Your personal AI field guide

Use the right tool.
For the right job.

This playbook maps your department's work to specific AI tools — with ready-to-run workflows, copy-paste prompts, and cross-tool hacks you can use starting today.

Department

Digital Transformation & E-Commerce

Data-heavy work. Structured analysis, performance tracking, cart behaviour, product content at scale.

Your AI Power Stack

Google AI Studio
Claude
NotebookLM
Gemini
ChatGPT
Google Gems
Gemini Slides
Cross-Tool Workflows
Data → Insight → Deck

The Performance Review Pipeline

Turn raw category data into a board-ready deck without leaving your browser.

Upload CSV → Claude
Get structured analysis
Paste into Gemini Slides
Claude finds the ASP compression signals and return rate anomalies. Gemini Slides turns the output into a 3-slide structure. Your job: edit the title and add your name.
Large Context Hack

Multi-Month Analysis with AI Studio

When your dataset is too large for standard tools, Google AI Studio handles it. Then bring the output into NotebookLM to query across time periods.

Google AI Studio
Copy insights output
Add as source in NotebookLM
Query across quarters
AI Studio processes your entire 6-month dataset at once. The output becomes a "smart document" in NotebookLM — query it in plain language: "Which category showed consistent ROAS improvement?"
Recurring Task Gem

Monthly Performance Summary Gem

Build this once. Use it every month. Paste your numbers in, get a board-ready paragraph out.

Build Gem in Gemini
Paste monthly data
Consistent structured summary
Your Gem holds the context (your categories, your KPIs, your format preference). You only paste the new numbers. The output is the same clean format every time.
Content at Scale

Product Description Batch Generator

200 product descriptions. One prompt. Claude writes them in your catalogue format.

Claude
Define format + upload SKU list
Bulk descriptions ready
Paste your SKU name, category, and key attributes. Claude writes in your tone. Review 10%, publish the rest. Time saved: 4–6 hours per catalogue refresh.
Department

Business Function & Marketing

Campaign briefs, partner pitches, business reviews, co-marketing proposals — output-heavy, relationship-driven work.

Your AI Power Stack

Claude
NotebookLM
Ask Gemini (Gmail/Docs)
Google Gems
ChatGPT
Gemini Slides
Cross-Tool Workflows
Intelligence Pipeline

Campaign Intelligence → Next Brief

Stop guessing what works. Upload past reports to NotebookLM, extract the pattern, then brief the next campaign on evidence.

Upload past campaign reports → NotebookLM
Ask: "What campaign type consistently outperforms?"
Use insight to prompt Claude for next brief
NotebookLM reads across all your quarters simultaneously. You get the pattern in 30 seconds. Feed that pattern into Claude: "Based on this finding, draft a campaign brief for Diwali 2026."
Partner Pitch Chain

First Draft → Gmail Polish → Send

Claude writes the structure. Gemini in Gmail tightens the tone for your specific partner. One final read, then send.

Claude — full pitch draft
Paste into Gmail draft → Ask Gemini: "Sharpen tone. Add a confident next step."
Review once. Send.
The division of labour: Claude is better at structure and content. Gemini in Gmail is better at email tone and brevity. Use both.
Brief Machine

Campaign Brief Gem — One Input, Full Brief Out

Build this Gem once. Never start a brief from scratch again.

Gem: "Objective, audience, budget"
Full brief: name, channel mix, timeline, KPIs
Paste into Gemini Slides → 3-slide deck
The Gem holds your VMart context permanently. You only provide the 3 inputs. The full brief is out in 45 seconds. Slide deck from that brief: 2 more minutes.
Review Shortcut

Email Thread → Summary → Action Plan

Long partner threads before a campaign deadline? Ask Gemini to cut through it.

Open Gmail thread
Ask Gemini: "Summarise and suggest a follow-up"
Reply in context, not from memory
Especially useful before weekly partner calls. Read the AI summary, not the 40-email thread. Walk into the call knowing exactly where things stand.
Department

Store Design & Visual Merchandising

Visual concept creation, layout planning, seasonal themes, trend synthesis, and creative direction at scale.

Your AI Power Stack

Google Flow
Gemini
NotebookLM
ChatGPT
Claude
Cross-Tool Workflows
Concept to Video

Seasonal Theme → Visual Concept Video

Describe your seasonal direction in Gemini. Refine it. Feed the refined concept into Google Flow and generate a visual reference video for your team.

