AI for Marketers — Cohort 1
Build a real AI-assisted marketing campaign or growth engine for your own brand. 6 live sessions across 4 weekends, completion certificate.
Led by D Patel — two decades accountable for revenue, not just campaigns, with deep hands-on strength in paid media, marketing automation, and ROI measurement: the parts of AI marketing that actually move the number.
What you walk away with
- A real AI-assisted marketing campaign for your brand
- Prompt engineering mastery with the Prompt Stack
- AI copywriting, image and short-video direction skills
- Campaign planning with AI research tools, including ads inside AI answers
- AI visibility audit across ChatGPT, Google AI Mode and Perplexity, including which brand AI shopping agents pick, with a 90-day plan
- Your first supervised marketing AI agent, with a human approval gate
- Dakshio × SuperDesi completion certificate
Who this is for
Working marketers across agencies, in-house teams, freelance, content, performance, brand, social, and growth roles. You should be comfortable with marketing fundamentals. No prior AI experience required.
Format
- 6 sessions across 4 weekends — starting Sunday 24 January 2027
- Session time: 10:30 AM–1:00 PM IST (2.5 hours each)
- 30-min live kickoff: Saturday 23 January 2027, evening
Schedule
- Weekend 1 · Session 1Sun 24 Jan
- Weekend 2 · Sessions 2–3Sat 30 & Sun 31 Jan
- Weekend 3 · Sessions 4–5Sat 6 & Sun 7 Feb
- Weekend 4 · Session 6 + capstoneSun 14 Feb
Mastery resubmissions close Sunday 28 February 2027.
Full Curriculum
Six live sessions across 4 weekends, 24 Jan – 14 Feb 2027, 10:30 AM–1:00 PM IST. Every session has a pre-work, a worked example, a live build, a written deliverable, and a pass-or-iterate rubric. This is what real instructional design looks like.
Backward Design
Every session starts from the outcome and the evidence.
Bloom's Taxonomy
Verbs climb from Understand → Create, session by session.
Mastery Learning
Each session ships a pass/fail rubric — no fuzzy completion.
Kolb's Cycle
Do → reflect → conceptualize → re-test, every week.
Gagné's 9 Events
Pre-work, recall, worked example, practice, feedback — all wired in.
70-20-10
70% live build, 20% peer + critique, 10% formal frameworks.
JTBD Outcomes
Each deliverable maps to a real job your role gets paid for.
Kirkpatrick L1–L4
We measure reaction, learning, behavior shift, and business result.
Become AI-Fluent
Outcome
Operate any current AI assistant with a documented Prompt Stack, and produce a personal Brand Brain your future sessions will reuse.
Pre-work
30-min Loom: 'How LLMs actually predict' + a 10-question diagnostic that calibrates your starting Prompt Stack score (0–5). Plus a one-page check of which AI assistant you already pay for, so the live build uses tools you actually have.
Diagnostic
Self-rated AI fluency baseline; you re-take it at the end of the cohort to see lift.
Worked example
We rebuild a fashion DTC founder's Brand Brain live — voice, offers, customers, do-not-say list — in under 25 minutes, then run it in two assistants side by side to show that context, not the model, drives on-brand output.
Live build
Build your own Brand Brain (1,200–1,800 words) inside ChatGPT Projects, Claude Projects or a Gemini Gem, and decide what the assistant should and shouldn't remember.
Deliverable
Brand Brain v1 (Project + exported markdown) + 5 prompt templates saved to your library, plus a dated Model Map: which assistant to use for which job.
Behavior shift
By Monday you stop pasting one-off context — every prompt starts from a stored Project.
Mastery rubric — pass to continue
- Brand Brain contains voice, ICP, offers, claims, and do-not-say — all 5 sections
- At least 3 prompt templates run end-to-end with reusable variables
- You can explain when to NOT use AI on a task
Reflection prompt
"Where in your last week of work would the Brand Brain have saved you 30+ minutes?"
AI Copywriting that Ships
Outcome
Generate on-brand long-form and short-form copy at 3–5× your current pace, with a critique loop that catches AI tells before publish.
