You have extracted the raw data you need. Your robot ran, the extraction looks good, and your table is full of product listings, lead contacts, or real estate prices.
But the data isn't quite ready to use.
Prices still have currency symbols and commas. Names are split across two columns. You need to flag which products are in stock and which aren't. And someone on your team is asking for profit margins, not raw numbers.
This is the part where most people export to Google Sheets or Excel and start writing formulas there. It works, but it adds steps, breaks your workflow, and means your data transformations don't carry over to the next extraction.
Browse AI now handles all of this inside the platform, with calculated columns and a new AI-powered formula assistant called Formula AI. You can clean, transform, categorize, and enrich your data right in your table, and it runs automatically on every extraction going forward.
Here are the most common data transformation problems our users solve, and how to tackle each one.
1. Cleaning prices and converting currencies
This is the most common data transformation we see. Scraped prices almost always come through with currency symbols, commas, or text that needs to be stripped out before you can do anything useful with the numbers.
The problem: Your Price column shows values like "CAD 1,299.99" or "$49.95 USD" and you need clean numbers.
With Formula AI: Open a new calculated column and type:
Clean up the price column to remove any text and convert to a number.
Formula AI generates a formula that strips non-numeric characters and converts the result to a number. You see the cleaned values instantly in the example results preview before you save.
Going further: Need to convert to a different currency? Follow up with: "Multiply the result by 0.73 to convert CAD to USD." Formula AI remembers the conversation context and builds on what it already created.
Manual formula example:
VALUE(REGEXREPLACE({Price}, "[^0-9.]", ""))
2. Categorizing products, leads, or listings
Once you have data, you usually need to segment it. Which products are budget vs. premium? Which leads are hot? Which listings are worth following up on?
The problem: You have a column with prices, review counts, or scores and need to label each row based on business logic.
With Formula AI:
"If the price is under $50, label it 'Budget'. If it's between $50 and $200, label it 'Mid-range'. Otherwise label it 'Premium'."
"If the review count is over 100, mark it as 'Popular'. Otherwise mark it as 'Standard'."
"If the company size is over 100 employees and budget is over $50,000, label as 'Hot Lead'. If either is above the midpoint, label as 'Warm Lead'. Otherwise 'Cold Lead'."
These kinds of multi-tier categorizations would normally require nested IF statements. Formula AI handles the complexity for you.
3. Combining or splitting text fields
Names split across first and last columns. Addresses that need merging. URLs you need to break apart. Text manipulation is a constant need when working with scraped data.
The problem: You have separate "First Name" and "Last Name" columns and need a full name. Or you have a URL and need just the domain.
With Formula AI:
"Combine first name and last name with a space between them"
"Extract just the domain name from the URL column"
"Get the first word from the job title column"
Manual formula examples:
CONCATENATE({First Name}, " ", {Last Name})
REGEXEXTRACT({URL}, "^(?:https?://)?([^/]+)")
4. Extracting emails and contact info from messy text
If you're scraping directories, profile pages, or contact sections, the data sometimes comes through as a block of text with the email buried inside it.
The problem: Your "Contact" column contains a mix of text, phone numbers, and email addresses. You just need the email.
With Formula AI:
"Extract the email address from the contact column"
Manual formula example:
REGEXEXTRACT({Contact}, "[a-zA-Z0-9._%+-]+@[a-zA-Z0-9.-]+\.[a-zA-Z]{2,}")
This works on HTML content too. If your scrape captured raw HTML with mailto links, Formula AI can parse those out just as easily.
5. Calculating margins, discounts, and percentages
Price comparison and competitive intelligence are among the most popular Browse AI use cases. Once you've scraped competitor prices or product costs, you usually need to calculate something from them.
The problem: You have an original price and a sale price and need the discount percentage. Or you have cost and selling price and need the margin.
With Formula AI:
"Calculate the discount percentage between the original price and sale price columns"
"Calculate profit margin as a percentage: selling price minus cost, divided by selling price, times 100"
Manual formula example:
ROUND(({Original Price} - {Sale Price}) / {Original Price} * 100, 1)
6. Creating stock status and availability flags
E-commerce monitoring often comes down to one question: is it in stock or not? Scraped data might give you a quantity, a text label, or a mix of both.
The problem: Your Quantity column has numbers and you need a clear status label for your dashboard or alert system.
With Formula AI:
"If quantity is more than 10, label 'In Stock'. If it's between 1 and 10, label 'Low Stock'. If it's 0, label 'Out of Stock'."
This pairs well with Browse AI's monitoring feature. Set up a monitor, and every extraction automatically gets the status label applied through your calculated column.
7. Validating and flagging incomplete data
When scraping at scale, some rows will inevitably have missing fields. Rather than manually scanning your table, you can create a calculated column that checks for completeness.
The problem: You need to quickly identify which rows are missing critical data like email, phone, or company name.
With Formula AI:
"If the email or phone column is blank, label the row as 'Incomplete'. Otherwise label it 'Complete'."
Manual formula example:
IF(OR(ISBLANK({Email}), ISBLANK({Phone})), "Incomplete", "Complete")
How Formula AI works
If you haven't tried it yet, here's the quick version:
- Go to your robot's Tables view and click add calculated column.
- Type what you want in the AI prompt field (the one with the sparkle icon).
- Review the generated formula and the Example results preview line, which shows real output from your data.
- Not quite right? Type a follow-up. The AI remembers your conversation and refines the formula.
- Click Add column when you're happy with the results.
The formula runs automatically on every row, including all future extractions. No exports, no spreadsheets, no code.
For a full walkthrough with screenshots, check out our help center guide: How to use Formula AI to create calculated columns.
Transform your data where you extract it
The traditional workflow of scrape, export, clean in a spreadsheet, re-import creates friction and breaks the moment something changes. Calculated columns and Formula AI keep everything in one place.
Your formulas run on every extraction. Your transformed data flows through to Google Sheets, Airtable, or any of Browse AI's 7,000+ integrations via Zapier. And anyone on your team can create formulas by describing what they need in plain English.
If you're already using Browse AI, open any robot's Tables view and try it. If you're new, sign up for free and see how far you can go without writing a single line of code.
