Summary

  • KEH eCommerce is best understood as the main Russian operating company behind Avito, not as a generic e-commerce shell. Registry, financial, procurement, network and media records all point to a company whose economics are tied to Avito's classifieds, advertising, delivery, professional seller and emerging financial-service surfaces.
  • The hard evidence is unusually strong for a private Russian platform. Public records show the Moscow legal entity, 2023 revenue above 100 billion rubles, profit above 39 billion rubles, a dividend payment, Avito-linked trademark and procurement activity, a dedicated electronic trading platform, and an autonomous-system footprint registered to KEH eCommerce or Avito.
  • The technology footprint does support profitable transactions, but it does not prove full end-to-end control. The entity appears to control a deep marketplace, advertising and internal platform stack; it still depends on banks, logistics partners, telecom upstreams, app distribution, regulators and user trust to finish the transaction loop.
  • The strongest economic case is marketplace density. Avito's official advertising page describes more than 72 million monthly visitors, mobile-heavy usage and roughly 10 transactions per second. That demand pool gives KEH pricing power over sellers, advertisers and service providers when discovery, messaging, trust and payment remain on platform.
  • Fulfilment is the weak edge of the model. Avito Delivery lowers distance and trust costs for users, but the ExMail dispute shows that logistics partners can become concentration and quality-risk points rather than pure variable-cost capacity.
  • The judgment should change if the company discloses weaker 2024 or 2025 margins, loses app or payment availability, sees material traffic compression, suffers regulatory limits on fees or financial products, or proves that new financial, AI and advertising products raise contribution without alienating sellers.

Start with a completed order. A buyer in one Russian city finds a used smartphone or a refrigerator part listed by a seller in another city. The buyer clicks into a listing, reads the seller's reputation signals, compares alternatives, chooses delivery, pays through the offered route, waits for the parcel, inspects it, and either accepts the item or returns it. The question for KEH eCommerce is not whether that looks like commerce. It clearly does.

The question is which resources this legal entity actually controls at each step, and whether those resources lower transaction cost enough to produce defensible profit rather than merely preserve a large classified-ad infrastructure.

On the evidence available, the answer is cautiously positive. KEH eCommerce appears to sit behind one of Russia's most valuable internet properties. It is not simply a brand licensor, a parked company, or a reseller of third-party demand. Public registries identify the company by its Russian registration number and Moscow address. Financial reporting cited by Interfax, RBC Companies and other Russian business databases ties the operating revenue of Avito to this entity. Network records identify AS201012 as KEH eCommerce or Avito.

Procurement sites show Avito buying power distribution, security software, storage, employee services, educational services, event logistics and office technology through the KEH eCommerce buyer perimeter. Avito's own advertising and delivery pages disclose a platform designed to monetise intent, behaviour, logistics and trust, not just banner traffic.

That is the base case. But the base case has limits. A completed order does not belong entirely to KEH. The legal entity can organise discovery, reputation, messaging, promotion, delivery choice, payment timing and data feedback. It does not manufacture most goods, own every fulfilment node, control every card rail, replace every telecom upstream, or fully eliminate off-platform bargaining. Avito's model is profitable precisely because it does not carry the inventory burden of a first-party retailer, but that asset-light structure also means the platform must keep sellers, buyers and partners inside a mediated relationship.

If too many high-value sellers see Avito as only a lead generator, they will pay for attention and move the closing conversation elsewhere. If buyers see delivery or refunds as unreliable, the platform loses the right to take a larger share of each transaction. If regulators treat embedded finance, advertising targeting or platform fees as systemically sensitive, the economics narrow.

The identity boundary matters because "Avito" is the reader-facing marketplace, while "KEH eCommerce" is the entity that appears in legal, financial, tax, routing, procurement and payment-adjacent records. Interfax described KEH eCommerce as Avito's main Russian legal entity when covering the 2022 sale from Prosus and Naspers to Kismet Capital Group. The same line appears in later financial reporting: the revenue and profit figures being discussed are those of the main legal entity of the Avito service.

