Step-by-step guides for real-world workflows. Follow along in your own Google Sheet.
Bulk sentiment analysis, categorization, and auto-response drafting.
Topic research to published content with SEO optimization.
Standardize messy contacts, addresses, and company names.
Score candidates, extract skills, and generate interview questions.
Advanced: output of one formula feeds the next for multi-step processing.
Turn a wall of unstructured customer feedback into structured, actionable data: sentiment scores, category labels, urgency ratings, and draft responses.
Create a new Google Sheet. Label Column A as Review Text. Paste your customer reviews into cells A2 through A1001 (one review per row). These can be product reviews, support tickets, survey responses, or social media comments.
Tip: Export reviews from your helpdesk (Zendesk, Intercom, Freshdesk) as CSV and paste directly. Or use Google Forms responses.
In cell B1, type the header Sentiment. In cell B2, enter this formula:
=IF(A2="","",AI("Reply with exactly one word: Positive, Negative, or Neutral. Review: " & A2))The IF(A2="",...) wrapper skips empty rows, saving your API quota. Drag B2 down to B1001.
Expected output: Every row in Column B now shows exactly one word. Positive, Negative, or Neutral.
Label C1 as Category. In C2:
=IF(A2="","",AI("Categorize this review into exactly one of: Pricing, Product Quality, Customer Support, Shipping, UX/Interface, Other. Reply with the category only. Review: " & A2))Drag down. Every row gets a category label you can filter, pivot, and chart.
Label D1 as Urgency. In D2:
=IF(A2="","",AI("Rate urgency 1-5 (5=most urgent), number only. Review: " & A2))Now you can sort by urgency and address critical issues first.
Label E1 as Suggested Response. In E2:
=IF(A2="","",AI("Write a brief, empathetic customer support response to this review in 2-3 sentences: " & A2))Review and edit the drafts before sending. The AI gets you 80% of the way there.
With all four columns populated, you now have a structured dataset. Use built-in Google Sheets features to analyze:
=COUNTIF(B:B,"Negative")| A: Review Text | B: Sentiment | C: Category | D: Urgency | E: Suggested Response |
|---|---|---|---|---|
| "Love the product but shipping took 3 weeks..." | Negative | Shipping | 4 | "We're sorry about the shipping delay..." |
| "Best purchase I've made this year!" | Positive | Product Quality | 1 | "Thank you so much for the kind words..." |
| "App keeps crashing on iOS 17" | Negative | UX/Interface | 5 | "We've identified this issue and our team..." |
Go from a list of topics to fully-formed content assets: SEO titles, meta descriptions, blog outlines, social media posts, and hashtag sets.
Column A: Topic/Keyword. Enter 20-50 content topics your audience cares about. Example: "remote work productivity tips", "AI tools for small business", "how to reduce customer churn".
=AI("Write one SEO-optimized blog title under 60 characters about: " & A2)Constrain to one title and 60 characters for search engine display.
=AI("Write a meta description under 155 characters for a blog post titled: " & B2)Notice Column C references Column B, not Column A. This is chaining. The SEO title feeds the meta description for consistency.
=AI("Write a 5-section blog outline with H2 headings for: " & B2)=AI("Write a LinkedIn post under 200 words promoting a blog about: " & B2)=AI("List 10 relevant hashtags, each starting with #, comma-separated, for: " & A2)Result: From 30 topic keywords, you now have 30 SEO titles, 30 meta descriptions, 30 blog outlines, 30 LinkedIn posts, and 30 hashtag sets. A content calendar that would take a team 2 weeks just appeared in 10 minutes.
| A: Topic | B: SEO Title | C: Meta | D: Outline | E: LinkedIn | F: Hashtags |
|---|---|---|---|---|---|
| remote work productivity | 7 Remote Work Habits That Double Output | Discover proven habits that help... | 1. Morning Routine... | Working from home? Here are 7... | #RemoteWork, #Productivity... |
Transform messy, inconsistent contact data into clean, standardized records ready for your CRM import.
Column A: Raw Contact Data. Paste entries from your CRM export, email inbox, business cards, or lead forms. Each cell can contain mixed formats like "john smith, CEO - Acme Corp (john@acme.com) 555.123.4567".
