Task Placement and Execution Guide for Sheet Workflows

Overview

This guide explains how tasks work in Google Sheets workflows, focusing on where to place them, how they execute, and the critical role of the header row.

Table of Contents

1. Sheet Structure and Task Placement
2. The Header Row - Your Control Center
3. How Tasks Execute
4. Task Spreading and Copying
5. Workflow Creation Patterns
6. Task Placement Guidelines

Sheet Structure and Task Placement

The Grid Layout

• Row 1 (Header Row): The template row containing task formulas or column headers
• Rows 2 and below: Data rows where tasks actually execute
• Columns: Each column represents a step in your workflow or contains data

Essential Placement Rules

1. Task formulas belong in Row 1 - this is where you define what the column will do
2. Data starts from Row 2 - this is where the actual work happens
3. Each column is independent - each can have its own task type
4. Tasks can reference other columns - using references like {A} or {B2}

The Header Row - Your Control Center

Understanding Row 1

Row 1 is special - it's both a template holder and a control panel:

1. Template Storage: Row 1 holds the "master formula" that defines what operation will happen in that column
2. Batch Trigger: Clicking "Run" on a Row 1 cell processes ALL data rows in that column at once

How the Header Row Works

When you run a task from Row 1:

1. The system recognizes you're in the header row
2. It looks down the column to find all rows with data
3. It copies the task formula to each data row
4. It executes each row in sequence

Example:

Cell C1 contains: =TASK_LM("Process", {A}, "Summarize this")
Cells C2-C10: Currently empty
Cells A2-A10: Contain text to process

Action: Click Run on C1
Result: The task formula spreads to C2-C10 and each executes

Model Dropdown Feature

Some columns in Row 1 might show a dropdown menu of AI models instead of a formula. When this happens:

• The system finds a task that uses the "{model}" placeholder
• It substitutes your selected model into the formula
• The formula then spreads to all data rows
• The dropdown remains in Row 1 for easy model switching

How Tasks Execute

Single Cell Execution

When you run a task in a single data cell (Row 2 or below):

1. If the cell is empty, the task formula is copied from Row 1
2. The task parameters are read from the cell
3. The task executes with those parameters
4. The result appears in the designated output location

Batch Execution (Running from Row 1)

When you run a task from the header row:

1. The task formula spreads to all data rows in that column
2. Cell references automatically adjust for each row
3. Each row executes in sequence from top to bottom
4. Results appear as each row completes

Column-by-Column Processing

When processing multiple columns:

• The system completes all rows in Column A before starting Column B
• This ensures dependent calculations work correctly
• Processing speed can be controlled with rate limits

Task Spreading and Copying

How Tasks Spread from Row 1

When a task spreads from the header row to data rows:

1. The formula in Row 1 is treated as a template
2. It copies down to each row that contains data
3. References automatically adjust for each row
4. Row 1 remains unchanged as the master template

Reference Adjustments

When formulas copy down, references change intelligently:

• {A} in Row 1 becomes the value from A2 in Row 2, A3 in Row 3, etc.
• {B2} stays as B2 in all rows (absolute reference)
• $A$2 remains exactly as written (spreadsheet absolute)
• {Header Name} finds the column with that header and uses the current row

Tasks That Support Batch Processing

These tasks can be placed in Row 1 and automatically spread to multiple rows:

• TASK_LM
• TASK_URL_READER
• TASK_WEBHOOK
• TASK_TRANSCRIBE
• TASK_DOC_TO_MARKDOWN
• TASK_FORMULA
• TASK_MARKDOWN_TO_DOC
• TASK_TEMPLATE
• TASK_API
• TASK_ROW_UPDATE
• TASK_SPLIT
• TASK_SHEET_UPDATE
• TASK_COPY
• TASK_CLEAR
• TASK_CONCAT
• TASK_FORMAT
• TASK_SEARCH

Reference Adjustment

When copying formulas, the system:

1. Preserves relative references: A2 in Row 1 becomes A3 in Row 2
2. Maintains absolute references: $A$2 stays as $A$2
3. Adjusts column-only references: {B} becomes B2 in Row 2, B3 in Row 3
4. Handles placeholders: {Header Name} references are resolved

Workflow Creation Patterns

Pattern 1: Simple Linear Workflow

Column A: Input Data
Column B: =TASK_LM("Process", {A}, "Transform this data")
Column C: =TASK_FORMAT("Format", {B}, "bold,italic")

Pattern 2: Multi-Step Processing

Column A: URLs
Column B: =TASK_URL_READER("Fetch", {A})
Column C: =TASK_LM("Extract", {B}, "Extract key points")
Column D: =TASK_TEMPLATE("Output", "{C}", "Report: {content}")