This matters for economic analysis because the attractive public story - a ubiquitous marketplace where people buy, sell, hire, rent, advertise and compare - only becomes investable or governable when it is attached to the entity that receives revenue, bears costs, contracts with suppliers and can distribute cash.

The 2023 numbers are the clearest signal. Interfax reported that KEH eCommerce's revenue under Russian accounting standards reached 100.9 billion rubles in 2023, up 56.7 percent from 2022, while net profit rose to 39.3 billion rubles from 11.9 billion rubles. RBC Companies and Checko carry the same broad financial picture, including cost of sales, gross profit and balance-sheet scale. AK&M added useful colour from the reporting: promotion services and listing services were important revenue categories, while site maintenance and advertising were among the largest expense lines.

Those figures do not provide a modern unit-economics table by vertical, order, advertisement, seller cohort or delivery method. They do show that, at the company level, the platform can convert attention and transaction infrastructure into very high accounting profit.

The simplest calculation is informative. A company with about 100.9 billion rubles of revenue and about 39.3 billion rubles of net profit is running close to a 39 percent net margin on the reported 2023 numbers. A reported gross profit of roughly 59.1 billion rubles on the same revenue implies a gross margin near 59 percent. Those are not retailer margins. They are platform margins, and they fit the observed model: charge for visibility, placement, seller tools, advertising formats, and a growing set of transaction services while leaving inventory capital mostly outside the company.

The economics are much less like Ozon or Wildberries carrying fulfilment complexity at marketplace scale and more like a classified marketplace moving gradually toward managed transaction rails.

This is why a single completed order is such a useful test. If the order starts and finishes inside the Avito environment, KEH can monetise more than one layer. It may receive value from the original listing or professional seller package; from paid promotion or advertising; from behaviour and conversion data; from delivery commission or service fees where applicable; from future targeting; and eventually from embedded financial-service origination through Avito Finance or related structures. If the order only uses Avito to discover a phone number, the monetisation is much thinner.

The gap between those two outcomes is the whole economic debate.

The evidence points to a company deliberately narrowing that gap. Avito's advertising platform does not sell only generic reach. It sells intent, audience scenarios, behavioural signals and formats inside search results, product cards, recommendations and the homepage. It advertises more than 72 million monthly visitors, a mobile-heavy audience, a minimum click price from one ruble, and case studies built around leads, applications and return on marketing investment. The public claim is effectively that Avito knows when a user is moving, hiring, buying a car, planning a repair, starting a business or preparing a major purchase.

That is a stronger monetisation surface than a horizontal classified board whose only product is a search box.

The platform's official delivery page makes the same strategic move from a different angle. It tells buyers and sellers that they can transact across cities, use delivery partners, keep money protected until pickup, inspect the product at receipt and return it if necessary. Those promises are not ornamental. They are the way a classified site turns distance into addressable gross merchandise value. Without delivery, the best transactions are local, cash-like and easy to leak off platform.

With delivery, Avito can participate in a national second-hand and small-business market where trust, payment timing and return rights are part of the product.

The caveat is that fulfilment is where the asset-light model becomes exposed. Avito can orchestrate delivery, but partner quality can still define the customer experience. The ExMail dispute is the clean warning. Retail and legal press reports said the logistics operator sued KEH eCommerce after Avito ended a delivery partnership. Avito's side reportedly argued that the cooperation ended after client complaints about damaged and lost parcels were higher than average; ExMail disputed the quality characterisation and alleged harm to its business and franchisees.

Later reporting said the Moscow arbitration court rejected a very large claim by ExMail. The final legal result favoured Avito, but the strategic lesson is less comfortable: a platform that wants to own the transaction experience cannot treat logistics partners as invisible plumbing. Partner concentration, franchisee expectations, claims handling and parcel-loss rates become direct contributors to whether sellers accept commissions and whether buyers trust remote purchases.

