Skip to main content

Programmatic Ads in Russia

RMAA helps international brands run programmatic ads in Russia through data-driven media buying, local DSP platforms, audience targeting, verified inventory, and transparent campaign control.

Contact our Team

What Are Programmatic Ads?

Programmatic advertising buys ad impressions automatically, using audience data, digital inventory, bidding logic, and verification tools to reach the right users.

Data-Driven Media Buying

Programmatic advertising is an automated media buying method that uses audience data, local ad platforms, and verification tools to buy digital ad impressions and reach specific users in Russia.

Audience, Not Just Placement

Programmatic advertising focuses on who sees the ad, not only where it appears. It helps brands reach users by behavior, interests, location, purchase signals, or other available audience data.

Local DSPs and Inventory

Programmatic campaigns in Russia can run through local DSPs and programmatic platforms connected to websites, mobile apps, video inventory, CTV, publisher networks, and other digital advertising placements.

Flexible Targeting Options

Programmatic advertising allows brands to build audience segments using retail, fiscal, telecom, contextual, behavioral, and first-party data, depending on campaign goals and data availability.

Verification and Control

Independent verification helps monitor campaign delivery, viewability, traffic quality, brand safety, fraud risks, and transparency across programmatic advertising campaigns.

How Programmatic Advertising Works in Russia

Programmatic advertising is useful when standard digital platforms are not enough for the campaign goal. In Russia, brands usually use programmatic ads for two main reasons: to scale media buying beyond core platforms or to reach a more specific audience with additional data signals, local inventory, and verification tools.

How Programmatic Ads Works

 

 

When Standard Digital Platforms Are Already Optimized

 

Programmatic advertising becomes relevant when a brand has already tested and optimized Yandex, VK, Telegram, or other core digital platforms, but still needs more reach. At this stage, programmatic ads help expand the campaign through additional websites, apps, video inventory, CTV, DOOH, mobile placements, and local programmatic platforms.

 

 

Image
pie charts

Key Reasons to Use Programmatic at this Stage:

  • expand reach beyond Yandex, VK, Telegram, and other core channels;
  • access extra inventory across websites, apps, video, CTV, and DOOH;
  • reduce dependence on one advertising ecosystem;
  • test new audiences and support awareness, launches, or market entry.
Audience Targeting in Programmatic Ads Desktop

 

 

When the Brand Needs a Highly Specific Audience

 

Programmatic advertising is useful when a brand needs to reach an audience that cannot be described only by age, gender, city, or broad interests. In Russia, programmatic campaigns can use purchase behavior, retail categories, fiscal data, telecom data, location signals, website behavior, CRM audiences, or look-alike segments.

 

 

Image
pie charts

Key Reasons to Use Programmatic for Specific Audience Reach:

  • reach category buyers and users with relevant purchase signals;
  • build segments by behavior, interests, location, or product category;
  • use retail, fiscal, telecom, contextual, behavioral, or first-party data;
  • reach niche audiences for premium, B2B, medical, financial, or specific products.

How Programmatic Ad Buying Works in Russia

 

Programmatic ad buying connects a brand’s campaign with users who match selected audience criteria. DSPs, data platforms, ad exchanges, SSPs, and publishers decide in real time whether to buy and display the ad. The process is automated, but audience logic, inventory, formats, budget rules, and verification still require agency control.

 

 

Image
Programmatic Ads Ecosystem Flow Desktop

Programmatic Ad Formats

Display Banners for Reach and Traffic Campaigns

Display banners are a common programmatic ad format in Russia. They help brands reach audiences across websites, mobile pages, apps, publisher networks, and available display inventory.

Online Video for Programmatic Campaigns in Russia

Online video uses video creative in programmatic campaigns across websites, apps, video platforms, and digital video inventory. It is useful when a brand needs visual storytelling, product explanation, or stronger message recall.

Mobile In-App Ads for Mobile-First Audiences

Mobile in-app ads help brands reach users inside mobile applications. This format is relevant for mobile-first campaigns, app-related products, gaming, e-commerce, and location-based communication in Russia.

CTV and Smart TV Advertising for High-Impact Video Reach

CTV and Smart TV advertising delivers video ads in connected television environments. In Russia, this format can help brands combine large-screen visibility with digital media planning.

Digital Audio for Screen-Free Audience Reach

Digital audio advertising helps brands reach users while they listen to music, podcasts, online radio, or other audio content. It can add frequency when screen-based formats are not enough.