Pattern 3: Parallel Processing

Column A: Source Text
Column B: =TASK_LM("Summary", {A}, "Summarize in 50 words")
Column C: =TASK_LM("Keywords", {A}, "Extract 5 keywords")
Column D: =TASK_LM("Sentiment", {A}, "Determine sentiment")

Pattern 4: Batch Processing with Header Trigger

Row 1, Column C: =TASK_API("API Call", "{A}", "endpoint")
Rows 2-100, Column A: Input parameters
Rows 2-100, Column C: (empty - will receive API responses)

Action: Click Run on C1
Result: Processes all 99 rows automatically

Task Placement Guidelines

When Creating a New Workflow

1. Choose your starting point:

   • Find the first empty column, or
   • Start where the user specifies

2. Set up Row 1 (Header Row):

   • Place each task formula in Row 1 of its column
   • Add descriptive labels as the first parameter
   • Ensure formulas reference the correct columns

3. Prepare data rows (Row 2+):

   • Input data goes in the leftmost columns
   • Processing columns can start empty
   • Output columns will be filled when tasks run

4. Configure references correctly:

   • Use {A} to reference "same row, column A"
   • Use {A2} for a specific cell that never changes
   • Use {Header Text} to reference a column by its header

Column Organization Strategy

Recommended Layout:

1. Input Columns (leftmost): Raw data, URLs, text, etc.
2. Processing Columns (middle): Tasks that transform data
3. Output Columns (rightmost): Final results, formatted data

Spacing Considerations:

• Keep related columns together
• Leave empty columns between different workflows
• Group similar operations in adjacent columns

Working with Multiple Sheets

When a workflow needs multiple sheets:

• Use TASK_COPY to transfer data between sheets
• Keep consistent column structures across sheets
• Name sheets descriptively for easy reference

Best Practices for Creating Workflows

1. Before Placing Tasks

• Check if columns are empty or contain data
• Verify the sheet exists and is accessible
• Ensure Row 1 is available for task formulas

2. Protect Existing Work

• Never overwrite existing formulas without permission
• Ask before replacing data in columns
• Consider creating a new sheet for complex workflows

3. Set Up References Properly

• Use {A} for row-relative references
• Use {A2} when you always need that specific cell
• Verify all referenced columns exist before placing tasks

4. Prevent Common Issues

• Don't create circular references (A depends on B, B depends on A)
• Ensure data exists before placing processing tasks
• Test with a few rows before running on hundreds

5. Make It User-Friendly

• Add clear labels to each task (first parameter)
• Use descriptive column headers
• Document what the workflow does

How to Trigger Execution

Running Tasks Manually

For Single Cells:

1. Click on the cell with the task
2. Click the "Run" button in the sidebar

For Entire Columns:

1. Click on the Row 1 cell of that column
2. Click "Run" to process all data rows

For Multiple Columns:

1. Select the range covering multiple columns
2. Click "Run" to process them sequentially

Execution Behavior

Processing Order:

• Rows process from top to bottom
• Columns process from left to right
• Each cell completes before the next begins

Visual Feedback:

• Processing cells may show loading indicators
• Errors appear with red borders
• Completed cells show their results

Speed Control:

• Some tasks support rate limiting (requests per minute)
• This prevents overwhelming external services
• The system respects these limits automatically

Key Concepts

When creating workflows, remember:

The Power of Row 1

• Row 1 is the command center - all task definitions go here
• It's also the batch trigger - clicking Run on Row 1 processes the entire column
• It remains the template - Row 1 never changes when tasks execute

Reference Intelligence

• {A} is smart - it knows to use A2 when in Row 2, A3 when in Row 3
• Headers are flexible - {Customer Name} finds that column automatically
• Mixing is allowed - combine {A}, B2, and {Header} in one formula

Workflow Building Blocks

1. Input columns: Where users place their data
2. Task columns: Where processing happens (Row 1 has the formula)
3. Output columns: Where results appear

The Execution Flow

1. User places data in input columns (Row 2+)
2. User clicks Run on a Row 1 task cell
3. Task formula copies to all data rows
4. Each row executes with its own data
5. Results appear in designated locations

Summary

The Sheet Workflows system is built around a simple but powerful concept: Row 1 defines what happens, Rows 2+ are where it happens.

When creating a workflow:

• Place task formulas in Row 1
• Use smart references that adjust per row
• Let users trigger batch processing from Row 1
• Keep columns independent but connectable
• Build workflows incrementally, column by column

This design makes workflows visual, understandable, and easy to modify - users can see exactly what will happen by looking at Row 1, and execute everything with a single click.