Describe theme in Gemini
Refine aesthetic direction
Feed prompt into Google Flow
Concept video for vendor brief
Use Google Flow to generate a 5–10 second moodboard video. It becomes the reference point your vendors, fixturing teams, and in-store teams all work from — no lengthy written briefs needed.
Trend Intelligence

Trend Reports → Creative Brief

Upload trend decks, competitor references, and lookbooks into NotebookLM. Ask it to extract the visual themes emerging across all sources.

Upload trend docs → NotebookLM
Ask: "What visual themes are consistent across Q3 sources?"
Use output as creative brief input
NotebookLM cross-references everything at once. You get a synthesis no one on the team had time to write manually. That synthesis becomes your seasonal creative direction.
Copy at Scale

Store Theme Copy Generator

Need 12 in-store headline banners for the Diwali fixture? Claude writes all 12 from one brief in under 2 minutes.

Describe theme + tone → Claude
Get 12 headline variants
Pick 3, brief the design team
Prompt: "Write 12 in-store display headline options for our Diwali 2026 theme. Tone: warm, celebratory, value-forward. Each headline under 6 words." Pick the best 3. Done in 3 minutes total.
Department

Store Operations

SOPs, briefings, staff coordination, process documentation, manager communications — speed and accuracy matter here.

Your AI Power Stack

Ask Gemini (Docs)
Ask Gemini (Gmail)
NotebookLM
Claude
Google Gems
Cross-Tool Workflows
SOP Builder

Bullet Points → Full SOP in Google Docs

Open a Google Doc. Type your bullet points — what needs to happen, in what order. Ask Gemini to turn it into a formatted Standard Operating Procedure.

Bullet points in Google Docs
Ask Gemini: "Turn this into a formatted SOP with sections and numbered steps"
Review and publish
Your domain knowledge is the input. Gemini handles structure, formatting, and language. Time from raw notes to publishable SOP: under 10 minutes.
Manager Briefings

Key Points → Crisp Briefing Email

Before a campaign launch or seasonal change, brief your store managers in 100 words — not 500. Ask Gemini in Gmail to compress your notes into a sharp briefing.

Draft key points in Gmail
Ask Gemini: "Format as a crisp manager briefing. 100 words. What, when, what action needed."
Send with confidence
Shorter briefings get read. Gemini in Gmail compresses without losing the instruction. Your store managers get clarity, not noise.
Training Resource

Operations Manuals → Instant FAQ

Upload your existing SOPs, guidelines, and training manuals to NotebookLM. It becomes a searchable knowledge base your team can query in plain language.

Upload SOPs + manuals → NotebookLM
Team asks: "What's the procedure for X?"
Instant, sourced answer
NotebookLM cites which document the answer came from. Especially useful for onboarding new staff — instead of reading 80 pages, they ask questions and get precise answers in seconds.
Department

Digital Transformation & E-Commerce

Data-heavy work. Structured analysis, performance tracking, cart behaviour, product content at scale. You're the most technical group — and the internal AI champions post-training.

Your AI Power Stack

Google AI Studio
Claude
NotebookLM
Gemini
ChatGPT
Google Gems
Gemini Slides
Your advantage: You already understand data. Your AI outputs will be dramatically better than other departments because you know what to look for — and you can embed that domain knowledge directly into your prompts. The better your hypothesis, the better the AI output.
Cross-Tool Workflows
Data → Insight → Deck

The Performance Review Pipeline

Turn raw category data into a board-ready deck without leaving your browser.

Upload CSV → Claude
Structured analysis with hypotheses
Paste into Gemini Slides → 3-slide deck
Claude finds the ASP compression signals and return rate anomalies. Gemini Slides turns the output into a 3-slide structure. Your job: edit the title and add your name.
Large Context Hack

Multi-Month Analysis with AI Studio

When your dataset is too large for standard tools, Google AI Studio handles up to 1M tokens. Then bring the output into NotebookLM.

Google AI Studio — full dataset
Copy insights output
Add as source in NotebookLM
Query across quarters
AI Studio processes your entire 6-month dataset simultaneously. The output becomes a "smart document" in NotebookLM — ask it in plain language: "Which category showed consistent ROAS improvement?"
Recurring Task Gem

Monthly Performance Summary Gem

Build this once. Use every month. Your numbers in — board-ready paragraph out.

