Cost per lead lies: the economics behind an inquiry

“What should a lead cost in our category in Russia?” I get this question in the first call more often than any other, and there is no honest answer to it. Not because the market is opaque, but because the question is built wrong.

A lead is not a unit of anything. It is a label stuck on events that have nothing in common. A PDF download is a lead. A callback request is a lead. A site visit booked by an engineer who will quote a six-figure contract is also a lead. The first costs almost nothing and means almost nothing. The third is rare and drags a large invoice behind it, with add-ons. Comparing their prices is comparing weight in kilograms against length in metres.

When a company arrives with “get us to the market rate for a lead”, the rate in question came from a chat group, a conference slide, or a competitor’s case study. It describes someone else’s niche, someone else’s site and someone else’s sales team. About your business it says nothing at all.

The number that should govern your budget is not in any benchmark. It is inside your own economics. Here is where it sits and how to get from it to targets you can actually manage.

Why a benchmark from your home market travels badly

Take two companies in the same product category. Their conversion rates and their cost per inquiry can differ several times over, and the reason is rarely the advertising account. It is the landing infrastructure. One has dozens of narrow pages, each built for a specific edge of demand. The other has a single page trying to speak to everyone at once. The campaigns can be configured almost identically and the economics still diverge by a multiple.

For a company entering Russia this gap is usually wider than at home, and for a predictable reason. The Russian site is almost always the last thing built and the first thing economised on. It gets translated from the English original, keeps the English page structure, and inherits a menu designed around a different buying habit. Meanwhile the paid traffic arriving at it is priced by a live auction against local competitors whose sites were built for this audience from the start.

So the benchmark you carry over describes a funnel you do not have. And a cheap lead is often an empty one. Someone left a phone number in passing, nobody has a reason to warm them up, and they will not answer the call. An expensive lead can be a single event that closes into a contract with add-ons. The price on its own means nothing. What means something is the pairing: what it costs, and what stands behind it.

The pass-through price of a deal, and who knows it

Exactly one figure genuinely governs the budget: the pass-through price of a deal. How much you can pay to acquire a signed contract and still have that contract be profitable.

No industry chat knows it. No agency knows it, including me on the first call. Only you know it, and it comes out of your unit economics: average order value, margin, the rate at which inquiries become sales, the share of add-on revenue.

Which is why the first money conversation is not about campaign setup. It is about the funnel, broken into steps. Conversion into an inquiry, which is the site and the ads working together. Conversion into a qualified inquiry, meaning how many survive the first contact. Conversion from inquiry to sale, which is about lead quality and about your sales team. Without those three numbers the target cost per lead is picked out of the air, and the expectations built on it will not survive contact with reality.

There is a Russia-specific wrinkle in the second step that foreign companies consistently miss. A large share of qualified contact here happens in messengers rather than by email, and a meaningful part of it happens on WhatsApp and Telegram rather than on the phone. If your tracking only counts form submissions and calls, your conversion into a qualified inquiry is understated, sometimes badly, and you will conclude the traffic is worse than it is. Fix the measurement before you judge the number.

If a client insists on “just launch it, we will look at the economics later”, that is not saved time. It is a postponed argument. A month later they will say the inquiries are wrong. They will be right by feeling and wrong by cause, and nobody measured the cause at the time when it could still be measured.

A target CPA is derived from below, not announced from above

The common mistake: someone names the cost per action they want and waits for the account to produce it. It does not work that way, because cost per action is not a thing you control. It is a result.

In an auction you hold one real lever: the bid. Through the bid you influence the click price. The click price, combined with the conversion rate of the site, produces the cost of an inquiry. That is the whole chain, and it runs upward from the bottom.

So the calculation runs the same direction. First you establish the target click price your business can carry. Then, on an account with accumulated history, you look at how bids actually translate into click prices in your categories: what the ratio is and how it moves week to week. Through that ratio you learn which bids produce which click prices. Only at the end do you arrive at the target cost of an inquiry, by multiplying the target click price by the conversion rate of the site.