Native Ads for Contextual and Consideration Campaigns

Native ads fit naturally into the content environment of a website, app, or publisher platform. They are useful for softer communication, contextual relevance, traffic generation, and product education.

Rich Media for Interactive Brand Engagement

Rich media formats add interactivity, animation, expanded creative space, or gamified mechanics to programmatic campaigns. They help brands create more engaging experiences than standard static banners.

Best For

  • Broad reach and ongoing visibility
  • Website traffic
  • Retargeting campaigns
  • Product reminders
  • Tactical offers and promotions

Typical Placements

  • Websites and mobile web pages
  • Publisher networks
  • Mobile apps
  • Static and animated banners
  • Responsive and standard IAB display formats

Why It Works

  • Easy to scale across digital inventory
  • Flexible for different campaign stages
  • Supports awareness and retargeting
  • Can use audience data and frequency control
  • Provides measurable campaign reporting

Best For

  • Brand awareness
  • Product launches
  • Storytelling
  • Audience education
  • Product demonstrations and consideration campaigns

Typical Placements

  • In-stream video
  • Out-stream video
  • In-feed video
  • Mobile and vertical video
  • Short-form video and selected CTV inventory

Why It Works

  • Combines motion, sound, and visual messaging
  • Reaches users in video environments
  • Delivers more context than static ads
  • Can be planned with audience targeting
  • Supports measurable delivery and reporting

Best For

  • Mobile reach
  • App installs
  • Gaming campaigns
  • E-commerce promotion
  • Local offers and location-based targeting

Typical Placements

  • In-app banners
  • Interstitial ads
  • Rewarded video
  • Native in-app units
  • Playable and full-screen mobile formats

Why It Works

  • Reaches users in focused mobile environments
  • Supports interactive and full-screen formats
  • Can use mobile behavior and location signals
  • Works with app category targeting
  • Supports awareness and performance goals

Best For

  • Brand awareness
  • Product launches
  • Premium positioning
  • Video storytelling
  • Household-level communication and TV-like impact

Typical Placements

  • Connected TV applications
  • Smart TV environments
  • Streaming video services
  • Available CTV inventory
  • Large-screen video placements

Why It Works

  • Creates a high-impact viewing experience
  • Reaches users in high-attention environments
  • Combines TV-like visibility with digital planning
  • Supports audience segmentation
  • Allows frequency control and campaign measurement

Best For

  • Brand awareness
  • Regional campaigns
  • Frequency building
  • Audio branding
  • Seasonal promotions and daily routine-based reach

Typical Placements

  • Streaming audio services
  • Online radio
  • Podcast environments
  • Music platforms
  • Mobile audio apps and companion banners

Why It Works

  • Reaches users when visual formats are not active
  • Fits commuting, work, study, and daily routines
  • Strengthens multimedia campaign frequency
  • Complements video, display, and mobile ads
  • Supports memorable audio branding

Best For

  • Content promotion
  • Product education
  • Lead nurturing
  • Traffic generation
  • Consideration and trust-building campaigns

Typical Placements

  • Recommended articles
  • In-feed placements
  • Content widgets
  • Sponsored cards
  • Promoted content on publisher platforms

Why It Works

  • Feels less intrusive than standard display formats
  • Matches the surrounding content environment
  • Attracts users open to reading or comparing
  • Supports editorial-style communication
  • Helps explain complex products or services

Best For

  • Brand engagement
  • Product demonstrations
  • Interactive storytelling
  • Gamified campaigns
  • Product exploration and high-impact launches

Typical Placements

  • Expandable banners
  • Interactive banners
  • Swipeable units
  • Playable ads
  • Gamified banners and rewarded video

Why It Works

  • Encourages users to interact with the ad
  • Increases engagement time
  • Helps improve brand recall
  • Communicates product features clearly
  • Creates a stronger experience than static formats

Need Help Choosing The Right Programmatic Setup in Russia?

Send Request

Programmatic Ads vs Standard Digital Advertising Platforms in Russia

Standard digital platforms are usually the first layer of online advertising in Russia, including Yandex Direct, VK Ads, Telegram Ads, and other platform-based tools. Programmatic advertising adds a more flexible media buying layer, helping brands use audience data, additional inventory, cross-platform placements, and independent verification for more controlled campaign delivery.