The infrastructure evidence is more positive. AS201012 is repeatedly identified by public routing databases as KEH eCommerce or Avito. BGP.tools shows the network as active, RIPE-allocated, with 16 originated IPv4 prefixes and no IPv6 prefixes, operating in Russia and using upstreams such as High Load Lab/Qrator, RETN, VimpelCom and Telecom-Birzha. Hurricane Electric's BGP page similarly shows 16 IPv4 prefixes, no IPv6 prefixes, valid RPKI status, observed peers and 2,816 originated IPv4 addresses.

IPinfo and IP2Location give slightly different address counts and classifications, which is normal for third-party network intelligence sources using different methods, but both connect AS201012 to KEH eCommerce, Russia and Avito. The RIPE record includes the organisation entity and abuse or routing contact references.

That network evidence does not prove that every Avito transaction touches a KEH-operated server. It does prove that the entity has more than a purely outsourced web presence. A platform that originates its own routes, maintains a public autonomous system and appears in RIPE records has a measurable operating surface. For a marketplace with tens of millions of users, that matters. Search latency, listing freshness, messaging reliability, image delivery, anti-fraud checks, delivery-status flows and advertising auctions are all infrastructure questions before they become commercial questions.

If the platform is slow, unreachable or easily abused, gross contribution leaks into support, refunds, fraud, seller churn and paid-acquisition replacement.

The software evidence points in the same direction. Avito's engineering posts on Habr describe a long migration path from monolith to microservices and Kubernetes, a data bus built on Kafka, multi-data-centre replication, Kubernetes node auto-healing, testing-as-a-service with quality gates, DBaaS work, large server fleets, Puppet-managed infrastructure, and streaming data systems built with Flink and Kubernetes operators. These are not audited management accounts, and they are written by engineering teams partly for recruiting and reputation. Still, they are directly relevant.

They show a company trying to industrialise deployment, data movement, service reliability and resource allocation. In a marketplace, those internal systems are not back-office luxuries. They determine how cheaply the company can launch pricing tests, seller tools, recommendation logic, fraud controls, advertising products and delivery integrations without turning every change into an outage risk.

The Puppet post is especially useful because it describes scale in operational rather than promotional terms: more than 15,000 servers and about 10 million requests per minute handled by HTTP balancers at a moment in time. Those numbers should be treated as engineering-blog claims rather than securities filings, but they align with the observed audience scale. The Kafka and multi-data-centre posts show the risk management logic behind such scale: if all services depend on one data bus and one data centre, a local failure can become a national marketplace interruption.

The node-healing and DBaaS posts show that Avito is trying to automate repetitive reliability work, which matters because human operations do not scale linearly with traffic.

The cost implication is ambivalent. A large in-house infrastructure stack can lower marginal transaction cost if it spreads fixed engineering expense across a very large user base. It can also become a permanent capital and talent burden. Procurement sources show real buying for Avito: power distribution equipment, storage-related items, security products such as PT Sandbox renewals, Mac Mini hardware, educational services and event-asset logistics.

X-Com's case material says it created new workplaces for KEH eCommerce in 2017, while TAdviser describes a 1C ERP/Holding Management implementation intended to give KEH eCommerce a unified system interface. Those are mundane facts, but mundane facts are often where the economic truth sits. A platform this large is not just a website. It is offices, servers, cybersecurity, enterprise resource planning, language training, employee tools, advertising procurement, supplier portals and governance routines.

That expense base is justified only if Avito's demand density remains unusually high. The official advertising page claims more than 72 million unique monthly visitors and roughly 10 transactions every second. The Avito B2B-Center portal describes the company as having five main business directions - Goods, Auto, Services, Jobs and Real Estate - with monthly audience above 60 million, yearly audience of 100 million users, more than 150 million listings and more than 6,000 employees.

Similarweb's public page for Avito shows continued engagement indicators such as long visit duration and low bounce rate, while its competitor page names Youla, Farpost, Ozon and Wildberries among comparable or competing destinations. ADPASS reported Avito's active listings had exceeded 150 million in April 2023. A later insurance-industry item, citing Avito's public positioning, described more than 230 million active listings in 2024 and a 72 million monthly audience. The exact audience and listing numbers shift by source, method and date. The directional fact is stable: Avito has national marketplace density.