Pre-work
Submit 3 pieces of your existing brand copy, plus screenshots of any ads where Meta or Google rewrote or generated your text. We'll critique them live.
Diagnostic
AI-Tells audit on your prework — count the giveaways in your current AI use, including 'platform tells': ad text the platform generated that your brand never approved.
Worked example
A 600-word landing page rewritten 4 times using the Critique-Refine loop — we watch quality climb measurably. Then we run Meta's auto-generated text variations through the same loop to see what would have shipped unreviewed.
Live build
Ship a complete copy set for your brand: 1 landing page hero, 5 ad variants, 3 emails, 10 social posts (one in an Indian language your audience reads, checked by a native speaker).
Deliverable
A 'Copy Pack' — the full set above, plus 3 critique-loop templates you can re-run on any future copy, and a restricted words and claims list ready for Meta and Google brand settings.
Behavior shift
Your team stops asking 'is this on brand?' — your loop already answered it, including for text the ad platforms generate.
Mastery rubric — pass to continue
- Zero unedited AI tells in final pack (we test against the AI-Tells checklist)
- Brand voice match scored ≥4/5 by a peer using the rubric
- Critique loop documented and reproducible by someone else
Reflection prompt
"What did the critique loop catch that you would have shipped on day 1 of the cohort?"
AI Visuals & The Ship Gate
Outcome
Direct AI visual tools with the precision of a junior art director, and apply the Ship Gate framework to publish-decisions.
Pre-work
Save 10 reference images you admire from your category. Bring 1 visual brief that's stalled and 1 real product photo to edit with AI.
Diagnostic
Score your last 5 AI images using the Ship Gate rubric. Most score <2/5 on first attempt.
Worked example
We take a stalled visual brief, run it through the 4-prompt direction process, and ship-gate the outputs live. Then we turn the approved hero into a 6–8 second clip and ship-gate that too.
Live build
Produce 6 publish-ready visual assets for your brand: hero, 2 product/lifestyle, 3 social. Stretch: one social asset can be a short AI video clip, held to the same gate.
Deliverable
Visual Pack — 6 ship-gated assets + your prompt library with seed/style refs documented, plus a dated Visual Tool Map (best tool per job, free tier, commercial-use terms).
Behavior shift
Designers/agencies stop being the bottleneck for first drafts — you arrive with directed options.
Mastery rubric — pass to continue
- Each asset passes all 5 Ship Gate criteria (brand, anatomy, type, claims, IP)
- Prompts are reproducible: a teammate can regenerate consistent results
- You can articulate what you chose NOT to ship and why
Reflection prompt
"Which Ship Gate criterion did you fail most often, and what's the fix?"
AI-Assisted Campaign Planning
Taught from real P&L and automation experience — not theory.
Outcome
Produce a defendable AI-assisted campaign plan with research, offer, channels, creative, and measurement — that survives a CMO challenge.
Pre-work
Pick one real business objective (yours or your employer's) you'll plan a campaign for. Bring last quarter's numbers if you have them, and note which ad automation (AI Max, Advantage+) is already switched on.
Diagnostic
Walk through your current planning process — we identify where AI compresses 4 hours into 40 minutes.
Worked example
We plan a launch campaign live for a real SuperDesi-portfolio brand: deep research → ICP → offer → channel mix → creative brief in 50 minutes, marking which AI claims were checked against a source and which were thrown out.
Live build
Draft your own one-page campaign plan using the Dakshio Plan-on-a-Page template, AI-supported at every step.
Deliverable
Campaign Plan v1 — research appendix, ICP, offer, channels, creative brief, measurement plan. Includes one AI-assisted paid-media creative + brief (Meta or Google Ads), a call on whether to test ads inside AI answers (ChatGPT, Google AI Mode), and what the platform's AI is and isn't allowed to change.
Behavior shift
Your next planning meeting opens with a draft, not a blank doc.
Mastery rubric — pass to continue
- Plan answers all 7 sections of the Plan-on-a-Page
- AI contribution is documented per section (what you used, what you discarded)
- Peer in your cohort can poke holes; you respond with data, not opinions
Reflection prompt
"Where did AI accelerate you, and where did it confidently mislead you?"