 

Comparison Point
Programmatic Ads
Standard Digital Advertising Platforms
Buying Logic
Builds a campaign across data sources, DSPs, publishers, apps, video inventory, CTV, DOOH, and other available placements.
Works mainly inside one ecosystem, such as Yandex Direct, VK Ads, or Telegram Ads.
Audience targeting
Uses audience data, purchase signals, location, behavior, contextual data, first-party data, and custom segments.
Uses targeting options available inside the platform interface, usually based on platform data and advertiser settings.
Flexibility
Suitable for complex media plans, niche audiences, cross-platform reach, and custom data-driven strategies.
Easier to launch and manage, but less flexible for advanced audience planning or external data use.
Formats
Display, video, CTV, audio, native, DOOH, in-app
Mostly display and video on one platform
Verification and Control
Can include third-party verification, viewability checks, anti-fraud tools, brand safety control, and transparent reporting.
Usually relies on the platform’s own reporting and internal campaign measurement tools.
Best Use Cases
Best for scaling beyond core platforms, reaching specific audiences, and building broader programmatic media campaigns in Russia.
Best for core digital campaigns, basic performance activity, retargeting, and ongoing promotion inside major platforms.

Audience Targeting Opportunities with Programmatic Ads

Category Buyers

Programmatic ads can reach users who are likely to buy specific product categories based on retail, fiscal, marketplace, or behavioral data. This approach is useful for FMCG, beauty, healthcare-related products, electronics, pet care, household goods, and other consumer sectors.

High-Intent Audiences

High-intent audiences include users who recently showed interest in a product, service, or category through search behavior, content consumption, website visits, or purchase-related signals. These segments help brands reach people who are closer to a decision.

Geo-Based Audiences

Geo-based targeting helps brands reach users in specific cities, regions, districts, or selected locations. Depending on the setup, campaigns may use location signals, mobility patterns, or telecom data to support local reach, store traffic, or regional media planning.

Look-Alike Audiences

Look-alike audiences help brands find new users who are similar to existing customers, website visitors, app users, or CRM segments. This is a practical way to scale campaign reach while keeping audience relevance.

Lifestyle Segments

Lifestyle segments group users by interests, habits, content behavior, and consumer preferences, such as travel, family, beauty, automotive, finance, premium products, or gaming. These segments help align campaign targeting with broader audience profiles.

Retargeting Audiences

Retargeting audiences include users who already interacted with a brand through a website, mobile app, product page, ad campaign, or customer database. Programmatic retargeting helps re-engage these users with more relevant messages and supports consideration, conversion, or repeat purchase goals.

We Work Closely with Brands and Companies While Also Collaborating with Advertising and Media Agencies

Free Whitepapers on Programmatic & Digital Media in Russia

Russian Digital Market Overview

Russian Digital Market Overview

The Current Ecosystem of the Russian Digital Market, including statistics, popular marketing tools, and media channels

How Does the Media-Buying Market in Russia Work?

How Does the Media-Buying Market in Russia Work?

Market report for global media buyers engaging with the Russian media landscape

Russian-speaking Influencer Marketing Overview

Russian-speaking Influencer Marketing Overview

Further insights and trends in the Influencer Industry within Russia & the CIS

Frequently Asked Questions About Programmatic Advertising in Russia

Programmatic advertising is the automated, data-driven purchase of ad impressions through real-time bidding (RTB). Instead of buying a slot on a single website, the system evaluates each individual user, their behavior, interests, device, and location, then decides in milliseconds whether to show them an ad and at what price. In short, programmatic is about who sees the ad, not just where it appears.

Programmatic accounts for roughly 83.2% of all digital ad spend in Russia (RMAA, 2024). It is no longer a niche channel. For most large advertisers in the country, it is the default way to buy digital inventory.

After Google and Meta* left the market, the ecosystem became dominated by local platforms: Yandex (Direct, AdFox), VK Ads, MTS Ads, Hybrid, Soloway, Between Exchange, and several specialized DSPs. Verification and analytics are handled by Weborama, AdRiver, Admon.ai, myTracker, and MediaHills.

* Meta is recognized as an extremist organization in the Russian Federation.

