starting point every later improvement is measured against.
- Record volumes per object in scope
- Results for all five dimensions
- The method behind each number
- Repeatable on the same filters
Duplicate Analysis
Duplicate groups by object and by matching rule, separated into certain and needs-review.
- Groups per filter, exact to loose
- Lead-to-contact and lead-to-account overlap
- Sample pre-merge reports
- The entry points creating new duplicates
Field Health Map
The fields your processes rely on, and how well each one holds up.
- Fill rate by record type and stage
- Variant counts for standard values
- Validity rule failures per field
- Unused fields and picklist values
Accuracy Sample
A verified sample, stated as a sample, with the source of truth named for each field.
- Sample size and selection method
- Fields checked and their source
- Error patterns found in the sample
- Which errors need people, not rules
Scored Findings Register
Every finding, its evidence and its score, in one list your owners have reviewed.
- Impact, reach, trend and effort
- Fix First, Plan Next or Monitor
- Source and owner for each
- Dependencies between findings
Prioritised Roadmap
The order of work, and which kind of work each step is.
- Sequenced remediation steps
- Source fixes before backlog fixes
- Cloudingo plan fit for the work ahead
- Re-measurement points against the baseline
A roadmap that hands off to the right next step
The audit stops at the roadmap. Each roadmap item names the kind of work that fixes it, and your team, Twopir or another partner can pick it up with the evidence already attached.
Set up or repair Cloudingo
When the audit shows the plan, permissions or filter set cannot support the work ahead, the first step is Cloudingo implementation and configuration.
Merge the duplicates
Duplicate findings move to Cloudingo deduplication services with the filters, counts and pre-merge reports from the audit as the starting evidence.
Standardize and complete
Consistency, completeness and stale-record findings become a Cloudingo data cleansing plan, often run before the related merges.
Keep it from coming back
Where the audit shows nobody owns the data, ownership, standards and a review routine belong to Cloudingo data quality consulting.
Measurement, scoring and a plan
Scope, access, the baseline, the five dimensions, the accuracy sample, source tracing, the scored findings register, the readout and the roadmap.
The org is left exactly as we found it, so the baseline describes your data, not our changes.
Any change to your records
Merges, mass updates and deletes, imports, new filters running on a schedule, validation or Flow changes, and fixes to forms and integrations.
Those are remediation. They are scoped from the roadmap as separate work, so the scope of the audit never grows with what it finds.
Two cleanups that started with a diagnosis
Both customer stories below are published by Cloudingo, and the figures are Cloudingo's, not Twopir results. They are here because each turns on a question an audit answers before the fix: what order to work in, and where the bad data comes from. Beside each, what we take from it into our own audits.
Audit Lesson
The story begins with an audit of systems and processes, and the cleanup began with a question about the hierarchy of cleansing. The worst duplication came from one application creating an account on every card transaction. That is a Fix First finding twice over: an active source, and a problem that would undo any merge until it was closed.
Twopir Consulting Our reading of a customer story published by Cloudingo
Docker, Inc.
A credit-card processing app created a new account for each transaction, and the sales team stopped relying on Salesforce.
60 Accounts possible for one company
$50K–$1M Yearly cost of bad data, as reported
4 Stages in the cleanup order
Audit Lesson
Here the problems arrived with a migration that moved the data without cleaning it, into an org where new records land every day. An audit after a migration measures the inherited backlog and the daily inflow separately, because they need different fixes and a different cadence.
Twopir Consulting Our reading of a customer story published by Cloudingo
Hotze Health & Wellness Center
Migrated from Microsoft Dynamics with no attention to data quality, so the new Salesforce org inherited the old problems.
~30,000 Patients on the roster
Daily New records and Cloudingo runs
Monthly Review of what did not merge
An audit read by people who build Salesforce orgs
A tool can count duplicates. Explaining why they exist, which process they break and what to fix first takes people who know how leads, accounts, integrations and automation fit together in a real org.
We trace problems to their source
Forms, imports, integrations and automation are where bad records start. We read the org the way its builders would, so a finding names the entry point as well as the count.
Our scoring is written down
Every finding carries its count, its method and its score on the same four factors. You can challenge a ranking, re-run a measurement, or hand the register to anyone.
We keep the audit read-only
No merges, no scheduled jobs, no mass updates during the audit. The baseline describes your data as it is, which is the only way a later improvement can be measured honestly.
Your owners agree the impact
Impact scores are set with the people who run the process, not by us alone. That is why the roadmap survives the first budget meeting.
The roadmap is not a sales funnel
Some findings belong in the Monitor tier, some fixes are a single validation rule your admin can add. We say so. When you want help with the rest, ongoing Salesforce support is a standing offer.
Data quality audit questions, answered plainly
A Salesforce data quality audit is a fixed-scope assessment that measures the condition of the records in your org at a point in time and ranks what to fix. Twopir Consulting measures duplicates, completeness, consistency, accuracy and validity on the objects in scope, records a dated baseline, scores every finding, and delivers a prioritised roadmap. It is diagnostic only: nothing is merged, updated or deleted.
No. Cloudingo filters behave like reports and change nothing until a merge is submitted, and during the audit we submit no merges, schedule no automation jobs and run no mass updates, deletes or imports. The org is left as we found it, so the baseline describes your data rather than our changes. Any fix is scoped afterwards as separate work.
No. Cloudingo provides a dashboard with duplicate counts, filters that report duplicate groups, field analysis, pre-merge reports and, on every plan, an analysis of the current health of your Salesforce data. It does not produce a scored audit report. The scoring and prioritisation in a Twopir audit are our own method, which weighs business impact, reach, trend and effort for every finding.
If you already run Cloudingo, we use your connection. If you do not, we agree how it is licensed for the audit window during scoping. Cloudingo offers a 10-day free trial and says it shows how many duplicates you have within minutes. Cloudingo's custom-object merging, which covers Opportunities, Cases, Campaign Members and true custom objects, needs the Professional or Enterprise plan and is enabled by Cloudingo on request; where those objects are in scope without it, we measure them with Salesforce reports instead.
It depends on the number of objects in scope, record volumes and how many process owners need to agree the impact scores, so Twopir fixes the duration when the scope is agreed, before the audit starts. An audit of Leads, Contacts and Accounts with one owner per process is the shortest. Custom objects, several business units and an accuracy sample across many fields add time.
Cloudingo runs in the cloud and connects to Salesforce through an OAuth connection over the API, and it states that customer data is neither stored nor cached. Cloudingo also lists SOC 2 Type II compliance and 256-bit SSL encryption. We check the connected user's permissions before measuring, and filter exports and accuracy samples are kept in storage your team controls.
You receive the baseline, the duplicate analysis, the field health map, the accuracy sample, the scored findings register and the prioritised roadmap at a readout with your owners and admins. Each roadmap item names the work that fixes it: Cloudingo setup, deduplication, cleansing, a source fix or an ongoing data quality program. Your team, Twopir or another partner can take it from there, and the baseline lets anyone measure the result.
Next Step
Measure it first, then decide the fix
A first conversation covers the objects that matter, the teams that disagree about the data, how records arrive, and whether Cloudingo is already in place: enough to scope the audit and say what it will answer. No obligation to proceed.
Talk to a Salesforce architect who audits with Cloudingo