Build "Monthly Analyst" Gem
Paste new month's data
Consistent format every time
The Gem holds your categories, KPI definitions, and format preference permanently. You only paste the new numbers. Same clean output, zero setup, every single month.
Content at Scale

Product Description Batch Generator

200 product descriptions. One prompt. Claude writes them in your catalogue format.

Claude — define format + SKU list
Batch generation with attributes
Review 10%, publish the rest
Time saved: 4–6 hours per catalogue refresh. The prompt does the heavy lifting — the more specific your attribute list, the less editing you do.
Ready-to-Use Prompts
E-Commerce Data Analysis
Use in: Claude · ChatGPT · Google AI Studio
You are a senior e-commerce analyst at VMart, India's leading value fashion retailer. Here is monthly GMV, orders, and return rate data for 8 categories over 6 months. Identify the following: 1. 2 categories with declining GMV but stable orders — flag this as an ASP compression signal and explain what it means 2. Any category with a rising return rate over 3 consecutive months — flag the risk and suggest one hypothesis 3. One actionable recommendation per finding Format: Board-ready report with clear headings, numbered findings, and bullet-point recommendations. No filler sentences.
Pro tip: The better your hypothesis going in, the better the output. Say "I suspect X" in your prompt — AI will validate or challenge it, which is more useful than a generic insight.
Cart Abandonment Insight Report
Use in: Claude · ChatGPT
You are a conversion rate analyst. I'm sharing cart abandonment rate data by category and month for an Indian value fashion e-commerce platform. Analyse this data and provide: 1. Which month and category combination has the highest abandonment — and your hypothesis for why 2. Whether abandonment correlates with higher average order value or seasonal timing 3. One A/B test hypothesis I could run to address the top finding Format: 3 clearly labelled sections. Keep it under 250 words. Write for a digital head who reads fast.
Pro tip: After getting the output, follow up with: "Now give me the same analysis as a 5-bullet executive summary I can paste into a WhatsApp group."
Google Gems — Monthly Performance Analyst
Use in: Google Gems (save once, use always)
You are a senior e-commerce performance analyst at VMart, a value fashion retailer in India. Our key categories are: Women Ethnic, Women Western, Men Casual, Footwear, Kids, Accessories, Home Decor, and Beauty. When I give you monthly performance data (GMV, orders, return rate, ROAS, cart abandonment), generate a structured monthly performance summary with: - Top 2 performing categories and why - Bottom 2 categories with a risk flag - One anomaly worth investigating - One recommendation for next month Format: Structured report with clear section headings. Tone: direct and analytical, written for leadership review. No filler. Under 300 words.
How to use: Go to gemini.google.com → Gems → New Gem → paste this as the instruction → Save. Every month, open the Gem and paste your numbers. Same clean output, zero setup.
Department

Business Function & Marketing

Campaign briefs, brand partnership pitches, co-marketing proposals, business reviews. Output-heavy, relationship-driven work. AI removes the production overhead so you can focus on the strategy.

Your AI Power Stack

Claude
NotebookLM
Ask Gemini (Gmail)
Ask Gemini (Docs)
Google Gems
ChatGPT
Gemini Slides
The mental model: AI doesn't replace your strategy. It removes the production overhead — the 2-hour brief that becomes 20 minutes, the draft proposal that takes 3 rounds of email into 1. Your relationship knowledge and strategic layer stay yours.
Cross-Tool Workflows
Intelligence Pipeline

Campaign Intelligence → Next Brief

Stop guessing what works. Upload past reports to NotebookLM, extract the pattern, brief the next campaign on evidence.

Upload past reports → NotebookLM
Ask: "What campaign type consistently outperforms?"
Feed insight into Claude → next brief
NotebookLM reads across all your quarters simultaneously. You get the pattern in 30 seconds. Feed that finding into Claude: "Based on this, draft a campaign brief for Diwali 2026."
Partner Pitch Chain

First Draft → Gmail Polish → Send

Claude writes the structure. Gemini in Gmail tightens the tone. One final read, then send.

