Learn AI for marketing and sales

Two courses, three levels each, 37 three-minute lessons in all. Each is a few quick questions with something to do at work this week. When a topic runs deep, a card links to the wiki. Free, no sign-up, and your progress stays on your device.

Marketing22 lessons Sales15 lessons

Using AI in marketing: research, content, search, ads and automation.

Start lesson 1 Not sure where to begin? Find your level in a minute

Basic

Start here

What AI is, how to ask it well, and how to catch it making things up.

You have tried a chatbot but you are not sure what it is doing.

0 of 7 done

What AI actually is A well-read guesser that writes what sounds likely. Up next

Why it forgets, goes stale and varies Context, cutoffs and one-off good answers. Up next

Asking well: the four-part prompt Role, context, task, format. Up next

When it makes things up Spot a hallucination before it reaches a customer. Up next

Sort every task by the cost of checking Cheap to check, costly to check, never ask. Up next

What never to paste Keep customer data and secrets out of the chat. Up next

Your first three useful jobs Where to start so it saves time this week. Up next

Medium

Working week

Research, drafts, search, ads and a way to prove it saved time.

You use AI now and then and want it in your weekly routine.

0 of 7 done

Research in an hour, not a morning Give it the sources, ask for evidence. Up next

Research in your customers’ own words Bring the real material and ask for exact phrases. Up next

Drafts that do not sound like AI Voice samples, a ban list, then a human edit. Up next

Search when people ask AI Be the clear answer, not just a ranked link. Up next

Ads: more variants, better tests AI makes volume. You choose the hypothesis. Up next

Where AI belongs in your week Delegate, collaborate, do it yourself, do not touch. Up next

Prove it saved time A simple before-and-after you can show anyone. Up next

Advanced

Automate it

Workflows, agents, guardrails and whether the automation pays.

You already use AI weekly and want it running without you.

0 of 8 done

Write the workflow down first If you cannot write the steps, you cannot automate them. Up next

Workflow or agent? Fixed path or a system that decides its own steps. Up next

Build a weekly research brief Trigger, collect, summarise, deliver. Up next

Where agents pay, and where they do not High volume, low stakes, easy to undo. Up next

Guardrails and testing Make it safe to leave running. Up next

Judge a tool, a claim or a demo A demo is a selected case. Run your own pilot. Up next

Does it pay for itself? Hours saved against cost and risk. Up next

What you keep Judgement, taste and who is accountable. Up next

Basic

Start here

How a business sale works, who is in it, and how to ask the right questions.

You are new to selling to businesses, or you come from marketing.

0 of 5 done

Who is in a business sale Six roles, one deal, and where it gets dropped. Up next

Nuisance or priority? Four tests for whether a buyer will really act. Up next

Four kinds of question Situation, problem, implication, need-payoff. Up next

A demo is not a tour Show three things, not thirty. Up next

Where AI helps a seller Prepare faster, write notes, never guess. Up next

Medium

Run a real deal

Objections, proofs of concept, buying committees, APIs and the economics of a motion.

You can discover and demo, and now you are running deals with several people.

0 of 5 done

Objections are information The first one is rarely the real one. Up next

A proof of concept, not a free trial Write the finish line before anything is built. Up next

Read the buying committee Titles are not roles. Up next

APIs in plain words The menu, the waiter and four status codes. Up next

Fit the motion to the deal The way you sell has to match what you sell. Up next

Advanced

Run the numbers

Sales process, deal arithmetic, and automating follow-up without losing the buyer.

You run deals and want the pipeline to be honest.

0 of 5 done

A stage is something the buyer did Why "proposal sent" fills with hope. Up next

Five good chances make one modest one The survival model for a committee. Up next

Cover every seat Threads as arithmetic. Up next

Test the plan, not the payback Which input moves first? Up next

Automate follow-up, keep the buyer Guardrails for AI in a deal. Up next

Too deep for a lesson? Look it up

The course wiki has 38 short entries on the ideas behind the lessons, from context windows and prompt injection to buying committees and deal arithmetic. Lessons link to it when a card needs more room, and every entry links back to the lessons that use it.

Browse the wiki

Want the reading version?

Five longer tracks with the articles behind these lessons, what I tried myself, and something to do this week. Each one says how much of it I have run myself.

Track 1 AI basics for marketers What AI can and cannot do, in plain words, so you stop guessing.

Track 2 Marketing with AI Research, content, SEO, social and reporting: what to hand to AI and what to check.

Track 3 AI automation in marketing Write the workflow down, automate it, and check that it pays for itself.

Track 4 AI in performance marketing Ads, creative tests and measurement, now that Google and Meta run much of the buying.

Track 5 AI around marketing The parts of the business next to marketing: sales and RevOps, pricing, rules, costs and jobs.

Find your first AI workflow · The full archive