One value, five spellings
Texas, TX and tx in the same state field. Phone numbers with and without brackets. A picklist that was free text for years. Reports split one group into five and filters miss the rest.
Salesforce · Cloudingo Data Cleansing
Cloudingo data cleansing services that keep Salesforce clean
Salesforce data cleansing with Cloudingo means standardising inconsistent values, filling and correcting incomplete records, retiring stale ones and stopping bad data at import, using Cloudingo's mass update, mass delete, transform and import tools. Twopir Consulting writes the data standards, runs each change on a tested sample first, and hands over a maintenance cadence your admins can keep. Records your reports, routing and automation can trust.
12+
500+
250+
40+
Trusted by 500+ organizations — including clinical, legal and professional-services businesses whose routing, reporting and outreach depend on clean Salesforce data.
What We Cleanse
- Salesforce Partner
- Leads
- Contacts
- Accounts
- Person Accounts
- Custom Objects
- Import Files
- U.S. Addresses
Bad data rarely arrives all at once
Most Salesforce orgs do not need rescuing from one disaster. They drift: one import with its own formats, one integration that leaves a field blank, one team that keeps test records. Each is small. Together they break routing rules, reports, territory assignment and every filter that relies on a value being written one way. These are the six patterns we clean most often.
Company names nobody agrees on
Acme, Acme Inc. and The Acme Company on three accounts. Account-based reporting, territory rules and any integration keyed on the name all treat them as different businesses.
The fields that drive routing are blank
Industry, region, country or lead source left empty by a form or an integration. Assignment rules fall through to a default queue, and segmentation quietly leaves those records out.
Stale and test records kept forever
Inactive leads untouched for years, test records from an old project, contacts at companies that no longer exist. They inflate counts, skew conversion rates and take up storage you pay for.
Imports that undo last quarter's cleanup
An event list or purchased file loaded straight in, with its own casing, its own country names and no check against existing records. A cleanse without import hygiene has a short shelf life.
Records owned by people who left
Accounts and open leads still assigned to inactive users. Nobody follows up, nobody updates them, and every dashboard grouped by owner carries a row of records that belong to no one.
Definition
Cloudingo data cleansing, defined
Cloudingo data cleansing is the work of making existing Salesforce records consistent, complete and current with Cloudingo's data-maintenance tools: standardising and transforming values, filling missing data, reassigning ownership, validating U.S. mailing addresses, deleting stale records in bulk, and checking new files against existing records before they are imported. It works on leads, contacts, accounts, person accounts and custom objects, and it is ongoing maintenance, not a one-off project.
Three parties are involved. Cloudingo, an AppExchange app from Symphonic Source, runs in the cloud outside your org, connects over OAuth and provides mass update, mass delete, find and replace, transforms, find-and-export, field analysis, the import wizard and the U.S. address validation add-on. Salesforce stores the data and enforces what is enforced at entry: validation rules, required fields, picklists and your automation, which runs on every record a mass change touches. Twopir Consulting decides what clean means for your org, writes it down as standards, and runs and hands over the cleanse. The tool's own description of these features is on Cloudingo's features page.
Cleansing and deduplication are related but different jobs. Cleansing fixes the values on each record; deduplication decides which of two records survives a merge. We usually standardise first, because matching filters work better on consistent values, then hand duplicate groups to our Cloudingo Salesforce deduplication services. If you are using a different tool for the same work, we also run Validity DemandTools cleansing.
New to the product? Our getting-started guide to Cloudingo covers the basics. Installing, permissioning and configuring the app itself is covered by our Cloudingo implementation services.
