Guide 2.5 — Skills Lab
Workflow Lab
One task per workflow. Exact steps. Specific tools. What you produce at the end. Built for Indian B2B marketing teams that need results, not theory.
How to use the Workflow Lab
These AI marketing workflows for B2B teams in India are built to be run end to end — each workflow covers one task, with exact steps, tools, and a defined output at the end. No theory, no filler. According to HubSpot’s State of Marketing report, marketers who follow structured AI workflows produce more consistent, higher-quality outputs than those improvising prompt by prompt. This lab builds on the function guides in the AI Skills Lab — work through those first if you have not already.
Every workflow in this lab covers one specific marketing task. Not a framework. Not a guide to thinking about the task. The exact sequence of steps, the tools involved, the prompts to use, and the output you should have at the end. Read time is under five minutes per workflow. You should be able to start within ten.
These workflows are built for lean Indian B2B teams. Tool recommendations include INR pricing where relevant. Market adjustments for NA, Europe, and APAC buyers are noted inline.
8
workflows at launch
<5 min
read time each
10 min
to first result
The workflows
Workflow 01
12 minutes
Turn a founder voice note into a LinkedIn post
Otter.ai / Whisper · Claude / ChatGPT
1
Record a 5-10 minute voice note explaining your POV on any topic relevant to your buyers
2
Transcribe using Otter.ai (free tier) or paste into Claude directly if under 2,000 words
3
Use the thought leadership prompt from the AI for B2B Content guide, with the buyer persona and market filled in
4
Review the draft. Rewrite the opening line in your own voice. The AI opening is almost always too generic
5
Add one specific data point or customer example that the AI could not know
Output
A 600-800 word LinkedIn article draft ready for final edit, or a 200-word post with a strong opening hook
Workflow 02
20 minutes
Build a 5-touch outbound sequence in 20 minutes
Clay (free tier) · Claude / ChatGPT · Apollo / LinkedIn Sales Nav
1
Define your ICP for this sequence: job title, company type, geography, and one specific trigger (funding, hiring, intent signal)
2
Pull 20-30 accounts from Apollo or LinkedIn Sales Nav that match the ICP and trigger
3
Use the outbound sequence prompt from the AI for Demand Gen guide, filling in market context explicitly
4
Generate the sequence. Review step 1 (the opening line) for each variant. This is where personalisation matters most
5
Load into your sequencing tool (Apollo, Instantly, or HubSpot sequences)
Output
A 5-email sequence with subject lines, body copy, and CTAs, adjusted for your target market
Workflow 03
25 minutes
Competitive battlecard from G2 reviews
G2.com · Claude / ChatGPT
1
Go to G2 and find your top competitor. Filter reviews by "most recent" and export or copy 10-15 reviews
2
Identify the 3 most common complaints in the negative reviews. These are their weaknesses
3
Identify what reviewers praise most. This is where they genuinely win
4
Use the battlecard prompt from the AI for Product Marketing guide with the review data as input
5
Have one salesperson review the draft before distributing. They will catch any claim that does not match their experience
Output
A one-page battlecard with win/loss situations, top objections, and specific responses
Workflow 04
30 minutes
Monthly pipeline narrative in 30 minutes
HubSpot / Salesforce · Claude / ChatGPT · Google Sheets
1
Export your pipeline report for the month from CRM. Pull: MQLs, SQLs, pipeline value by stage, top sources, win rate, closed lost reasons
2
Paste the data into a structured format (the pipeline narrative prompt in AI for Revenue Ops has the exact structure)
3
Run the prompt. Review the output for accuracy. The AI will sometimes misread a directional trend from ambiguous data
4
Add the one thing the data does not show: what is the sales team saying about deal quality this month?
5
Send as a Slack message or email, not a dashboard link. Leadership reads prose, not dashboards
Output
A 300-word pipeline narrative ready to send to CFO and sales leadership
Workflow 05
35 minutes
Case study from a customer call recording
Otter.ai / Fireflies · Claude / ChatGPT
1
Get the transcript of a recent customer success or QBR call. Most video tools auto-transcribe
2
Skim the transcript for: the problem they had before, why they chose you, what they implemented, and any specific result they mentioned
3
Use the case study prompt from the AI for B2B Content guide. Paste the relevant sections of transcript, not the full thing
4
The AI will produce a draft. The most important edit: check that every metric and quote is verbatim from the transcript. Do not let the AI embellish
5
Send the draft to the customer contact for approval before publishing. Keep the approval request simple: "Does this accurately reflect your experience?"
Output
A 400-600 word case study draft with situation, approach, result, and one verified customer quote
Workflow 06
20 minutes
Landing page copy for a specific buyer persona
Claude / ChatGPT
1
Choose one persona for this page. Do not write for everyone. A page written for a VP of Engineering and a Head of Revenue is written for neither
2
List the three things this persona cares most about in a purchase decision. If you do not know, check your last five sales call notes for this persona
3
Use the landing page persona prompt from the AI for Demand Gen guide
4
Review the headline. It should name their specific priority, not a generic benefit. Rewrite it if it sounds like it could be on any SaaS website
5
Test the CTA. "Request a demo" is the weakest option. "See how [specific outcome] works" converts better for most B2B audiences
Output
Complete landing page copy: headline, subhead, three benefit bullets, social proof, and CTA
Workflow 07
15 minutes
Repurpose one article into five assets
Claude / ChatGPT
1
Take any published article or long-form piece. It needs to be at least 800 words with a clear argument
2
Paste the full text into Claude or ChatGPT
3
Use this prompt: "From the article below, produce: 3 LinkedIn posts (each a different angle, each for a different buyer persona), 1 email newsletter section (200 words, one CTA), 5 pull quotes formatted for design, 1 two-minute SDR talk track on the article topic. Article: [PASTE]"
4
Review each asset. The LinkedIn posts will need the opening line rewritten. The AI defaults to weak openers
5
Schedule the LinkedIn posts across three weeks. Do not post all three in the same week
Output
3 LinkedIn posts, 1 newsletter section, 5 pull quotes, 1 SDR talk track
Workflow 08
45 minutes
Positioning statement using the April Dunford framework
Claude / ChatGPT · Customer interview notes
1
Before running this workflow, you need inputs from real sources: at least 3 customer interviews, or 5 closed-won deal notes from your CRM, or both
2
Extract: who your best customers are, what alternatives they considered, what made them choose you, and one specific outcome they got
3
Use the positioning prompt from the AI for Product Marketing guide with these real inputs
4
The AI will produce a positioning statement and three value themes. Share these with two salespeople. Ask: "Would you say this in a discovery call?" If the answer is no, the positioning is still too marketing-facing
5
Iterate once with feedback. Do not iterate more than twice. Positioning that survives two rounds of sales review is good enough to ship
Output
A positioning statement, three value themes, and a list of five things not to say in your positioning
Previous
← AI for revenue ops
Get the Digital Uncovered newsletter
B2B marketing intelligence for India. No padding. No filler. Straight to the point.