Most link campaigns are bought on publisher metrics and lost on topic. The rate sheet tells you DR, traffic and indexation; it rarely tells you what the article will be about. Then the vendor sends ten pieces with titles that could have been written for any client on any site, and the campaign produces links that rank for nothing and send no one.

The topic is the part that compounds. A link is a signal; an article is an asset that keeps getting read, cited and re-shared long after the link was placed. Here is how to brief it.

Start with the query, not the keyword list

A useful guest post topic answers a question your buyer actually asks — in the words they use, at the stage they ask it. Before approving any topic, ask for the search intent behind it:

  • Who types this? A founder evaluating vendors and a junior marketer doing a school report both search “PR agency.” Only one is a buyer.
  • What else would they need after reading? The next question a reader has is the topic of the next placement.
  • Can this claim only credibly be made by your client? A generic “10 tips” piece can be written by anyone, which means it should not be written by your client. A piece with data, in-market experience, or a contrarian operating view earns the byline.

If the vendor’s topic list has no reasoning attached — just titles — that is a content factory, and the titles will read like one.

The four-filter test

Run every candidate topic through these before it goes into a brief:

  1. Adjacency. Does the subject sit naturally inside the publisher’s beat? A careers index can plausibly publish on hiring markets and remote-work policy; it cannot plausibly publish on logistics software. Topical adjacency is the strongest single predictor of whether a placement survives, ranks and sends referrals.
  2. Durability. Will this be true in eighteen months? “Evergreen” is overused, but the underlying test is real: date-stamped trends decay, and a dead article on a live URL is worse than no article. Mechanisms, frameworks, comparisons and “how to evaluate X” topics age well; news cycles do not.
  3. Ownability. Does the piece say something the publisher’s regular editorial wouldn’t, but that the publisher’s audience wants? The best placements fill a gap in the site’s coverage — which is exactly why a real editor accepts them.
  4. Linkability. Can the sentence containing your link read naturally? If the only honest way to insert the link is “for more, see [client]” at the bottom, the topic is wrong. The right topic has a place where a citation belongs — a definition, a benchmark, a tool comparison.

A topic that passes all four is rare. One that passes three is worth running. One that passes adjacency alone is a link donation.

What to refuse from a vendor

Three topic patterns should come back every time:

  • “Ultimate guide” to a term your client doesn’t own. You’ll be competing with the publishers’ own editorial and the SERP’s established guides, with no authority advantage.
  • Listicles with no point of view. “7 ways to…” pieces get no links, get no shares and read like ads, which is the opposite of why you’re paying for placement.
  • Anything written around the exact-match keyword. The topic was chosen to force an anchor, and both will read that way. Choose the topic a human would click; fit the keyword into the body where it belongs.

The bilingual reality

On a dual-language publisher, a single topic becomes two editorial decisions — and they often differ.

Search behaviour in Traditional Chinese (Hong Kong, Taiwan) is not a translated mirror of English. Query phrasing, format preference and even the maturity of the commercial intent can be different for the same subject. A “vendor comparison” angle that works in English can feel presumptuous in a market where buyers expect to shortlist through referrals; a career-advice angle that reads generic in English can perform strongly in Chinese where structured guidance is a bigger search category.

So the correct brief is not one topic translated twice. It is one thesis, tested against two editorial calendars, rewritten natively for each. If a vendor can’t discuss the Chinese angle as an editorial decision rather than a translation step, they’re selling you one placement dressed up as two.

Sequence the topics like a campaign

One good article is a line item. Six is a narrative. The strongest authority campaigns we see mapped out look roughly like this:

  1. Open with an evaluation piece — “how to choose / how to vet X.” It’s the highest-intent topic, it earns links naturally, and it sets your client up as the person defining the standard.
  2. Follow with a mechanism or data piece — something only your client can write credibly, drawn from real operations.
  3. Then adjacency plays — a topic from the neighbouring category where the client has a defensible opinion.
  4. Leave gaps between placements. Two articles a month on different publishers, sequenced by topic, reads like authority. Eight in a week reads like a footprint.

Track referral traffic and ranking movement on the target page per article, not across the campaign. You’ll learn within a quarter which topic patterns work in which markets, and the second round of briefs gets sharper than the first.

The short version

Choose topics by adjacency, durability, ownability and linkability — in that order. Refuse anything written around a keyword. Treat the Chinese version as a second editorial decision, not a translation. And sequence the year so each placement sets up the next. The buyers reading your guest posts don’t know they’re in a link campaign, and that’s precisely the point: the article has to be worth reading on its own.

PressInAsia publishes permanent placements across six established bilingual publishers — including hkese.net at Ahrefs DR 55 — with topics briefed natively in English and Chinese for each masthead. Request rates and see live samples.