Density is the moat, but not an absolute moat. For a buyer of second-hand goods, substitutes include Youla, regional boards, social networks, Telegram channels, direct store websites, Ozon, Wildberries and local offline channels. For auto, substitutes include Auto.ru, Drom, dealer software and direct dealer traffic. For real estate, Cian and other vertical portals compete. For jobs, HeadHunter, SuperJob, VK and direct employer pages compete. For services, local search and social recommendations compete.

For a professional seller, the economically relevant question is not "Is Avito large?" It is "Does Avito deliver incremental qualified demand at a cost below the seller's gross margin after commission, promotion, fulfilment, returns and labour?" KEH can charge aggressively only while enough sellers answer yes.

The observed 2023 margin suggests many did answer yes. But margin quality depends on where growth came from. If growth came from better conversion, more paid seller tools and higher transaction completion, it is durable. If it came from fee increases, mandatory monetisation of previously free features, or a temporary post-2022 traffic shift after foreign platforms and advertising alternatives withdrew, it is less durable. The available public record does not fully separate those drivers.

AK&M's note that promotion services and listing services were large revenue components supports the idea that monetisation is still heavily tied to attention and placement, not only completed transactions. That is economically rational, but it means the completed-order thesis remains partly unproven.

Payment and embedded finance are the next test. The Bank of Russia included Avito Finance in the register of financial platform operators in March 2025. Interfax and ComNews reported that Avito Finance was created in 2024 as a 100 percent subsidiary of KEH eCommerce and that Avito described its model as natively embedding financial services in the platform while not planning to become a bank. This is a material signal. A marketplace that knows user intent around cars, real estate, business equipment, jobs and major purchases has attractive financial leads.

Credit, instalments, insurance, deposit products or partner offers can raise revenue per user and make the transaction more complete.

But finance also increases the regulatory and reputational load. If Avito remains a lead-generation or application layer for licensed financial institutions, it can capture economics without carrying full bank balance-sheet risk. If users start to perceive financial products as confusing, too aggressive, or insufficiently separated from ordinary classified listings, regulators and consumers may push back. The Central Bank's own public communications about marketplace sales of financial instruments distinguish between merely posting an advertisement and actually executing a financial transaction.

That distinction is the right lens for Avito Finance. The economic upside is meaningful, but so is the need for clean control boundaries.

Ownership adds another layer. Naspers and Prosus sold Avito to Kismet Capital Group for 151 billion rubles in 2022, according to Naspers, Interfax and other transaction coverage. Interfax later reported that Kismet had closed the acquisition. In 2025, Forbes and Finam reported that a structure under the management of Rosselkhozbank acquired 50 percent of Avito, with Avito saying the deal closed in April 2025 and that the parties were considering a possible public-market exit in the future. The transaction terms were not disclosed.

The state-bank connection does not automatically impair operations, and financing involvement was part of the earlier 2022 deal context. It does, however, shift the risk profile for international partners, sanctions-sensitive suppliers, app stores, investors and any future listing venue.

The geopolitical risk is therefore not abstract. Avito is a Russian domestic platform with Russian users, Russian telecom dependencies, Russian regulatory exposure and a post-2022 ownership structure. That domestic focus may be a strength: the company is less dependent on cross-border consumer demand than an exporter would be, and the retreat of foreign alternatives can strengthen local incumbents. It can also narrow optionality. Hardware, cloud, security tools, developer ecosystems, mobile distribution and payment rails are all affected by the external environment.

Procurement records showing security software, storage hardware and office equipment purchases are ordinary, but they also remind the reader that a large platform must keep refreshing technical capacity in a constrained supplier universe.

There is also a governance question inside the legal structure. Saby and other contractor databases identify Avito Management as a management company connected with KEH eCommerce. Public sources also show trademarks and multiple affiliated entities. None of that is inherently concerning; it is normal for a large platform group. But for economic analysis, it means one should not assume every Avito-related asset, cost, employee or subsidiary sits directly inside KEH eCommerce.