Yandex Direct will let you skip all of this and simply type a target cost per action into an automated strategy. It will accept the number. It will then attempt to reach it with the only lever it has, and if your number implies a click price that loses most auctions in your category, the campaign will go quiet instead of getting cheap. Quiet campaigns are how a demanded CPA announces that it was never achievable. Nothing in the interface tells you that is what happened.

Cheaper clicks hit a wall, conversion does not

There is a standing temptation to lower the cost of an inquiry by lowering the cost of a click: bid less, pay less.

That path has a hard ceiling. To pay less you have to lose more auctions, appear less often and lower on the page, and that cuts traffic volume and conversion count at the same time. Push too far down and you lose position and end up worse off overall. Cheapening clicks is movement along a wall that your competitors built.

Raising the conversion rate of the site has no such ceiling. It scales the return on every click you have already bought: at the same click price, a higher conversion rate automatically yields a lower cost per inquiry. And the effect accumulates. Relevant pages, worked out for specific segments of demand, deliver a gain that stays with you and keeps working.

Over a couple of years, a durable fall in the cost of an inquiry almost always walks hand in hand with a rising conversion rate rather than with sharp bid manoeuvres. The economics of a project on a long horizon are decided by the pairing of advertising and landing page, not by the account alone. This is also why I treat Yandex Ads and the pages behind them as one job rather than two.

Saved spend is not profit

The classic argument: should we switch off ads on our own brand, given that we already rank first organically? Why pay for a click that arrives free?

The logic sounds airtight and it has one weak joint. It sees the saved expense and does not see the forgone revenue.

Brand advertising holds the top of the page, pushes out competitors and the intermediaries intercepting demand you created, and gives you full control over the headline, the sitelinks and the landing page. Give up the slot and it opens for someone else’s ad. The saving appears immediately, in the expense line. The lost income appears almost nowhere. That is what makes a loss-making decision look like a win.

In Russia this asymmetry is sharper than in most Western markets, because the paid slots above your organic listing are competitive real estate for aggregators, marketplaces and resellers who are entirely happy to sell your product to your customer at their margin. A foreign brand that turns off its own brand campaign here usually discovers the effect a quarter later, in the sales figures, with no line in any report explaining it.

The same holds for any “economising” inside the account. Cutting the budget, switching off a segment, lowering a bid is easy, and the spend figure obediently drops. The question is always the same: what happened to the income. If there is no answer to that, it is not optimisation. It is guessing.

What to work out about your own inquiries this week

A short route you can walk right now:

1. Define what counts as a lead. One word is being used for several different events. Reduce it to specifics: a call, a form, a messenger conversation, a site visit, a contract.
2. Extract the pass-through price of a deal. Average order value, margin, share of add-ons, and how much you can pay for a deal and stay in profit.
3. Break the funnel into steps. Into an inquiry, into a qualified inquiry, into a sale. Wherever a number is missing, there is a hole your decisions are leaking through. Check that messenger conversations are counted.
4. Build targets from below. From click price to cost per inquiry through the conversion rate of the site, not the other way round.
5. Find where the lever is. If the cost of an inquiry is high, answer honestly which is cheaper to move: the click you have nearly squeezed dry, or the conversion rate you have not touched.

A cost per lead is not a number from someone else’s chat and not a line on a price list. It is a derivative of your economics, and until the economics are laid out, any figure will either frighten you or reassure you without either reaction being earned.

The counting starts not with advertising but with how much the business can pay for a deal. Everything in the account is downstream of that.

Frequently asked questions

Sources

• Yandex Direct help, bidding strategies and target cost per action: yandex.com/support/direct/en/strategy
• Yandex Metrica help, goals and conversion tracking: yandex.com/support/metrica/general/goals
• Russian version of this article on belokrylovo.ru: Цена лида врёт: считаем экономику заявки