Five trends stand out:
  • Media inflation of 8–15% per year, driven by reduced inventory supply and growing demand.
  • A shift toward closed ecosystems. Yandex, VK, and MTS are each building self-contained stacks.
  • Mobile video and Smart TV/CTV growing roughly 20% year over year.
  • Local AdTech vendors filling the gap left by Western platforms.
  • Generative AI moving into creative production, predictive targeting, and bid optimization.

When a user opens a page or app, an anonymized impression request is sent to an Ad Exchange. Within 100–200 milliseconds, DSPs from competing advertisers evaluate how well that user fits their audience and submit bids. The highest bid wins, the creative is rendered, and the user sees the ad before the page even finishes loading. RTB is the engine behind most programmatic buying.

The most common are Weborama, AdRiver, Admon.ai, myTracker, and MediaHills. They track viewability, traffic quality, brand safety, fraud, and overall campaign delivery. Running programmatic without independent verification is the single biggest source of wasted budget.

Two requirements matter most. First, every ad creative must be labeled with an ERID token under the Federal Law on Advertising. Non-compliance leads to fines from Roskomnadzor. Second, any audience targeting that relies on personal data must comply with Federal Law No. 152-FZ, which requires explicit user consent for data processing.

Contextual advertising is the entry level. You set keywords and basic demographics inside a single platform, and that platform serves the ads. Programmatic sits one layer above: cross-channel targeting through DMPs, RTB auctions across tens of thousands of sites, plus access to mobile apps, Smart TV, audio, DOOH, and partner networks. Put differently, contextual buys placements, while programmatic buys audiences.

Usually no. Programmatic only pays off when you already have meaningful audience data, a working analytics setup, and a monthly budget large enough for the algorithms to learn. Smaller advertisers are almost always better off starting with Yandex Direct and VK Ads, then layering in programmatic once those channels hit their reach ceiling.

When you have already exhausted Yandex Direct, VK Ads, and Telegram Ads but still need more qualified reach. At that point programmatic opens up additional inventory: premium publishers, mobile apps, in-stream video, Smart TV, DOOH, and audio. None of that is available through the basic platforms.

Yes, and that is one of its core strengths. If your audience cannot be described by age, gender, and city alone, programmatic lets you combine purchase behavior, fiscal data from retailers, telecom signals, geolocation, on-site behavior, CRM uploads, and look-alike models. The more granular the segment, the bigger the advantage over contextual.

The market keeps growing. The exit of Google Ads and Meta* Ads pushed budgets into local AdTech, and those platforms have responded with custom integrations and better transparency. Even with annual price inflation of 8–15% and stricter labeling rules, programmatic still captures about 83% of digital budgets, which makes it structurally hard to replace.

* Meta is recognized as an extremist organization in the Russian Federation.

A working monthly budget usually starts around 300,000 to 500,000 RUB. Below that, the algorithms do not get enough impressions to optimize, and fixed costs (verification, creative production) eat a disproportionate share. For small advertisers, anything under 100,000 to 200,000 RUB per month is almost always better spent on Yandex Direct or VK Ads.

It works well, especially when paired with retargeting, dynamic banners, and behavioral data such as category buyers or abandoned-cart segments. The largest Russian e-commerce players (Ozon, Wildberries, Yandex Market) rely on programmatic for both acquisition and retention, which is the clearest signal of its performance utility.

Skip it if any of the following is true. You are just starting paid digital. Your monthly budget is below 100,000 to 200,000 RUB. Your basic analytics (Yandex Metrica, goals, e-commerce tracking) is not configured. Or your target audience is so broad that Yandex Direct alone can reach it. In those cases programmatic adds complexity without adding return.

Four issues come up repeatedly:
  • Wasted spend on fraudulent traffic when no verification is in place.
  • Wrong conclusions due to broken attribution and missing end-to-end analytics.
  • Overpaying for inventory because of poorly configured bid strategies.
  • ERID labeling mistakes that lead to fines from Roskomnadzor.
For a first launch, working with an agency that has hands-on experience with Russian DSPs is far cheaper than learning these on your own budget.

The clearest signals are these. Yandex Direct and VK Ads have plateaued but your audience is not saturated. You need data sources you cannot access through them, such as telecom, fiscal, or retail. You want true cross-device reach, for example the same user on mobile, desktop, and Smart TV. Or you need non-standard formats: out-stream video, rich media, audio, DOOH.

Because it depends on infrastructure most beginners do not yet have. Large data volumes are needed for the algorithms to learn. Integration with DMP, CRM, and call tracking is required for honest attribution. Fraud protection and a budget big enough to keep optimization stable round out the list. Strip any of these away and programmatic stops being effective.

