有些话，机构是不能说的。

Most marketing problems in this industry are not invented in a marketing department. They are written by a counsellor under pressure, in a message to one family, trying to keep a student who is being courted by three other offices. The sentence works. It also becomes the thing the agency is held to when the outcome is different.

The categories below are the ones that recur, and each of them has the same shape: a claim about an outcome controlled by somebody else.

Outcomes that belong to a third party.

Visa results.

No agency decides a visa. A statement that a visa is guaranteed, assured, or ninety-nine per cent certain is a statement about a decision the agency does not make, and it converts a refusal into a broken promise. The defensible version describes the process and the agency's own work: what will be prepared, what the known requirements are, and what the agency will do if the answer is no.

Admission.

Conditional offers are conditional and institutions change entry requirements between cycles. Presenting an application as a formality is the same error one step earlier in the process.

Scholarships.

Awards are decided by institutions on criteria they control and frequently revise. An agency may accurately describe what has been awarded in the past. It cannot promise what will be awarded, and the past figure should be described as what it is rather than as an expectation.

Claims about the agency itself.

The second family of claims is about status rather than outcome, and it is easier to make accidentally.

• Describing a relationship with an institution in terms stronger than the agreement supports. Recruiting for an institution is not the same as being its official or exclusive representative, and the word official is the one that gets used loosely.
• Using an institution's name, logo or imagery without the permission the agreement actually gives, which is often narrower than the office assumes.
• Implying an endorsement or accreditation that has not been granted, including by design: placing crests in a row under the word partners is a claim made in layout rather than in words.
• Quoting a success rate without saying what it measures, over what period, out of how many cases.

Numbers are claims too.

A statistic in a brochure is a promise about a measurement. If the office cannot produce the underlying figures on request, the number should not be published. That includes the round ones that feel harmless: thousands of students placed, a success rate quoted to two decimal places, a partner count that includes institutions the agency has never sent a student to.

The rule worth adopting is that every published number has a note behind it saying what it counts and when it was calculated, and that the note is written before the number is used rather than after somebody asks.

Where the exposure actually sits.

Not on the website. Websites get reviewed. The exposure is in the places nobody reviews: messages sent by individual counsellors, slides used at a school fair by a part-time recruiter, posts made by a sub-agent in another city, and the caption under a photograph of a smiling student on a social account run by whoever has the password.

Which is why the practical control is not a policy document. It is a short list of sentences that may not be used, given to everyone who speaks for the agency, including the people who do not think of themselves as marketing.

The test before it goes out.

One question, asked of any claim: if a family holds us to this exactly as written, in nine months, can we show that it was true when we said it. If the answer needs a qualification, the qualification belongs in the sentence rather than in the reply to the complaint.

Most of the time the qualified version still wins the enquiry. Families comparing four offices are not choosing the one that promises most; they are choosing the one whose account of the process sounds like it came from somebody who has done it before. A sentence that says what the agency will do, rather than what the consulate will decide, is both the safer claim and usually the more persuasive one.