KEH appears to be the main operating and revenue company, yet group economics can still move through management companies, subsidiaries, intercompany charges and platform-specific vehicles such as Avito Finance. The public financial statements are useful; they are not a full consolidated investor pack.

The most attractive part of the model remains the absence of inventory risk. Traditional retail margins are squeezed by purchasing goods, forecasting demand, warehousing, markdowns, shrink, returns and last-mile operations. Avito can capture value from intent without owning the item. Even Avito Delivery is framed as delivery through partners, with safe payment timing and inspection rights rather than a fully owned logistics network. In a country as geographically wide as Russia, that is a powerful design choice.

The platform can expand the practical market radius for used goods and small merchants without making a capital commitment to every warehouse lane. The economic cost is that delivery quality becomes partner-governed, and the platform's brand absorbs a user-facing part of the failure.

Seller economics are similarly mixed. The official advertising and third-party seller guides show a sophisticated market around Avito promotion, tariffs, pay-per-view or pay-per-response models, advertising formats and analytics. That supports KEH's monetisation. It also raises a question: when sellers must pay for visibility in a crowded marketplace, is Avito lowering customer-acquisition cost or simply taxing access to demand it controls? The answer depends on seller category. In low-ticket used goods, even small fees can push users off platform.

In auto, real estate, jobs and professional services, a qualified lead can justify a meaningful spend. Avito's best economics likely sit in high-intent verticals where the value of a converted lead or completed transaction is large enough that sellers tolerate platform take rates and advertising spend.

Customer concentration is not disclosed in a way that allows a precise answer. The revenue base is probably diversified across millions of users and many seller categories, which is good. But supplier and partner concentration can still matter. The ExMail case shows logistics dependence. Telecom and anti-DDoS dependencies appear in the upstream list. Advertising clients may cluster in financial services, real estate, auto, retail, telecom or education depending on the economic cycle. The official advertising cases name brands and verticals, but they do not disclose revenue concentration.

A downturn in cars, property or small-business formation could weaken high-value lead categories even if total visitor traffic remains healthy.

The platform's unofficial market signals are important but must be treated as signals, not proof. Consumer complaint pages and forum posts frequently mention KEH eCommerce because receipts, refunds or delivery-related disputes identify the legal entity. Some users describe confusion around refunds or commissions. Seller discussions complain about delivery fees or changing commission structures. These posts do not prove systemic misconduct; complaint channels overrepresent negative experiences.

They do prove that the legal entity is visible to ordinary users at the moment money moves, and that Avito's attempt to capture more of the transaction exposes it to higher expectations. A classified board can disappoint users with spam. A transaction platform disappoints them with money, parcels and refunds.

The strongest negative interpretation is that KEH eCommerce is harvesting a dominant attention market. Under that view, Avito's technology footprint is necessary but not a proprietary transaction advantage: users come because everyone else is there; sellers pay because traffic is concentrated; the company adds paid promotion, delivery and finance where it can; and high margins reflect market power more than operational excellence. That reading is plausible, especially because financial disclosure is not granular enough to show contribution by product.

The stronger positive interpretation is that KEH has transformed a horizontal classified service into a managed transaction and advertising infrastructure at national scale. Under that view, the technology footprint is not decorative. The AS footprint, microservice estate, data buses, testing platforms, procurement perimeter, delivery partner orchestration, advertising audience segmentation and finance subsidiary all serve one goal: reduce search, trust, payment, fulfilment and marketing friction so that more transactions can begin and end in the Avito environment.

The reported 2023 margin then becomes not just a fee story, but evidence that fixed platform investments are spreading across enough high-intent activity to generate operating leverage.

My judgment is closer to the positive interpretation, with two reservations. First, KEH has proved platform profitability at company level, not disclosed order-level resilience. A 39 percent net margin in 2023 is impressive, but it does not show how margins behave if delivery mix rises, seller acquisition becomes more expensive, regulation limits advertising targeting, or users resist higher fees. Second, KEH's control is broad but incomplete. It controls an important marketplace layer and real infrastructure; it does not control the whole commerce stack.