Claude — full pitch draft
Paste into Gmail → Ask Gemini: "Sharpen tone. Add a confident next step."
Review once. Send.
Claude handles structure. Gemini in Gmail handles email tone and brevity. The combination beats either tool alone — use both.
Brief Machine

Campaign Brief Gem — One Input, Full Brief Out

Build this Gem once. Never start a brief from scratch again.

Gem input: objective, audience, budget
Full brief: name, channels, timeline, KPIs
Paste brief → Gemini Slides → 3-slide deck
Brief out in 45 seconds. Slide deck from that brief: 2 more minutes. Total time from nothing to leadership-ready brief and deck: under 5 minutes.
Email Intelligence

Thread Summary → Sharp Reply

Long partner threads before a deadline? Ask Gemini to cut through it before your call.

Open Gmail thread
Ask Gemini: "Summarise and suggest a follow-up"
Walk into the call knowing where things stand
Read the AI summary, not the 40-email thread. Especially useful before weekly partner calls — you arrive briefed, not overwhelmed.
Ready-to-Use Prompts
Brand Partnership Pitch
Use in: Claude · ChatGPT
You are a senior brand partnerships manager at VMart, India's leading value fashion retailer. Write a partnership pitch for a mid-market Indian ethnic wear brand. Target reader: Their marketing head Goal: Co-branded festive season campaign Include: - Why VMart (reach, footfall, digital presence) - What we offer as a retail partner - What we're asking for (co-marketing budget, exclusivity period) Format: 3 short paragraphs. Tone: confident and collaborative, not salesy. End with one clear next step for the reader to take. Keep the entire pitch under 200 words.
Customise: Replace "mid-market Indian ethnic wear brand" with the actual brand name and category. The more specific, the less editing you'll need to do on the output.
Executive Business Review Summary
Use in: Claude · ChatGPT
You are a senior marketing manager at VMart preparing a business review for leadership. I'm sharing campaign performance data from the last quarter. Write a 150-word executive summary that includes: 1. What worked — top 1–2 performing campaigns and why 2. What didn't — bottom 1 campaign with an honest reason 3. One clear recommendation for next quarter Tone: Direct and data-backed. Written for leadership who reads fast and expects opinions, not just numbers. No passive voice. No filler phrases like "it is recommended that..." — just say what needs to happen.
Pro tip: Paste the AI output into a Google Doc and use Ask Gemini to "rewrite this in a sharper, more confident tone" — then compare the two versions before picking the one to send.
Google Gems — Campaign Brief Assistant
Use in: Google Gems (save once, use always)
You are a senior marketing manager at VMart, a value fashion retailer in India. When I give you a campaign objective, target audience, and budget, generate a complete campaign brief. The brief must include: - Campaign name - Objective (one sentence) - Target audience (specific demographics) - Key message (one line) - Channel mix (digital + in-store breakdown) - 4-week timeline with weekly milestones - Success metrics (2–3 KPIs) Tone: Strategic and concise. Use clear headings. No filler. The brief should be ready to share with a brand partner or internal team without editing.
How to test it: After saving the Gem, try this input: "Objective: drive footfall for Diwali. Target: women 25–45 in Tier 2 cities. Budget: ₹50L." A complete brief should appear in under 30 seconds.
Department

Store Design & Visual Merchandising

Visual concept creation, seasonal theme development, layout planning, creative direction for vendors and fixturing teams — at scale, faster than before.

Your AI Power Stack

Google Flow
Gemini
NotebookLM
Claude
ChatGPT
Google Flow: Google's AI video generation tool. Describe a visual concept in text — it renders a short video reference. Use it to brief vendors, align fixturing teams, or present seasonal themes to leadership without needing a design agency.
Cross-Tool Workflows
Concept to Video

Seasonal Theme → Visual Concept Video

Describe your seasonal direction in Gemini. Refine the aesthetic. Feed into Google Flow. Get a visual reference video for briefing vendors and teams.

Describe seasonal theme → Gemini
Refine aesthetic direction with Gemini
Feed refined concept → Google Flow
Concept video for vendor brief
A 5–10 second visual concept video replaces 3 pages of written brief. Your vendors, fixturing teams, and in-store staff all work from the same visual reference point — no ambiguity.
Trend Intelligence

Trend Reports → Creative Brief

Upload trend decks, competitor references, and lookbooks into NotebookLM. Ask it to synthesise the visual patterns across all sources.

