Class 71 closed the community flywheel: content attracts people, members contribute, contributions become searchable assets. Class 72 opens a different engine. Instead of earning attention slowly or waiting for search, you buy a measured amount of attention to answer one specific question. This lesson makes that engine legible before you spend anything.

Paid means rented, measured, and stoppable

Paid advertising buys placement in front of people you have not yet earned. You pay a platform or publisher to show your message, and in return you get speed: exposure within days instead of months. You do not get trust, relevance, or forgiveness. Those still have to be earned by the message and the destination.

Learn the working vocabulary in one pass, because every later lesson in this class reuses it:

• Paid advertising is purchased exposure with a stated objective and budget cap.
• An impression is one delivery of your ad to a screen. It does not mean anyone noticed.
• An audience is the set of people eligible to see the ad, defined by place, interest, behavior, or context.
• A placement is where the ad appears: a feed, a search result, a video slot, a newsletter slot.
• A bid is what you are willing to pay for an outcome; a budget is the most you will spend while learning.
• A campaign holds the objective and budget. An ad set or ad group holds the audience and placement choices. The creative is the actual message people see.
• A landing page is where the click arrives. A conversion event is the one useful action you count — booked call, sample request, completed template.

The durable mechanic matters more than any dashboard layout. Before touching platform settings, browse real ads to calibrate your eye. The Meta Ad Library shows running ads across Meta surfaces with their copy and creative, and the TikTok Creative Center shows trending formats and top-performing patterns. Use both as observation decks, not templates: what promise does each ad make, who is it for, and where would the click go?

The five-piece system

Every paid test is five pieces working together:

audience + message/creative + offer + destination + measurement

Change one piece at a time or you learn nothing. If you change the audience, the headline, the offer, and the page at once and results improve, you cannot say why — and you cannot repeat it.

Walk it through the local business launch, this part's running example. A tailoring shop with same-week repair capacity considers a small paid test:

• Audience: people within roughly ten minutes searching or browsing for tailoring, alterations, or event preparation.
• Message/creative: "Same-week hemming before your event — real pickup date at booking." One honest sentence, one real garment photo.
• Offer: a bookable fitting slot with a stated price range and turnaround.
• Destination: the booking page, not the homepage, not the social profile.
• Measurement: completed bookings from that page, counted for one week, against a fixed spend cap.

If bookings do not happen, the five-piece frame tells you where to look instead of concluding "ads don't work." No reach means the audience or placement definition is wrong. Reach without clicks means the message does not earn attention. Clicks without bookings mean the offer or destination fails. Tracking without a counted event means measurement was never wired. Each diagnosis points to a different fix.

When paid is the right route

Paid is useful when five conditions hold: a clear offer exists, the audience is reachable in an identifiable place, the destination page is usable today, the budget covers enough exposure to learn something, and the question is specific. "Will nearby people with an event deadline book a same-week fitting from this page at this price?" is a testable question. "Let's get our name out there" is not.

Paid is the wrong first fix for an unclear product. If the offer is vague, the page is confusing, or nobody can say what success means, buying traffic multiplies confusion at cost. Fix the offer and destination first — Lesson 72.2 exists for exactly that reason — then buy attention to test the fixed system.

For Sonariq, the Research Desk example, the same discipline applies. A sensible first question is not "promote the product" but "will analysts who value evidence click a promise of a source-linked sample brief and complete the sample?" The audience is narrow, the offer is a sample rather than a subscription, and the conversion event is a completed first brief. Narrow and specific beats broad and hopeful.

Exercise: write the first paid test

Create FIRST-PAID-TEST.md with these fields, each in one or two sentences:

• Objective: the single question this test answers.
• Audience: who, where they are, and why this placement reaches them.
• Offer: what they get and what it costs in money, time, or attention.
• Conversion action: the one counted event and where it is recorded.
• Creative variants: two versions differing in one element (headline or image, not both).
• Budget cap and duration: maximum spend and end date, stated up front.
• Review date and decision rule: when you read results and what each outcome triggers (continue, change one piece, stop).

A worked miniature: objective — "Do nearby event-goers book same-week hems from the booking page?" Audience — adults within 10 km searching alterations or engaging local event content. Offer — fitting slot, stated price band, pickup date at booking. Conversion — completed booking. Variants — same photo, two headlines ("Ready before Friday" versus "Same-week hemming, pickup date guaranteed"). Budget — fixed small amount, seven days. Decision rule — fewer than a threshold of bookings with healthy clicks means fix the page, not raise the budget.

Finish line: a FIRST-PAID-TEST.md test card a teammate could run without asking what success means.

Verify quickly: hand the card to someone unfamiliar with the project. If they can name the audience, the offer, the one counted action, and the stop date, the card is ready.

Common failure mode: launching a "follow us" or "check us out" campaign with no offer and no conversion event. Attention arrives, nothing is asked, nothing is learned, and the spend teaches only that vague ads produce vague results.

Check your understanding

1. Name the five pieces of the paid system and explain why changing all of them at once teaches nothing. 2. Why is a vague "get our name out" campaign untestable, and how would you rewrite it as a question? 3. No reach, reach without clicks, and clicks without bookings each point to which piece of the system?

Next

A test card is only as good as the place the click lands. Lesson 72.2 separates the ad's job from the destination's job, so your booking page or sample brief finishes the work your creative starts.