Banks, logistics partners, telecom providers, app stores, search alternatives, government policy and user trust remain outside its direct command.

The facts that would change the view are specific. The bullish view would strengthen if later financial statements show revenue growth with stable or improving margins; if Avito Finance produces fee income without regulatory friction; if delivery complaint rates fall while transaction penetration rises; if seller tools increase retention rather than just monetise existing demand; if the company adds IPv6 and stronger routing diversity; and if a future IPO prospectus discloses diversified revenue across verticals rather than dependence on a few paid-placement products.

The bearish view would strengthen if 2024 or 2025 accounts show margin compression from advertising, site maintenance, logistics or finance compliance costs; if traffic falls materially against Youla, Ozon, Wildberries or vertical specialists; if sellers shift to off-platform closing to avoid commissions; if regulators scrutinise platform fees, financial-product distribution or advertising data; if payment or app-store access becomes unstable; or if major logistics partners demand risk premia after disputes.

It would also change if the group structure moves profits out of KEH eCommerce in ways that make the entity less representative of Avito's real economics.

There is a further test that is easy to miss because the company is private: the quality of gross contribution by use case. Avito's total numbers can look excellent while individual product lines have very different economics. A promoted listing in a crowded service category may have almost no fulfilment cost and only ordinary moderation, ranking and support expense. A delivered consumer good can require payment timing, parcel tracking, partner reconciliation, claims handling and customer support. A real-estate lead may have high advertiser willingness to pay but lower transaction visibility if the deal moves into offline negotiation.

A financial product can produce attractive referral or platform income but needs identity, consent, disclosure, regulator-facing controls and partner-bank coordination. The right question is therefore not whether Avito is profitable in aggregate. It is which extra layer of control actually increases profit after all the extra operational burden.

That distinction helps explain why advertising remains strategically important. Advertising may be less satisfying than a fully closed commerce loop, but it is often the cleanest margin. Avito can sell attention around real intent without taking responsibility for the entire underlying transaction. The official ad platform's language about life events and behavioural signals shows why. A user looking for a car, an apartment, a telecom connection, a tutor, a renovation specialist or business equipment is not casual entertainment traffic. It is commercial intent with context.

If Avito can price that intent to banks, retailers, developers, telecom operators, education companies and local service providers, it can build a high-margin advertising business on top of marketplace activity. The risk is saturation. When the platform monetises too many surfaces, users see clutter and sellers conclude that organic reach has been deliberately weakened. The art is to monetise intent without making the marketplace feel paywalled.

The same logic applies to professional seller tools. A small merchant does not buy an Avito package because it likes software. It buys a package if the software reduces the cost of finding demand, answering leads, renewing listings, measuring conversion and presenting inventory credibly. Avito's procurement and engineering evidence suggests the company has built the machinery to offer those tools at scale. The economic burden is that professional sellers become more analytical over time. They know cost per view, cost per lead, response quality, return rate and conversion.

If Avito's pricing rises faster than realised gross profit, these sellers can shift spend to Ozon, Wildberries, VK, Telegram, search, regional classifieds or direct repeat customers. A consumer brand can tolerate a brand campaign with imperfect attribution; a used-car dealer or local service provider will eventually ask whether the next ruble spent on Avito produces a profitable call.

This is where Avito's horizontal shape is both strength and weakness. A vertical marketplace can optimise one transaction type very deeply. Auto.ru can organise around car data, dealer stock, financing, insurance and vehicle history. Cian can organise around real-estate search, broker workflows, maps and mortgage leads. HeadHunter can organise around resumes, vacancies and recruiter tools. Avito has all of these categories and more, which gives it extraordinary cross-category demand but makes product governance harder. The same account may be a buyer, seller, employer, tenant, freelancer and advertiser in different months.

That creates rich behavioural data, but it also requires category-specific trust rules. Fraud in used electronics is not the same as fraud in apartment rentals. A bad job listing harms users differently from a delayed parcel. A misleading car history is a different economic harm from a noisy service lead. KEH's technology stack must let Avito enforce category-specific rules without fragmenting the user experience.