Upload trend docs → NotebookLM
Ask: "What visual themes are consistent across Q3 sources?"
Use synthesis as your creative brief
NotebookLM cross-references all your uploaded documents at once. The synthesis it gives you — themes, colours, motifs — becomes your seasonal creative direction. No one on the team had time to do this manually.
Copy at Scale

Store Theme Copy — 12 Variants in 2 Minutes

Need 12 in-store banner headlines for a seasonal fixture? Claude generates all 12 from one brief. You pick 3.

Brief Claude: theme + tone + constraints
12 headline variants in 2 minutes
Pick 3, brief the design team
Total time from blank screen to 3 shortlisted headlines for your designer: under 5 minutes. No agency needed for the copy phase.
Ready-to-Use Prompts
Google Flow — Seasonal Concept Prompt
Use in: Google Flow · Gemini
A retail store interior for a Diwali 2026 festive season. The aesthetic is warm celebration — deep marigold yellows, terracotta oranges, rich gold accents on white fixtures. The space is bright and airy, not cluttered. Clothing displays at the centre feature women's ethnic wear — kurtas, dupattas in deep jewel tones. Subtle diyas and string lights frame the ceiling. The atmosphere is premium but accessible — celebratory without being overwhelming. Camera moves slowly through the entrance, past the display tables, and settles on a hero fixture wall.
How to use: Paste this into Google Flow to generate a concept video. Adjust colour palette and season for other occasions (End of Season Sale, Republic Day, etc.).
In-Store Headline Variants Generator
Use in: Claude · ChatGPT
Write 12 in-store display headline options for VMart's Diwali 2026 festive season fixture. Tone: Warm, celebratory, value-forward. Not premium or luxury — accessible and joyful. Audience: Women 25–45, shopping for family and self during the festive season. Format: Each headline must be under 6 words. No punctuation. No exclamation marks. Avoid: Clichés like "Big Sale", "Lowest Prices", "Don't Miss Out" Organise the 12 options in 3 groups: - 4 headlines for the entrance fixture (aspirational) - 4 headlines for the main floor display (product-focused) - 4 headlines for the checkout zone (value reinforcement)
Pro tip: After you pick your 3 favourites, ask Claude: "Rewrite these 3 in Hindi, keeping the same feeling and word count." Use the Hindi versions for Tier 2 store fixtures.
Department

Store Operations

SOPs, manager briefings, staff coordination, process documentation — speed and accuracy are everything. AI handles the drafting so you can focus on execution.

Your AI Power Stack

Ask Gemini (Docs)
Ask Gemini (Gmail)
NotebookLM
Claude
Google Gems
The key insight for operations: Your domain knowledge is the most valuable input. AI handles structure and language — you provide the "what needs to happen." The combination produces SOPs and briefings faster and more consistently than any manual process.
Cross-Tool Workflows
SOP Builder

Bullet Points → Full SOP in 10 Minutes

Open Google Docs. Type what needs to happen. Ask Gemini to turn it into a formatted Standard Operating Procedure.

Your bullet points in Google Docs
Ask Gemini: "Format as SOP with sections and numbered steps"
Review. Publish.
Your operations knowledge is the input. Gemini handles structure, formatting, and professional language. Raw notes to publishable SOP: under 10 minutes.
Manager Briefings

Key Points → 100-Word Briefing Email

Before a campaign launch or seasonal change, brief your managers in 100 words — not 500. Shorter briefings get read.

Draft key points in Gmail
Ask Gemini: "Format as a crisp briefing. 100 words. What, when, what action needed."
Send with confidence
Gemini in Gmail compresses without losing the instruction. Your store managers get clarity, not noise — and they actually read it.
Knowledge Base

Manuals → Instant FAQ for New Staff

Upload your SOPs, training manuals, and guidelines to NotebookLM. It becomes a searchable knowledge base your team queries in plain language.

Upload SOPs + manuals → NotebookLM
Staff asks: "What's the procedure for X?"
Instant, sourced answer
NotebookLM cites which document the answer came from. Especially useful for onboarding — instead of reading 80 pages, new staff ask questions and get precise answers in seconds.
Ops Gem

Pre-Shift Briefing Gem

Build a Gem that generates consistent pre-shift briefings from your daily inputs. Every shift starts with the same quality of information.