Use it, but do not outsource your strategy to it. Modern DSPs (Yandex Direct, MTS Ads, Hybrid) offer generative creative, auto-targeting, predictive analytics, and real-time bid optimization. Yandex’s “Master Campaigns,” for instance, often outperform manual dynamic campaigns. But the AI still needs clean goals, validated data, and a human-authored creative strategy. Otherwise it optimizes confidently in the wrong direction.

A user opens a site or app. The publisher’s SSP sends an impression request, with anonymized user data, to an Ad Exchange. Connected DSPs evaluate the user, decide how much that impression is worth, and submit bids. The highest bidder wins, the ad renders, and impression data flows back into the analytics layer. All of this happens in a fraction of a second.

ComponentRole
DSP (Demand-Side Platform)The advertiser’s interface: campaigns, targeting, bids
DMP (Data Management Platform)Stores and segments audience data
Ad ExchangeHosts the real-time auction
SSP (Supply-Side Platform)The publisher’s side: sells inventory
Verification toolsWeborama, AdRiver, and others check impression quality

Between 100 and 200 milliseconds, from the moment the page starts loading to the moment the ad appears. The user perceives no delay.

There are three main models:
  • Open RTB. Open auction, broad inventory, lower price, higher fraud risk.
  • Private Marketplace (PMP). Invitation-only auction with selected premium publishers.
  • Programmatic Direct. Guaranteed inventory at a fixed price and volume.
In Russia, a common pattern is to combine PMP for premium reach with Open RTB for scale.

Viewability is the share of impressions that actually appear on-screen long enough to be seen. The IAB standard is at least 50% of the ad’s pixels in view for one second (display) or two seconds (video). Tools like Weborama, AdRiver, and Adloox integrate with the DSP to report this. If viewability drops below industry benchmarks, swap placements rather than chasing CPM.

Use three layers, not one:
  • Analytics layer. Yandex Metrica flags anomalies like 100% bounce rate and zero session time.
  • DSP-level filters. Hybrid, MTS Ads, and others have built-in anti-fraud.
  • Third-party verification. Weborama, AdRiver, myTracker run independent checks.
On top of that, maintain a whitelist of trusted publishers and block suspicious IP ranges.

Services such as Calltouch, CoMagic, and Callibri generate dynamic phone numbers per traffic source. When a programmatic visitor lands on the site, they see a unique number. When they call, the system ties that call back to the DSP, campaign, and creative, then passes it into the CRM. This is what lets you measure offline conversions and real revenue from impressions.

End-to-end analytics stitches together every touchpoint (impressions, clicks, calls, chats, site sessions, CRM leads, payments) into a single picture of the customer journey. Programmatic in particular needs this, because it usually contributes assists rather than last-click conversions. Without end-to-end attribution you will systematically undervalue programmatic and over-invest in lower-funnel channels.

Track them in layers. Operational metrics: CTR, CPC, CPM, impressions, clicks. Quality metrics: viewability, share of valid traffic, completion rate for video. Business metrics: ROMI, CPA, Search Lift, Visit Lift (both available in Yandex). The mistake to avoid is judging programmatic on CTR alone. It is almost always the wrong KPI for upper-funnel formats.

More often than contextual. A practical cadence: in the first 2 to 3 days, watch reach and frequency; by the end of week 1, review CTR, viewability, and fraud; in week 2, look at conversions and ROMI by segment; from there, tune bids and creatives every 3 to 5 days. DSP machine learning handles the routine. Humans should focus on strategic moves like killing weak placements and refreshing creatives.

Display banners are the workhorse format, best for broad reach, retargeting, tactical offers, and product reminders. They run across websites, mobile web, publisher networks, and apps, in standard IAB sizes (static or animated). Strengths: easy to scale, flexible across funnel stages, simple to measure, and they support frequency capping and audience targeting natively.

In-app advertising targets users inside mobile applications, which is where most attention now sits. Common formats are banners, interstitials, rewarded video (the user opts in to watch in exchange for an in-app reward), native units, and playable ads. It is especially strong for app installs, gaming, e-commerce, hyperlocal offers, and any campaign that benefits from precise geo-targeting and mobile behavioral data.