One field, one way of writing it
Standardisation sets one accepted format per field; normalisation rewrites the existing values into it. We agree the standard with the people who report on the field, then apply it in bulk with Cloudingo and lock it in at entry with Salesforce, so the same variants do not come back.
| Field | Standard and how we apply it |
|---|---|
| State and country | One form, code or full name, chosen per field. Existing values are rewritten with transform or find and replace; Cloudingo's import transforms convert names to codes, or codes to names, on the way in. |
| Phone | One layout, with extensions held one way. Import transforms format brackets, dashes, spacing and extension prefixes; existing numbers are normalised in bulk. |
| Names and titles | Consistent casing for people, job titles and cities. Text transforms set upper, lower or title case, so an all-caps list stops looking like a different dataset. |
| Company names | One agreed form per company, with legal suffixes treated the same way across accounts. Cloudingo's mass update is documented for normalising company names. |
| Free-text picklist values | Old free-text values mapped to the current picklist with find and replace, before the field is restricted in Salesforce. |
| U.S. mailing addresses | Standardised and appended by Cloudingo's address validation, an add-on on every plan. Non-U.S. addresses need a separate service, which we scope with you. |
Standardised values also make duplicate matching more reliable: a filter comparing state or phone finds far more true matches once every record is written the same way. The standards document is the first artefact we hand over, and the reference for every change that follows.
Fill it, fix it, or retire it
Every record that fails a standard gets one of three outcomes: the missing value is filled from a source you trust, the wrong value is corrected, or the record is retired. The decision is made per segment and written down, not made record by record during the update.
Fill missing values
Blank industry, region or source fields filled in bulk with Cloudingo's mass update, from a file, a parent account or a rule your team agrees.
Correct wrong values
Values known to be wrong replaced with find and replace or a targeted mass update, scoped by filter so only the affected records change.
Reassign ownership
Records held by inactive users or catch-all queues reassigned in bulk to the right owner, by territory, account or the rule sales operations sets.
Validate U.S. addresses
Mailing addresses standardised and appended with Cloudingo's address validation add-on, so territory, tax and shipping logic read one format.
Match a file to records
Find and export matches a spreadsheet, such as a list of names from finance or an event, to Salesforce records and returns their IDs or other fields for the update.
Retire stale records
Test records, long-inactive leads and worthless records removed with mass delete, after the agreed retention check and an export of everything deleted.
Analyse field usage
Cloudingo's field analysis reports the number, type and usage of fields on any object, which shows the fields worth cleaning and the ones to retire from layouts.
Custom objects too
The same work on the custom objects your business runs on, within what your Cloudingo plan supports and what the connected user is allowed to edit.
Bulk changes with a way back
A mass update can change thousands of records in minutes, and every one of them fires the flows, triggers and validation rules on that object. Cloudingo's Undo & Restore is documented for merges, so for updates and deletes the way back is the export we take before the change.
| Operation | Use and safeguard |
|---|---|
| Mass update | Fill, correct or reassign a filtered set. We run it on a small sample first, check the records and the automation that fired, then run the rest in batches. |
| Find and replace | Swap known-bad values for the standard. The list of replacements is reviewed by the field owner, because one broad pattern can rewrite values that were correct. |
| Update from a file | The import wizard's update-only mode writes file values straight to records by Salesforce ID, without a duplicate scan. We check that every ID and every column is the one intended. |
| Mass delete | Remove test, inactive and worthless records. Retention rules, related records and integrations that reference them are checked first, and a full export is kept. |
| Find and export | Pull IDs and current values for a set of records. We use it to keep the before-state of every change, so any field can be written back. |
Clean data in, not cleaned up later
Cloudingo's import wizard checks a CSV file against existing records before anything is written, applies transforms as columns are mapped, and lets you preview every merge and every new record. We set it up so routine imports follow the same standards as the cleansed data.
