The Scenario
Someone on the content team built a research list three months ago — 30 industry blog post URLs in an Excel workbook — and then left the company. The work was handed to you with a note that said "we need this analyzed before the editorial calendar meeting."
You open the workbook. Thirty URLs. No titles, no summaries, no word counts. Just links.
You click the first one. The article is 2,400 words and takes eight minutes to read well enough to write a one-sentence summary. You have thirty of these. The meeting is Friday.
The bad version:
- Click each URL, wait for the article to load, read enough to write a summary — then open the workbook, type the summary, switch back, read the title from the browser tab, type that too
- Realize halfway through that some articles are gated — you get a preview paragraph and a subscribe prompt, so your "summary" for row 14 is just the lede
- Open a separate tab to count approximate word count for each article, note that the method varies by blog platform, and give up on consistency after row six
The editorial calendar meeting needs content analysis, not a transcription exercise. You were handed a research task, not a reading assignment. The difference matters when you have a stack of other things queued up.
The Easy Way: One Prompt in SheetXAI
SheetXAI is an AI agent that lives inside your Excel workbook. It reads your data, understands the URL column, and through its built-in ScrapingAnt integration it fetches each article as rendered Markdown — full text, clean structure — and extracts exactly the fields you need.
Open the SheetXAI sidebar and ask:
For each URL in column A, use ScrapingAnt to extract the page content as Markdown, then parse out the article title into column B, write a one-sentence summary into column C, and calculate approximate word count into column D
SheetXAI processes each URL through ScrapingAnt, retrieves the rendered content, extracts the title from the page, generates a one-sentence summary from the full text, and calculates word count from the Markdown. All three columns fill in one pass.
What You Get
- Column B: article title as it appears in the page heading
- Column C: a one-sentence summary drawn from the full article text
- Column D: approximate word count based on the extracted Markdown
- Rows where the page is paywalled or returns no body text flagged with "Content unavailable — manual review"
What If the Data Is Not Quite Ready
Some articles are on platforms that render content behind a scroll event
For each URL in column A, use ScrapingAnt with scroll emulation enabled so the full article body loads before extraction, then populate columns B, C, and D with title, summary, and word count
The workbook has a mix of English and Spanish articles — column E should note the language
Scrape all 30 URLs in column A with ScrapingAnt Markdown extraction, fill title in column B and summary in column C, and in column E write the detected language of the article body
Some URLs are duplicate posts — identify them after extraction
After filling columns B through D for all 30 rows, scan column B for duplicate article titles and write "DUPLICATE" in column F for any rows that share a title with another row
Full content audit in one shot: scrape, summarize, score, and flag
For each URL in column A, use ScrapingAnt to extract the full article as Markdown; write the title into column B, a one-sentence summary into column C, word count into column D, and in column E give a reading level estimate (Basic / Intermediate / Advanced) based on the text; flag any article under 500 words in column F as "Too short for editorial use"
One ask handles the scraping, the analysis, and the classification.
Try It
If you have a research list sitting in an Excel workbook with nothing but URLs — Get the 7-day free trial of SheetXAI and ask it to extract titles, summaries, and word counts from every row using ScrapingAnt. Also worth reading: how to enrich B2B lead lists with social links, or back to the hub for the full overview.