These are the best practices from experienced TexAu users. They will help you work more efficiently, save money, and keep your data clean.
Data quality: start right
If your input data is messy, your enrichment results will be too. Here is how to keep your data clean.
Clean your data before importing
Before you upload anything to TexAu, take 30 minutes to clean it.
-
Remove duplicates:
- Use your spreadsheet's "Remove Duplicates" feature before exporting to CSV
- Duplicates waste credits, and you do not want to enrich the same company twice
- This alone saves 10-20% in credits
-
Remove incomplete rows:
- Delete rows where critical fields are blank (company name, email, domain)
- An enrichment cannot succeed if the input data is missing
- Look for obviously wrong data (emails without @, phone numbers with letters)
-
Standardize formatting:
- Phone numbers: use the same format consistently (e.g., 123-456-7890)
- Emails: all lowercase, no extra spaces
- Company names: no extra symbols or misspellings
-
Validate URLs:
- Make sure website URLs start with https:// or http://
- Remove trailing slashes
- Test a few manually. Do they load in a browser?
-
Segment by quality tier:
- Grade your data: A (high confidence), B (medium), C (low/incomplete)
- Enrich A-tier first to build momentum
- Try different providers on B and C data
Tip: export a "before" version as backup before any cleanup. That way, if you accidentally delete something important, you can restore it.
Use semantic data types correctly
TexAu has 32+ semantic data types (Email, Company Name, Phone, etc.), and using the right ones matters a lot.
Why: different data types are enriched by different providers. If you label a column wrong, TexAu will not know how to enrich it.
How to apply them:
-
When uploading, TexAu auto-detects types. Review them:
- Email addresses should be labeled "Email"
- Company names should be labeled "Company Name"
- Domains/websites should be labeled "Company Website"
- Phone numbers should be labeled "Phone Number"
-
Edit if needed:
- Click a column header, then change the type
- Make sure the type matches the data inside
- Do not label a column "Email" if it contains company names
-
Create custom types for your business:
- If you track "Industry Code" or "Account Tier," create custom types
- This helps formulas and integrations understand your data
Common mistakes to avoid:
- Labeling a "Competitor List" column as "Company Name" (confuses enrichment)
- Putting multiple data types in one column (email AND phone in "Contact Info")
- Instead, create separate columns for each data type
Validate inputs regularly
Set up a quick validation process:
- Before enriching a batch, spot-check 20-30 rows
- Look for patterns in bad data (missing domains, typos)
- Fix the most common issues
- Then enrich
Use a simple formula to flag suspicious data:
IF(Email = "" OR Email NOT CONTAINS "@", "REVIEW", "OK")IF(Company_Name = "", "MISSING", "OK")- This highlights rows to review before enrichment
Enrichment strategy: be intentional
Start with small batches
Do not enrich your entire 100k row table on day one. Here is why:
- Learn what works: test with 100-1,000 rows first
- Catch data quality issues early: if 50% of results are empty, you need better input data
- Save money: small batches let you refine strategy before spending big
Your first enrichment should be:
- A subset of your best, cleanest data
- One enrichment type (just email, not email + phone + company)
- One data provider (not your full waterfall yet)
Review results:
- Success rate: aim for 70-85% on first try
- Cost per successful enrichment: track this to optimize later
- Data quality: does the enriched data look right?
Once you are comfortable, expand to larger batches and more complex enrichments.
Use waterfall for important data
Waterfall enrichment tries multiple providers in order, giving you the best chance of finding data while keeping costs down.
Set up your waterfall like this:
-
Order by accuracy, then cost:
- Try the most accurate provider first
- Then cheaper but slightly less accurate providers
- Then backup providers for edge cases
-
Example waterfall for email enrichment:
- Provider 1: High-accuracy premium provider
- Provider 2: Good accuracy, mid-tier pricing
- Provider 3: Fast and affordable backup provider
- Result: if Provider 1 finds it, done. If not, try Provider 2. If that fails, Provider 3 is your backup.
-
Why this works:
- Most emails are found by your primary provider
- A smaller percentage needs the second provider
- Only a few percent need the third provider
- Your blended cost is much lower than using the premium provider alone
-
Set limits:
- Use "Stop on First Success" if you just need ANY result
- Use "Max Providers to Try" to control costs (e.g., try max 3 providers)
- This prevents burning credits on low-value attempts
Real example: a sales team enriching 10,000 leads for email found that waterfall enrichment saved roughly 50% compared to using a single premium provider, with the same overall match rate.