| Stage | What Cloudingo does | What we set |
|---|---|---|
| Choose the object | Imports against leads, contacts, accounts, person accounts or a custom object enabled for dedupe, or scans contacts then leads in one pass, looking up accounts by name with city and state. | Which object each routine file belongs to, the default record type, and whether a new contact may create a new account. |
| Map and transform | Maps columns to fields and applies transforms: text case, phone format, extensions, ID lookups, and state or country names to codes. | Saved mappings per file source, with the transforms that match your data standards. |
| Preview and edit | Runs pre-validation checks on field values and allows in-line edits and row deletions before the scan. | What the importer must fix in the file and what can be corrected in the preview. |
| Scan for matches | Scans the file against a saved or one-time filter, with a rule deciding which values the import may overwrite. | The filter and overwrite rule per source, so a list cannot replace a better value already in Salesforce. |
| Review, then push | Shows unique records, Salesforce matches, duplicates within the file and errors; merges and pushes can wait for review, and file-only duplicates are left for a person by default. | Automatic merge and push left off for new file formats until they have been reviewed once. |
| Make it routine | Rapid Import adds shared import templates, permission-limited import users and automated deduplicated imports, on the Professional and Enterprise plans. | One template per recurring source, and import rights limited to the people trained on them. |
Rules that stop the drift coming back
A cleanse fixes the records you have. Rules protect the records you will get. Some rules belong in Salesforce, at the point of entry; some belong in Cloudingo, at import; and some are a routine a person runs. We put each standard where it is cheapest to enforce.
| Rule | Where it lives | What it prevents |
|---|---|---|
| Restricted picklists | Salesforce, including state and country picklists where your org uses them. | New free-text variants in fields that have a fixed list of answers. |
| Validation rules | Salesforce, written to the format in the data standards. | Values entered by hand or by an integration in the wrong format. |
| Required fields by stage | Salesforce validation rules and page layouts. | Records moving forward with the fields routing and reporting need still blank. |
| Import transforms | Cloudingo saved mappings and Rapid Import templates. | A new list arriving in its own casing, phone layout or country names. |
| Import rights | Cloudingo roles and permissions. | Untrained users loading files outside the agreed templates. |
| Source fixes | The form, integration or process writing the bad value. | The same gap reappearing every week, however often it is cleaned. |
Stopping duplicates at entry is a separate question, with its own tools and trade-offs. It is covered with matching and merge rules in our deduplication work, linked in the definition above.
Six steps from messy fields to a maintained org
Each step ends with something your team signs off before the next starts. Nothing is changed in bulk until the same change has been run on a sample and checked in Salesforce.
Profile & Prioritise
We profile the fields that matter to routing, reporting and integrations with field analysis and filters, and rank the fixes. If you need a scored baseline first, that is our Cloudingo Salesforce data quality audit.
DeliverablePrioritised cleansing backlog
Standards & Safeguards
Field owners agree the target formats and the fill, fix or retire decision per segment. We confirm your backup, review the connected user's field access and list the automation on each object.
DeliverableSigned-off data standards
Standardise
Formats and values normalised with transforms, find and replace and mass update, on a sample first, then in batches, with the before-state exported each time.
DeliverableChange log with exports
Complete & Correct
Missing values filled, wrong values corrected, owners reassigned and U.S. addresses validated where the add-on is licensed. Each run is checked against the backlog it closes.
DeliverableClosed backlog items
Retire
Test, inactive and worthless records deleted under the agreed retention rules, after a check for related records and integrations that still reference them.
DeliverableDeletion export and sign-off
Protect & Hand Over
Picklists and validation rules in Salesforce, import templates and permissions in Cloudingo, and the maintenance cadence below, written as a runbook your admins can follow.
DeliverableRules in place and admin runbook
A cadence your admins can actually keep
Data quality decays the day a cleanse ends unless someone owns the routine. Cloudingo's own blog cites the 1-10-100 rule, an industry rule of thumb that preventing a bad record costs less than cleaning it later, and far less than leaving it. The cadence below is how we turn that into a calendar.