Set up columns for each data point:
' B2 - Full Name
=AI("Extract only the person's full name from: " & A2)
' C2 - Company
=AI("Extract only the company name from: " & A2)
' D2 - Email
=AI("Extract only the email address from: " & A2)
' E2 - Phone (standardized)
=AI("Extract the phone number and format as (XXX) XXX-XXXX: " & A2)
' F2 - Title/Role
=AI("Extract only the job title from: " & A2)=AI("Standardize this company name to its official form (e.g., 'msft' becomes 'Microsoft'): " & C2)=AI("Classify the industry of " & C2 & " into one of: Technology, Healthcare, Finance, Retail, Manufacturing, Services, Other")| A: Raw Data | B: Name | C: Company | D: Email | E: Phone | F: Title |
|---|---|---|---|---|---|
| john smith CEO acme corp john@acme.com 5551234567 | John Smith | Acme Corporation | john@acme.com | (555) 123-4567 | CEO |
| Jane D. - VP Sales @ BigCo jane.d@bigco.io | Jane D. | BigCo | jane.d@bigco.io | N/A | VP Sales |
Score candidates against job requirements, extract key skills, flag concerns, and generate tailored interview questions. All from pasted resume text.
Cell B1: paste your Job Description text. Column A (starting A4): paste each candidate's Resume Text (one resume per row). This two-part setup lets every formula reference the same JD.
=AI("List technical skills mentioned in this resume, comma-separated: " & A4)=AI("Score this resume 1-10 against this job description. Reply with number only. JD: " & $B$1 & " Resume: " & A4)The $B$1 is an absolute reference. It always points to your job description, even as you drag down.
=AI("List any red flags (employment gaps, overqualification, skill mismatches) in one sentence: " & A4)=AI("Based on this resume and job description, write 3 targeted interview questions. JD: " & $B$1 & " Resume: " & A4)Sort by Column C (score) descending. Your top candidates are at the top, complete with extracted skills, flagged concerns, and custom interview questions. Share the sheet with your hiring team for collaborative review.
Build multi-step AI processing pipelines where the output of one column feeds into the next. This is SheetSense's most powerful pattern.
Normally, each =AI() formula references raw data in Column A. In a chain, Column C references Column B, which references Column A. Each column is a processing step.
' The chain:
' A2: Raw data (input)
' B2: Step 1: process A2
' C2: Step 2: process B2
' D2: Step 3: process C2
' Each step refines, transforms, or enriches the previous output' A2: Product features (raw input)
"Wireless Bluetooth 5.3 headphones, 40hr battery, ANC, foldable, USB-C"
' B2: Generate English marketing description
=AI("Write a compelling 2-sentence product description: " & A2)
' C2: Translate to Spanish
=AI("Translate to Spanish, preserve marketing tone: " & B2)
' D2: Translate to French
=AI("Translate to French, preserve marketing tone: " & B2)
' E2: Generate SEO title from the description
=AI("Write an SEO title under 60 chars for: " & B2)Result: One row of product features becomes a marketing description, two translations, and an SEO title. automatically.
' A2: Customer complaint text
' B2: Classify severity
=AI("Rate severity 1-5, number only: " & A2)
' C2: Draft response ONLY if severity >= 4
=IF(VALUE(B2)>=4, AI("Write an urgent support response: " & A2), "No action needed")This pattern uses the AI's classification output to conditionally trigger further AI processing. You only spend API calls on the rows that matter.
' A2: Original text (e.g., a marketing email)
' B2: Rewrite for clarity
=AI("Rewrite this email to be clearer and more concise: " & A2)
' C2: Score the rewrite 1-10
=AI("Score this email 1-10 for clarity and professionalism, number only: " & B2)
' D2: If score < 7, rewrite again with feedback
=IF(VALUE(C2)<7, AI("This scored " & C2 & "/10 for clarity. Rewrite to be even clearer: " & B2), B2)This is an AI reviewing its own output and self-correcting. Column D contains either the improved version (if it scored low) or the original rewrite (if it scored high enough).
=IF(B2="","",AI("..." & B2))$B$1 for values that should not change when dragging (like a job description or system prompt).