Paid media divides into two very different games, and most agencies only play one.
One game is volume: a consumer brand, a marketplace listing, a resort with rooms to fill. Thousands of buyers, short decisions, and success measured in cost per order. The other is scarcity: a few hundred people in the world can sign your contract, the cycle runs six months, and a cheap click from the wrong company is worse than no click at all.
We run both, and we do not use the same playbook for them.
For volume, the work is hypothesis testing and unit economics. A federal e-commerce retailer grew orders 39% year over year while cost per order fell 22% and cost per click fell 14% — no new budget, just a disciplined sequence of tests that killed what did not work fast enough to matter. For a Tyumen thermal spa, audience segmentation and a qualifying form doubled monthly leads from 66 to 125 and cut cost per lead 40%, in two months.
For scarcity, the work is qualification and patience: a narrow audience, strict exclusions, and a form that filters before it collects. THE HEIGHTS Bali is the clearest case — a premium resort complex selling $200,000 lots to a skeptical market, which produced warm leads from the first month at around 5,900 ₽ each, in a niche where the market cost per lead runs far higher.
What links them is refusing to report a number that does not correspond to money. Impressions are context. Clicks are context. The question is always what happened after.
Frequently asked questions
What budget do we need to start?
That depends far more on the market than on the number. In a scarcity market — a few hundred possible buyers, a six-month cycle — a modest budget goes a long way when targeting is strict, because the waste in that niche never comes from bidding. In a volume market you need enough to reach a readable result on tests inside a sensible timeframe. The honest answer comes after the account audit, not before it.
How soon will we see results?
Volume campaigns give a readable signal in weeks: the SPA resort cut its cost per lead by 40% inside two months. Considered purchases need a full quarter. At Skydent the largest gain arrived in month two, once statistics had accumulated; judged at week three it would have looked unfinished. We published the numbers from seven campaigns if you want the range.
Will you run the same playbook you use for everyone else?
No, and the two playbooks contradict each other. Volume work is hypothesis testing and unit economics; scarcity work is qualification and patience. Running a consumer playbook against a market of a few hundred buyers burns budget proving nothing. The first decision in any account is which of the two games it is.
What happens to the account if we stop working together?
You keep it, and everything in it: naming conventions, audiences, exclusion lists and the full test log. Accounts are built to be handed over. A tested account is worth more to you than a tidy report, and locking a client in with an account nobody else can read is not a business we want to be in.