Prioritize by ROI
Not all enrichments are equally valuable. Prioritize:
- High-touch accounts first: enrich your biggest customers or hottest leads before your long tail
- Critical fields first: email and company name are more useful than industry or company size
- Revenue-generating departments first: sales gets enrichment before marketing
- Test on small segments: before enriching all 50k records, test on 1,000
Ask yourself: "If this enrichment succeeds, what happens next?" If the answer is "nothing," maybe skip it.
Formulas: master the basics
Formulas let you create new columns by combining and transforming existing data. Here is how to use them well.
Start simple, build up
Do not jump into complex nested formulas. Build progressively:
-
Day 1: master basic functions
CONCAT(First_Name, " ", Last_Name)—Full nameUPPER(Company_Name)—UppercaseLEN(Email)—Email length
-
Day 2: add conditionals
IF(Revenue > 1000000, "Enterprise", "Mid-Market")
-
Day 3: combine multiple functions
IF(AND(Revenue > 1000000, Industry = "Tech"), "HVT", "Standard")
Each step should take a few minutes to learn and test.
Test with sample data
Before running a formula on your entire table:
- Create a test row (copy an existing row, change one detail)
- Run the formula on just that row
- Review the result —does it match what you expect?
- Fix any issues
- Then apply to your full table
This catches 90% of formula mistakes before they affect thousands of rows.
Use COALESCE for waterfall results
When you enrich a column with multiple providers, each provider creates its own result column. Use COALESCE() to pick the first non-empty result:
Example: you enriched email with three providers, creating three result columns. Create a final column called Final_Email:
COALESCE(Provider1_Email, Provider2_Email, Provider3_Email, "Not Found")
This uses the first provider's result if available, falls back to the second, then the third, and finally shows "Not Found" if all three are empty.
This is clean, reliable, and handles all edge cases.
Use IF for conditional logic
IF() statements let you categorize or flag data:
IF(Email = "" OR Email NOT CONTAINS "@", "INVALID", "VALID")
IF(AND(Revenue > 500000, Company_Age > 5), "QUALIFIED", "UNQUALIFIED")
IF(OR(Industry = "Tech", Industry = "Finance"), "HOT", "WARM")
Use these to:
- Flag data quality issues
- Segment leads by criteria
- Create scoring models
- Clean up data automatically
Team organization: scale across people
As your team grows, good organization prevents chaos.
Set up role-based access
Assign roles based on what people actually need:
-
Owner/Admin: only 1-2 people (usually ops manager + founder)
- Can manage team, change settings, delete tables
- They control costs and security
-
Editor: your working team (sales ops, marketing ops, analysts)
- Can create and run enrichments
- Cannot delete tables or manage team
- Prevents accidental disasters
-
Viewer: stakeholders who just need results (sales managers, directors)
- Can see data and reports
- Cannot edit or run enrichments
- Keeps them informed without risk
Use naming conventions
Establish rules for naming tables, columns, and formulas:
Table names:
- Good: "2024-Q1-Leads-Sales" (date, purpose, owner)
- Bad: "New Data" (too vague)
Column names:
- Good: "Provider1_Email", "Provider2_Phone", "Manual_Company_Website"
- Bad: "email1", "email2", "email3" (which provider is which?)
Formula columns:
- Good: "Final_Email", "Lead_Score", "Outreach_Ready"
- Bad: "Formula1", "Formula2"
When your team member inherits your table, they should understand it in 2 minutes, not 2 hours.
Organize by project or campaign
Structure your workspace logically:
-
Project 1: "Sales Prospecting Q1 2024"
- Table: "Clean Leads - Ready for Enrichment"
- Table: "Enriched Results"
- Table: "Follow-up Prospects"
-
Project 2: "Customer Success - Expansion"
- Table: "Existing Customers"
- Table: "Enriched with Firmographics"
This makes it easy to find what you need and archive old campaigns.
Cost optimization: spend smart
Use Stop on First Success
If you just need any result (you do not care which provider finds it), enable "Stop on First Success":
- Enrichment stops after the first provider returns data
- Prevents wasting credits on unnecessary fallback providers
- Use when: Any email is better than none, or you just need to confirm a company exists
Compare with "Try All Providers" (when you want the best result):
- Keeps trying until all providers are exhausted
- Use when: You need the most accurate email or highest confidence data
Savings: 20-40% on credits by choosing the right mode.