Digital audio works for brand awareness, building frequency, and reaching audiences during routines other formats cannot cover: commuting, exercising, cooking. Inventory includes music streaming services, online radio, podcasts, and mobile audio apps, usually with companion display banners. It is particularly useful for regional and seasonal campaigns.

Native ads blend into the content experience of the publisher: recommended articles, in-feed cards, content widgets, sponsored posts. They are designed for traffic generation, lead nurturing, and trust-building rather than hard direct response. Native typically delivers higher engagement than display, but lower CTR than retargeting banners, which is exactly the point.

Rich media is interactive advertising: expandable banners, swipeable units, gamified creatives, playable ads, rewarded video. Use it when you need genuine engagement, such as product exploration, brand games, or demos. Production costs are higher, but engagement and dwell time can be several times higher than static banners.

A quick mapping:
GoalRecommended formats
Mass awarenessOnline video (in-stream, pre-roll), CTV/Smart TV, display
Storytelling, educationOnline video (mid-roll), native ads
Frequency, reminderDigital audio, display, out-stream video
Mobile audiences and appsMobile in-app, rewarded video, vertical video
Premium positioningCTV/Smart TV, online cinemas (in-stream)
RetargetingDisplay banners, out-stream video, native

The standard toolkit covers six layers: category buyers (based on retail and fiscal data), high-intent audiences (search and product-view behavior), geo-targeting (down to the district level), look-alikes (modeled on CRM or site visitors), lifestyle segments (interests, habits, media consumption), and retargeting (users who already interacted with the brand). Most campaigns combine several of these rather than relying on a single one.

Category buyers are segments of users who have a high probability of buying within a specific product category, for example “baby food buyers” or “premium car enthusiasts.” The data usually comes from retail chains, fiscal data operators, and marketplaces like Ozon and Wildberries. These segments work particularly well for FMCG, beauty, electronics, and health categories where purchase intent is hard to infer from web behavior alone.

These are users who recently demonstrated active interest: searched for product information, read reviews, compared prices, or added items to a cart without completing the purchase. Programmatic lets you follow those users across other sites and apps, which is what makes it more powerful than single-platform retargeting.

Four main sources, with different accuracy levels:
  • Mobile device coordinates, accurate to a city district.
  • Anonymized telecom data from MTS, Beeline, and MegaFon.
  • Wi-Fi and GPS signals from mobile apps.
  • IP addresses, the broadest and least precise option, but useful as a fallback.
Combined, they support drive-to-store campaigns, service-area targeting, and hyperlocal promotions.

Start with a “seed,” for example 10,000 buyers from your CRM, 30 days of site visitors, or app users. Upload that seed into a DMP or DSP (MTS Ads, Hybrid, Weborama), and the platform models users with similar behavior, demographics, and interests. Done right, look-alikes let you scale beyond your first-party audience without a steep drop in quality.

Lifestyle segments group users by interests and habits: “frequent travelers,” “car enthusiasts,” “financially active,” “sports fans,” “gamers.” In Russia they come mainly from telecom operators (MTS Ads, Beeline AdTech), DMPs (Weborama, Soloway), or content-consumption data from VK and Yandex Audience.

Drop a pixel on your site or app and define the audience: visitors, cart abandoners, viewers of specific product pages. The DSP then serves reminder ads to those users across third-party inventory. In Russia this is usually done via Yandex Audience, the VK Ads pixel, or external DMPs feeding into the DSP.

Yes. You upload a hashed list of emails or phone numbers to a DMP or DSP. The platform matches them against its user profiles and uses the resulting audience for upsell, cross-sell, churn prevention, or for excluding existing customers from acquisition campaigns. The legal requirement is strict: user consent for personal data processing under FZ-152.

  • First-party data: your own CRM, site, app, and pixels.
  • Telecom data: MTS, Beeline, all anonymized, covering geo, demographics, interests.
  • Retail data: partnerships with Ozon, Wildberries, and similar.
  • Fiscal data: receipt aggregators.
  • Behavioral and contextual data from Yandex and VK.
  • Data brokers: Weborama, Segmento, Soloway.
All of it must be anonymized and collected with proper user consent.

You cannot target minors directly. You cannot target by medical diagnoses without special permissions. You cannot target by race, ethnicity, or religion. Any audience built on personal data requires user consent, typically captured through a cookie banner or service offer. Violations fall under FZ-152 and the Federal Law on Advertising, and they lead to fines from Roskomnadzor.