An online analytics school in Almaty sells a course for 25,000 ₸: 15 payments a week, 60-65 a month. The owner wants to "add some psychology" to the course page and to the Instagram ads: a "registration closes in 23:59:59" timer, a "3 places left" badge and a block of five-star reviews with the line "This changed my life". The timer from the site builder's template restarts every time the page is reloaded, nobody has counted the places, and nobody at the school remembers who wrote the review. The numbers are illustrative.
The timer and "3 places left" are scarcity, the reviews are social proof: two of Robert Cialdini's principles of persuasion. The principles describe what people in experiments responded to more often, but you can only show them on a page when something stands behind them. Below are all seven, each with an example for the course page and for an ad, and what has to back the example up.
Where the seven principles come from
In the first edition of "Influence" (1984), Cialdini had six principles: reciprocity, commitment and consistency, social proof, liking, authority and scarcity. He described the seventh, unity, in his book "Pre-Suasion" (2016): a shared identity, the sense of "we". In "Influence, New and Expanded" (Harper Business, 2021), unity joined the list as the seventh principle. Short descriptions of all seven are on the site of Influence at Work, Cialdini's company.
The principles grew out of experiments, and an experiment means particular people in a particular setting. In 2008 Goldstein, Cialdini and Griskevicius showed in one American hotel that guests reused towels more often when a sign said other guests did so than when it simply asked them to protect the environment. In 2014 Bohner and Schlüter repeated the experiment in two hotels in north-west Germany, and the sign about other guests did no better than the usual one. In the second hotel both signs raised towel reuse compared with a week without signs. This happens often enough with psychology experiments, which is why every principle on your page is a hypothesis to check against your own leads and payments.
Seven principles on the course page and in ads
| Principle | On the course page | In an ad | What has to be true |
|---|---|---|---|
| Reciprocity | A recording of an open lesson that is useful without buying | "Open lesson: we'll build a sales report in an hour" | The lesson delivers a result by itself, with no "the rest is in the course" |
| Commitment and consistency | First sign-up for the open lesson, then the course offer | The ad leads to the open lesson rather than straight to payment | Each next step is the person's own choice, with no pre-ticked boxes |
| Social proof | Graduate reviews with context, the number of graduates | A graduate in a video says what they did before the course and after | The review is real, the author agreed, the number comes from the CRM |
| Liking | The instructor talks about themselves and how they teach | The video_teacher video, where the instructor speaks in person | The person in the video actually teaches the lessons |
| Authority | The instructor's experience: where they worked, what they built, a profile link | "Taught by an analyst with retail experience" | The experience can be checked through the link |
| Scarcity | Start date, registration deadline, places left | "Starts 2 November, register by 30 October" | Registration closes after 30 October, places are counted from payments |
| Unity | The graduate club at 1,000 ₸ a month: reviews and partner job openings | "A course from Almaty analysts, followed by a graduate club with partner job openings" | The reviews run on schedule, the job openings come from real partners |
The open lesson the school planned to run live to test the mini-course idea can be posted as a recording afterwards. It covers two rows of the table at once: reciprocity and a small first step.
You don't need all seven blocks on the page. Start with whatever stops people from signing up. At the school, that's the course contents: 12 of the 30 WhatsApp questions over two weeks were about what the price includes, and the answer to those is the "What 25,000 ₸ includes" block from the hypotheses for a landing page, not a Cialdini principle.
A review with a person behind it
A review saying "This changed my life" may be genuine, but it tells a prospective student nothing: who wrote it, what they started with, what they got. A useful review answers the three questions in the picture: who the author is (a name and photo, or at least a job title), where they were when they started, and what they achieved over what period. For example: "I work as an economist at a trading company and used to compile sales in Excel by hand. During the course I built a sales dashboard, and now the report updates itself." Underneath go the name, job title and stream, and it is published with the author's consent, covered below in the section on the law. The result in a review is what the graduate wrote, and you can't add "landed a job as an analyst" on their behalf if they didn't say so.
Numbers in your evidence get the same treatment. The spring stream had 60 students. 24 answered the post-course survey, and 18 of them said they use the skills at work. A caption saying "75% of graduates apply what they learned" takes 18 of 24 and passes it off as a share of all 60. There's no data on the 36 who didn't answer, and the ones who stayed silent may be exactly those it didn't work for. An accurate caption reads: "24 of 60 graduates of the spring stream answered the survey; 18 of them use the skills at work." If you need a single figure, the lower bound is 18 of 60, or 30%.
Scarcity you can verify
A timer that shows 23:59:59 again after a page reload is what the FTC, in its report on dark patterns that also covers the default tick box, calls a baseless countdown timer, and "3 places left" while registration is wide open is a false low stock message.
