When Ads Stop Working: Three Cases of Rebuilding the Funnel, Prepping the Site, and Cleaning Up YANThematic roundup · digest for October 23
Poster-style editorial graphic showing three funnel rebuild paths for office partitions, auto lighting, and CCTV
Thematic roundup3 sections≈ 5 min

When Ads Stop Working: Three Cases of Rebuilding the Funnel, Prepping the Site, and Cleaning Up YAN

When an ad campaign refuses to bring in calls or lead costs blow past break-even, the average marketer starts endlessly shuffling settings: tweaking daily caps, switching bidding strategies, rewriting sitelinks, and nudging bids by 5%. It's a classic illusion of control. In reality, micro-tweaks are powerless when the core link between the buyer's real pain point, the offer's precision, and traffic quality is broken.

PPC expert Leonid Streltsov (author of the blog "Digital Cartel," with more than 200 campaigns under his belt) walked through three contrasting niche service projects to show how a ground-up funnel rebuild pulls marketing out of the red.

Pleated Blinds and Office Partitions

Why zero leads mean you need to change the offer, not twist the bids:

The first cautionary tale started with a flat-out flop. An interior glazing and shading company decided to push pleated blinds. They launched ads on Search and in display networks, spent around 140,000 RUB, and got… 0 leads.

At that point, a junior specialist usually blames Yandex or suggests hiring another PPC manager. Instead, the team dug deeper: the "pleated blinds → generic catalog on a broad website" setup offered no clear market value, and the average order size couldn't survive an overheated auction.

Rather than keep burning budget, they rebuilt the funnel around the product itself:

  • Dropped pleated blinds and pivoted to glass office partitions—a high-ticket segment with clear B2B demand.
  • Built a dedicated conversion landing page around buyer pain points: added a project calculator, committed to strict clean-installation deadlines, and showed transparent turnkey pricing.
  • Targeted ads strictly at high-intent commercial search queries from business buyers.

The second launch delivered: 4,980 impressions, 406 clicks, and 30 qualified leads on 288,464 RUB spend (VAT included). Cost per lead came in at 9,615 RUB with a 7.39% landing page conversion rate—easily profitable given the high ticket of office partitions. Even factoring in the failed first test (bringing blended CPL to 14,282 RUB), the business walked away with a scalable revenue asset instead of a money pit.

Auto Lighting Workshop

Rewriting the landing page around real customer symptoms before spending a ruble on traffic:

The second project was a specialized workshop for repairing and upgrading car headlights. Before launch, their website read like a dry technical parts catalog: "projector lens replacement," "polycarbonate polishing," "retrofit."

Leonid hit pause before traffic ever went live. An everyday driver doesn't think in terms like "retrofit." They drive home on a rainy evening and mutter: "My headlights barely light up the road—I'm driving blind," or "My headlight fogged up inside after the car wash; I need to fix this."

They rewrote the entire website in the language of driver symptoms before buying a single click:

  • The hero section hit visitors with direct questions: "Headlights dim at night?", "Condensation inside the lens?", "Don't rush to buy expensive new headlights—we restore factory brightness in 1 day."
  • Under each symptom, they placed clear before-and-after photos listing the exact car model and repair price.

Only after that overhaul did they launch a Yandex Advertising Network (YAN) campaign. Over two months (April 3 to June 1, 2026), it posted strong returns: 804,661 impressions, 6,869 clicks (0.85% CTR, 36.10 RUB average CPC), and 84 conversions at 2,952 RUB (VAT included) on 247,978 RUB spend. Site conversion hit 1.22%, backed by continuous manual placement pruning and a tight match between the banner hook and the hero screen. For more on how buyer intent rescues conversion even without fancy design, see our breakdown: Why Ads Pay Off Even on a Weak Website: When Audience Intent Beats Perfect Design.

Video Surveillance

Why commercial keywords and YAN autotargeting must never share a campaign:

The third case was turnkey CCTV installation. Here, marketers routinely make a methodological blunder: mixing exact commercial keywords and Yandex autotargeting inside the same YAN campaign. The stats blur together, making it impossible to tell which traffic actually pays off.

The team split the launch into two isolated parallel campaigns:

  1. A campaign running strictly on commercial keyword phrases ("CCTV installation price," "warehouse camera setup").
  2. A pure autotargeting campaign matched to the landing page.

The test uncovered a surprising paradox:

  • Total spend was 168,740 RUB (VAT included), yielding 448,520 impressions, 2,441 clicks, and 82 target conversions at an average of 2,058 RUB.
  • The keyword campaign bought clicks at 45.23 RUB (excl. VAT), converted at 2.51%, and delivered 46 conversions at 2,198 RUB.
  • Meanwhile, the autotargeting campaign paid nearly double per click—91.06 RUB (excl. VAT)! Disaster? Not at all: landing page conversion jumped to 5.91%, bringing in 36 leads at 1,879 RUB—14.5% cheaper than the keyword campaign.

Autotargeting bought pricier clicks from a far warmer, higher-intent audience. Had both run together in one campaign, a shallow CPC check would have killed autotargeting—cutting off the business's most profitable lead source.

Takeaways for Marketers

All three stories point to one rule: stop trying to fix broken marketing with cosmetic tweaks inside the ad account. If the offer doesn't fit the unit economics, change the product. If the site speaks dead technical jargon, translate it into real customer symptoms. And when scaling YAN, separate your traffic streams so you can see the true price of every decision.

Want a deep diveVotes help us choose what to break down in detail. This material has 0 votes.
Sources · 4