| Cadence | What runs | Who owns it |
|---|---|---|
| Every import | The saved template for that source, with its transforms, scan and review before push. | The trained importer for that source |
| Daily or weekly | Cloudingo's scheduled automation jobs for proven filters, and a review of the import and job activity on the dashboard. | Salesforce admin |
| Monthly | A sweep of the priority fields for new blanks and variants, owner reassignment for leavers, and fixes for whatever the rules let through. | Admin, with field owners |
| Quarterly | Stale and test record retirement, field analysis on the main objects, a review of Cloudingo Intelligence org-health alerts, and a check of record counts, since Cloudingo adds a record-count fee above 300,000 records. | CRM manager |
| On business events | A focused cleanse before a migration, a new integration, an acquisition or a territory change, and an update to the data standards after it. | CRM manager, with Twopir |
Who decides the standards, how exceptions are approved and how data quality is reported to the business is a governance question. That operating model is the subject of our Cloudingo Salesforce data quality consulting.
Cleansing run by people who build orgs
A bulk update is a change to production. The fields it touches feed routing, automation, integrations and reports, so the person running it needs to understand the org as well as the tool.
We know what each field feeds
Before a value changes we trace where it is used: assignment rules, flows, reports, integrations and outbound syncs. That is what keeps a cleanse from breaking something downstream.
Standards come from your teams
What counts as clean is a business decision. We run short sessions with the owners of each field and turn their answers into written standards before any tool is opened.
Every bulk change can be reversed
An export before each change, a sample run checked in Salesforce, then batches. If a value needs to go back, the before-state is there to write it back.
We fix the source, not only the records
Where a form, an integration or an import keeps writing the bad value, we change it there. Otherwise the same cleanse has to be repeated every quarter.
Independent of the vendor
Cloudingo is licensed from its vendor; we design and run the work around it. If a Salesforce validation rule or a different tool is the better fix for a problem, we say so.
Cloudingo cleansing questions, answered plainly
Field profiling and a prioritised backlog, written data standards agreed with your field owners, then standardisation, filling and correcting of missing and wrong values, owner reassignment, U.S. address validation where licensed, and retirement of stale and test records, each run on a sample first with an export taken. It ends with entry rules in Salesforce, import templates in Cloudingo and a maintenance runbook. Merging duplicates is a separate deduplication service.
Cleansing fixes the values on each record: formats, missing fields, wrong values, ownership and records that should no longer exist. Deduplication finds records that describe the same person or company and merges them into one. The two support each other, and we usually standardise first because matching filters find more true duplicates once values are written one way.
Cloudingo's Undo & Restore is documented for records merged in Cloudingo, not for mass updates or deletes. So before every bulk change we export the records and fields being changed, with their Salesforce IDs, and confirm your own Salesforce backup. If a value has to go back, the export is used to write it back.
Cloudingo's address validation standardises and appends U.S. mailing addresses and is an add-on on every plan. Phone numbers can be reformatted with its transforms, which is formatting rather than verification. For non-U.S. address validation, or for email and phone verification, we scope a separate service alongside Cloudingo.
With rules in three places. Salesforce enforces formats at entry with restricted picklists, validation rules and required fields. Cloudingo's import templates apply the same transforms and checks to every routine file, and import rights are limited to trained users. Then a maintenance cadence, from each import to quarterly reviews, catches what the rules let through, and any form or integration still writing bad values is fixed at the source.
It is when the deletion follows agreed retention rules. We define what stale means per object with the owners, check for related records and for integrations that still reference the records, export everything that will be deleted, and run a sample first. Cloudingo's mass delete then removes the rest, which also reduces the storage and record counts you pay for.
Continuously in small pieces rather than in one large project every few years. A practical starting cadence is a template-driven check on every import, scheduled jobs daily or weekly, a monthly sweep of priority fields and leavers' records, a quarterly retirement of stale records and field review, and a focused cleanse before any migration, new integration or acquisition.
Next Step
Bring the messy fields, we plan the cleanse
A first conversation covers which fields cause the most trouble, where the bad values come from, which Cloudingo plan and add-ons you hold and what has already been tried: enough to outline the standards, the order of work and what the cleanse would take. No obligation to proceed.
Talk to a Salesforce architect who runs Cloudingo data cleansing