The school has real scarcity. A curator checks every student's assignments, each curator has 20 students and there are three curators, so a stream has 60 places. The stream starts on 2 November and registration closes on 30 October. That's what goes on the page:
- the start date and the registration deadline, and after 30 October registration really does close and the page offers the next stream;
- the number of places left, if it's counted from payments in the CRM rather than typed in by hand;
- the reason for the limit in one line: "20 students per curator, because the curator checks every assignment".
Then even "3 places left" a few days before registration closes can be accurate: if 57 people have paid for the course.
What the law says
The Law "On Advertising" calls advertising unreliable when it contains false information about availability on the market (Article 7(2)(2)), about expected results (Article 7(2)(6)) or about the status or competence of the producer, the seller or those advertising them (Article 7(2)(14)), and when statistics are presented so they look better founded than they are (Article 7(2)(13)). As we read it, "3 places left" with open registration, "75% of graduates" from a survey of 24 people and an instructor's experience that never existed fit these points when they appear in advertising. The law doesn't say outright whether the course page itself counts as advertising, but its definition of advertising is broad (Article 3, subparagraph 1-2), and it applies the requirement to be truthful explicitly to advertising in e-commerce (Article 6(1)). We read a resetting timer as false information about the course's availability on the market (Article 7(2)(2)), and if the advertiser knows the timer resets and leaves it up, as we read it, this may also count as knowingly false advertising (Article 7(4)).
A review with a graduate's name and photo on the school's website falls under the Law "On Personal Data and Their Protection". It lists full name and an image of the face among the identifiers of personal data (Article 6), and allows personal data to be placed in publicly available sources only with the person's consent (Article 7(3)). So, as we read it, such a review, like a graduate's story in a post or video, is published only with their consent. Article 8(4) lists what the consent must contain, including whether the data may be published in publicly available sources (Article 8(4)(6)), and the school has to be able to prove it was given (Article 25(2)(5)), so the consent is kept. If the review goes out as sponsored content on an online platform, for example through a blogger, the Law "On Online Platforms and Online Advertising" requires recommendations and reviews in it to be genuine and verifiable (Article 21(3)) and allows other people's opinions, photos and videos in it only with their prior consent (Article 21(2)(2)). This is our reading of the laws as amended up to 17 May 2026 ("On Advertising"), 14 September 2026 ("On Personal Data and Their Protection") and 2 March 2026 ("On Online Platforms and Online Advertising"), not legal advice.
Every tactic is a hypothesis
Say the school started with the reviews and rewrote three of them as author, context, result. That's a hypothesis, and it gets written up with the same template as the graduate club hypothesis. When there are several ideas, the queue is ordered with ICE or RICE. The school lacks the traffic for an A/B test here, as with its other landing page changes: the change is made and observed through the log, and the A/B test waits until there is more traffic.
Leads aren't the only thing to watch. A tactic that lifts leads with a promise can damage what comes after them: fewer people get as far as a conversation with the manager and as far as paying, and more ask for a refund in the first 14 days. The school has already seen something similar: in the week the program carousel brought 30 leads instead of 14, the manager got through to only 11 of them, and none of them paid that week (the breakdown on the dashboard). The school hasn't found out why yet, but a rise in leads without payments is a reason to stop and look into it, not to add budget.
Assignment: the seven principles on your page
Take the landing page and ads of a project you run ads for, or the project from the Start track, and fill in a table with a row for each of the seven principles. The first row is the school's example.
| Principle | What's on the page and in the ads now | What backs it up | What to do |
|---|---|---|---|
| Scarcity | A "registration closes in 23:59:59" timer | Nothing: it resets when the page is reloaded | Remove it, show the registration deadline |
The last column takes one of three answers: keep; remove or rewrite straight away, with no test; or write a hypothesis with a guardrail metric (payments, refunds, or the share of leads the manager reached) if the principle isn't on the page. Below the table, list every number on the page and its denominator.
The review starts with the "What backs it up" column. If it says "everyone does it" or "the client said it was like that", there's no proof, and the block comes down until there is. A number without a denominator gets rewritten the way the school's "75%" was: how many answered, out of how many, and what exactly they said.
Finally, rewrite one line. The school removed the timer but left "Last chance to register" above the button, although after 30 October registration simply moves on to the next stream. What should the line say?
The look of scarcity went with the timer, but the false promise stayed: "last chance" says you won't be able to sign up later, and that's inaccurate. In an ad, as we read it, a line like that gives the same false information about the course's availability on the market as the resetting timer. The line should state the real condition: "Registration for the stream starting 2 November closes on 30 October". The school's scarcity is a date, and the word "last" adds nothing accurate to it.
Video: an overview of the book
A retelling of Cialdini's book from the Knizhny Ryad channel. As you watch, note for each principle what on your page could back it up.