The Scenario
The performance marketing review is tomorrow and you have 80 landing page copy snippets in column A of the 'Copy Variants' sheet in an Excel workbook — every variant from the last campaign cycle. Your job is to tell the team which pages are targeting which audiences, based on what the copy is actually about, not what someone on the creative team said it was about three months ago. You need the dominant IAB content category and the top named entity for each snippet, so you can map them against the audience segments in the media plan.
The bad version:
- Open the TextRazor API, paste the first copy snippet, read through the entity and category arrays in the response, note the top IAB category and the entity with the highest relevance score, enter both into columns B and C.
- Move to row 2. Notice that this snippet's top entity is a brand name you already know — not very useful — and start wondering whether you should filter out brand names from the entity list or just note the second entity.
- Make an ad hoc call and keep going, but now column C has a mix of brand entities and product entities, and the methodology isn't consistent across rows.
By row 20 you've introduced three different entity-selection rules that you're applying inconsistently. The report you hand to the team tomorrow will have methodology questions you won't be able to answer cleanly.
The Easy Way: One Prompt in SheetXAI
SheetXAI is an AI agent that lives inside your Excel workbook. It reads the copy snippets, understands the column, and through its built-in TextRazor integration can run full analysis across every row with a consistent extraction rule applied uniformly.
Analyze the copy in column A of the 'Copy Variants' sheet with TextRazor full analysis and for each row write the dominant content category, top 3 entities, and their confidence scores into columns B through E.
What You Get
- Column B receives the top IAB content category for each snippet — e.g., "Health and Fitness/Running and Jogging" or "Business and Finance/Investing."
- Columns C, D, and E receive the top three entities by relevance score and their confidence values.
- The extraction rule is applied identically across all 80 rows — no per-row methodology drift.
- The output is ready to join against the media plan's audience segment taxonomy without reformatting.
What If the Data Is Not Quite Ready
The top entity is always the brand name, which isn't useful
For each snippet in column A of the 'Copy Variants' sheet, use TextRazor full analysis, exclude any entity that matches the brand names listed in column A of the 'Brand List' sheet, and write the top remaining entity by relevance score into column B and the top IAB category into column C.
Some snippets are duplicated across variants
Identify and remove duplicate copy snippets from column A of the 'Copy Variants' sheet before running TextRazor analysis. For unique rows, write the top IAB category into column B and the top entity into column C. Note which rows were duplicates in column D.
Some snippets include HTML markup from a CMS export
For each snippet in column A of the 'Copy Variants' sheet, strip HTML tags before sending to TextRazor full analysis. Write the dominant content category into column B and the top entity into column C.
Full kill chain: clean, score, audience map
For each snippet in column A of the 'Copy Variants' sheet, strip HTML tags and trim whitespace, run TextRazor full analysis, write the top IAB category and top entity into columns B and C, and then add an 'Audience Map' sheet grouping the 80 pages by IAB category, showing the most common entities within each category, and flagging any category that appears fewer than 3 times as likely miscategorized.
The audience map is what the review needs — collapsing cleanup, scoring, and aggregation into one prompt means you show up with the answer ready.
Try It
Get the 7-day free trial of SheetXAI and open the Excel workbook with your landing page copy variants in the 'Copy Variants' sheet. Ask SheetXAI to run TextRazor full analysis on each row and build the audience category summary view. The IPTC tagging spoke covers that taxonomy if a different system needs it for the same pages.