Limit max providers to try
Instead of trying all 5 providers in your waterfall, limit to 2-3:
- Most data is found by the first 2 providers
- The 4th and 5th attempts rarely find new data
- Limiting prevents wasting credits on edge cases
Schedule enrichments during off-peak hours
TexAu runs faster when fewer people are using the system:
- Schedule large enrichments for evenings or weekends
- Your job completes faster
- Leaves daytime resources for real-time work
How to schedule:
- Go to your table, then open the Schedules section
- Create a schedule for early morning (e.g., 2am or 3am your time)
- TexAu automatically enriches new rows at that time
Use standard AI models for simple tasks
When you use AI-powered processing actions:
- Standard models are fast and cost-effective. Good for simple tasks like text classification, basic extraction, and straightforward summarization.
- Premium models have higher accuracy and are better for complex analysis, nuanced writing, and multi-step reasoning.
Smart allocation:
- Use standard models for simple text classification ("Is this a tech company?")
- Use standard models for basic summarization and medium-complexity tasks
- Reserve premium models only when you need the absolute best accuracy
This approach saves 30-50% on AI processing credits.
Automation: set it and forget it
Set up scheduled jobs for recurring enrichments
If you regularly import new leads or customer data:
- Go to your table, then open the Schedules section
- Create a schedule (daily, weekly, or custom interval)
- TexAu automatically enriches new rows on that schedule
- You only pay for new rows, not re-enrichment of existing data
Example: your sales team imports 500 new leads every Monday. Set a schedule for Monday at 3am. TexAu enriches all new leads before the team arrives. No manual work needed.
Use webhooks to push data to your tools
Instead of manually exporting and importing data, use webhooks:
- Enrich a lead in TexAu
- Webhook automatically sends it to your CRM
- Your sales team sees it instantly
Setup:
- Go to your table, then open the Webhooks section
- Add a destination URL (provided by your receiving tool)
- Map which columns to send
- Save
Now enrichment results sync automatically. One less manual step.
Combine formulas and processing actions
Use both together for more automated workflows:
-
Formula: clean and categorize data
IF(Revenue = "", "UNKNOWN", Revenue)—Fill blanksIF(Industry = "Software", "TECH", "OTHER")—Categorize
-
Processing action: use AI to analyze
- Summarize company descriptions
- Extract keywords from text
- Classify company types
-
Result: fully enriched, cleaned, categorized data ready for action
Security: protect your data
Use secure authorization over API keys
When connecting integrations:
- One-click secure authorization is safer, easier to revoke, and requires less maintenance.
- API keys are more flexible but riskier if exposed.
Use secure authorization whenever it is available.
Review audit logs regularly
Your audit log shows who did what and when:
- Go to Settings, then Audit Logs
- Review weekly or monthly
- Look for:
- Unexpected data exports
- Changes to team members
- Large enrichments you did not authorize
Catch issues early before they become problems.
Remove inactive team members
Every person with access to your workspace is a potential security risk:
- Go to Settings, then Team
- Remove anyone who does not actively use TexAu
- When people leave your company, remove them immediately
Do not let inactive accounts pile up.
Enable two-factor authentication
Add another layer of security to your account:
- Go to your profile settings and enable two-factor authentication
- Use an authenticator app on your phone for the strongest protection
- This prevents unauthorized access even if your password is compromised
Summary: your action plan
This week:
- Export your data and back it up
- Clean duplicates and obvious bad data
- Assign semantic data types correctly to one table
- Test a small enrichment (100-500 rows)
Next week:
- Establish a naming convention for your team
- Set up role-based access for team members
- Create your first waterfall enrichment strategy
- Set up a scheduled job for recurring data
This month:
- Build 2-3 formulas to automate data transformation
- Set up a webhook to sync data to your CRM
- Audit your spending and optimize cost per enrichment
- Document your processes for your team
You do not need to do everything at once. Start with data quality, then automation, then optimization.
Questions? Reach out to our support team.
Next steps
- Having a problem? Visit the Common Issues and Troubleshooting guide.
- Browse the Frequently Asked Questions for quick answers.
- Return to the Help Center home page for more guides.