Search & AI Visibility (GEO/AEO)
Outcome
Make your products the ones AI shopping agents choose. Audit how AI answer engines and shopping agents surface your brand, and ship a 90-day plan to improve visibility on the queries that matter.
Pre-work
Run 10 buying-intent queries about your category in ChatGPT, Google AI Mode (or AI Overviews), and Perplexity. Note who shows up and which sites get linked.
Diagnostic
Your brand's current AI Visibility Score (0–100) across the AI engines your customers use.
Worked example
Live audit of a real D2C brand: we lift their visibility score by identifying the 5 missing content assets, then check the same query in an Indian language. Then a short agent-shopper demo: we give an AI assistant a real buying task and watch which brand it picks, what it read to decide (price, stock, delivery, returns) and whether checkout happens in the chat or on the brand's site.
Live build
Audit your own brand across ChatGPT, Google AI Overviews and AI Mode, Gemini, and Perplexity — mentioned, accurate, recommended, cited — and, for product queries, whether the shopping agent chose your brand and where the purchase would finish. Run 3 of your 20 queries as shopping tasks, score one product page on the agent-readability checklist, then build the 90-day GEO/AEO plan with content briefs, feed and structured-data fixes, distribution, and tracking.
Deliverable
AI Visibility Audit + 90-day plan, with 5 prioritized content briefs.
Behavior shift
Your content calendar is now driven by AI-query gaps, not vibes.
Mastery rubric — pass to continue
- Audit covers ≥20 queries across ≥3 AI engines
- Each gap mapped to a specific content/authority/structured-data fix
- 3 queries run as shopping tasks, and one product page scored on the agent-readability checklist
- Plan has owner, deadline, and a measurable visibility KPI per item
Reflection prompt
"Which one fix, if shipped this month, moves your visibility most?"
Automate & Prove
Taught from real P&L and automation experience — not theory.
Outcome
Ship your first working agent — a supervised no-code marketing agent with a human approval gate — then integrate all six deliverables into one presented capstone with a measurement plan you can defend.
Pre-work
List 5 weekly marketing tasks that bleed time. Skim a 10-min explainer on triggers, actions, and webhooks, plus a 5-min 'automation vs agent' explainer. Import the Dakshio lead-agent starter template and connect Gmail and Google Sheets (Make's free plan is enough). Assemble your five prior deliverables into the capstone template, so the session only finishes it.
Diagnostic
Map each task on the Safe-to-Automate matrix (reversible × low-stakes) and pick one to build. Then map each tool your agent could use (look up, write, send): anything that sends a message or writes outside your team gets an approval gate.
Worked example
The 'Lead Triage and First-Reply Agent', built live: a consent check runs first, then the agent reads your Brand Brain, looks up the lead's website, scores fit against your ICP, drafts a first reply in your Copy Pack voice and logs it all to a CRM sheet. Nothing is sent until a human approves it in Slack or Gmail. Then we break it twice: an unclear lead goes to a human queue, and a model outage switches to the fallback.
Live build
Two parts: (1) about 30 minutes on your own supervised agent in n8n, Make or Zapier: your Brand Brain, at most 3 tools, a step limit, an approval gate on every send or external write, and one fallback path, run on at least 2 real inputs. If your task suits a fixed flow better, a plain automation with an AI step meets the same rubric. (2) about 20 minutes finishing your capstone and ROI one-pager, then present it in a pod of 5 (3 minutes each, recorded).
Deliverable
Running agent (or automation) + Ops Card + a presented capstone (recorded) + a one-page ROI/measurement plan.
Behavior shift
You stop seeing repetitive marketing tasks as work — you see them as supervised agents and flows waiting to be built, with proof attached. You decide which actions an AI may take alone and which wait for you.
Mastery rubric — pass to continue
- Flow or agent runs successfully on at least 2 real inputs
- Includes one fallback path and one human approval gate on a send or external write
- Ops Card written so a teammate could maintain it without you
- Capstone presented; measurement plan names a KPI, a baseline, and a target for each campaign element
Reflection prompt
"Across your capstone, which single metric would most convince a skeptical CMO you've leveled up?"
Ready to build real AI capability?
Limited to 40 learners. 7-day money-back guarantee.