Input: target, priority, escalations
Gem outputs formatted briefing
Share with team via WhatsApp or email
The Gem holds your store's context and briefing format. You give it today's variables — footfall target, focus area, any escalations — and it produces a briefing your entire team can follow.
Ready-to-Use Prompts
SOP Generator
Use in: Ask Gemini in Docs · Claude
You are an operations documentation specialist. I'm going to give you bullet points describing a store procedure. Turn these into a clean, formatted Standard Operating Procedure (SOP). The SOP should include: - Title and purpose (1 sentence) - Who this applies to (role) - When to use this SOP (trigger condition) - Step-by-step numbered instructions - What to do if something goes wrong (exception handling, 2–3 scenarios) - Sign-off or completion confirmation Tone: Clear, direct, no ambiguity. Written for frontline staff — no jargon. Each step must be one action only. Here are my bullet points: [PASTE YOUR BULLET POINTS HERE]
Pro tip: After generating the SOP, ask: "Add a checklist version of the same steps at the bottom that a supervisor can use to verify completion."
Google Gems — Pre-Shift Briefing Generator
Use in: Google Gems (save once, use daily)
You are a store operations assistant for VMart. Every day, I'll give you the day's key inputs and you'll generate a clean pre-shift briefing for my team. When I give you: - Today's footfall and sales target - Priority category or zone - Any promotions running today - Any escalations or issues to flag Generate a pre-shift briefing with: 1. Today's goal (one line) 2. Priority focus area and why 3. Active promotions and how to communicate them 4. Any open escalations or watch points 5. One motivating closing line (not generic — make it specific to today's context) Format: Bullet points under each section. Keep the entire briefing under 150 words. Tone: Direct, encouraging, ops-ready.
How to test it: Try: "Target: ₹4.2L. Priority: Women Ethnic — new arrivals in zone 3. Promo: Buy 2 get 10% off. Escalation: Fitting room queue was long yesterday — monitor." You'll get a complete briefing in seconds.

How to Write a Prompt That Actually Works

The quality of your AI output is directly proportional to the quality of your input. Here's the anatomy of a prompt that gets results the first time.

01
Role
Tell AI who to be. "You are a senior e-commerce analyst at VMart" gives it a frame that shapes every sentence it writes.
02
Context
Give it your world. The more specific the context (your industry, your data, your constraints), the less editing you'll do on the output.
03
Task
State exactly what you need. Not "give me insights" — "identify 2 categories with declining GMV and explain the risk." Specific tasks get specific outputs.
04
Format
Tell it how to structure the output. "Board-ready report with headings and bullets" vs "3 paragraphs under 200 words" — different formats for different readers.
Rules That Separate Good Prompts from Bad Ones
Embed your domain knowledge. The AI doesn't know your categories, your KPIs, or your partners. You do. Put that in the prompt.
State your hypothesis. "I suspect footwear return rates are rising because of sizing" gives AI something to validate or challenge — more useful than an open question.
Name your reader. "Write for a brand partner's marketing head" produces different output than "write for leadership." Same content, different register.
Set a word limit. Unconstrained AI writes long. "Under 150 words" forces it to prioritise. Use this for emails, briefings, and summaries.
Say what to avoid. "No filler phrases. No passive voice. No bullet points." Negative constraints are as powerful as positive instructions.
Compare tools on the same task. Run the same prompt in Claude and Gemini. Pick the output that needs less editing. That's your tool for that task.
Use follow-up prompts. "Now make it 50% shorter." "Rewrite this in a more confident tone." "Turn this into a 5-bullet WhatsApp message." One prompt is rarely the end.
Build Gems for recurring tasks. Any task you do more than once a week — campaign briefs, summaries, briefings — belongs in a Gem. Set it up once, stop starting from scratch.
Bad Prompt vs Good Prompt
❌ Bad Prompt
"Give me insights from this e-commerce data."
No role. No context. No specific task. No format. You'll get a generic summary that tells you what you already know.
✓ Good Prompt
"You are a senior e-commerce analyst at VMart. Here is 6-month GMV data across 8 categories. Identify 2 categories with declining GMV but stable orders — flag this as ASP compression. One hypothesis per finding. Board-ready format with headings."
Role ✓ Context ✓ Specific task ✓ Format ✓ Domain knowledge embedded ✓