The no-inventory model also has a hidden strategic cost: the platform cannot fully guarantee supply quality. It can verify users, rank sellers, hold payment, moderate listings, suppress scams, encourage reviews and remove bad actors. It cannot inspect every item, certify every service provider or force every seller to behave like a retailer. That is why trust tools are not ancillary. Phone masking, identity checks, ratings, safe payment, delivery inspection and refund workflows are revenue protection. Every major trust failure increases the share of users who treat Avito as a place to browse but not a place to pay.

The more KEH monetises completed transactions, the more it has to spend on the unglamorous controls that prevent the transaction surface from becoming a complaint engine.

The routing and infrastructure evidence gives the company an advantage here because trust tools are data- and latency-intensive. Fraud detection works better when behaviour, messaging, payment state, listing history, device signals and delivery events can be joined quickly. Search relevance works better when the platform can re-rank inventory under heavy load without breaking core flows. Advertising yield works better when impressions, clicks, conversions and audience segments are measured reliably. Delivery claims work better when the platform can reconstruct the chain of events.

A public AS and a large engineering platform do not prove superiority, but they make it more plausible that Avito can run these feedback loops inside a controlled environment rather than stitching together a fragile outsourced stack.

Still, the absence of IPv6 in the route-origin sources is a minor but real monitoring point. It is not a fatal flaw for a Russia-focused consumer platform in 2026, because IPv4 remains widely usable and many large services operate with complex dual-stack or translation arrangements. But the public route view showing no IPv6 originated by AS201012 suggests either that IPv6 is handled elsewhere, not prioritised, or not visible through this autonomous-system footprint. For a company arguing that its infrastructure lowers transaction cost, modern network posture matters.

A future shift toward clearer IPv6 coverage would be a small but useful sign of infrastructure modernisation. A deterioration in route validity, upstream diversity or abuse responsiveness would matter more.

Regulation is another cost centre disguised as a strategic option. Avito's finance move is economically logical because the platform sees moments when users need credit, insurance, instalments, deposits or other financial products. But once finance is embedded, the company has to manage disclosure, suitability boundaries, data-sharing permissions, identity flows and complaints that are not just e-commerce complaints. The same is true for employment listings, real estate, cars and advertising. A dominant horizontal marketplace becomes a quasi-infrastructure venue for many regulated economic activities.

That gives KEH bargaining power, but it also makes the company legible to regulators in more categories. The better the platform becomes at closing transactions, the more likely it is to be judged not only as a publisher of listings but as a market operator with duties.

The capital-market story is also double-edged. A possible public-market exit can discipline the company toward clearer reporting, more formal governance and a cleaner product narrative. It can also create pressure to demonstrate growth in monetisation before all trust and supplier systems are mature. If future owners or lenders value Avito on advertising yield, take-rate expansion and financial-service optionality, management has an incentive to push fee surfaces. If users and sellers experience that as extraction, the brand equity accumulated over many years can erode quietly before it appears in traffic statistics.

A marketplace does not usually fail all at once. It first becomes less trusted for expensive items, then less efficient for professional sellers, then more dependent on paid acquisition and defensive promotions.

The practical benchmark is contribution after avoided alternatives. For a household seller, Avito must be easier and safer than a Telegram chat, a social network post or a local offline sale. For a professional merchant, it must be cheaper than acquiring the same customer through search ads, marketplace storefronts or repeat CRM. For a bank or advertiser, it must deliver intent that cannot be bought with the same precision elsewhere. For a logistics partner, it must offer volume without making one client a survival risk.

For telecom and security suppliers, it must be a stable enterprise customer rather than a high-pressure buyer that transfers too much operational risk. The public record shows KEH has the scale to make all those counterparties care. It does not yet show that all of them receive enough surplus to keep the ecosystem balanced.

This is why the ExMail episode deserves weight even after Avito's reported court win. The issue is not whether ExMail's damages theory prevailed. The issue is what happens when Avito's transaction layer becomes large enough that a partner builds capacity around it. If a partner underperforms, Avito must protect customers and the brand. If Avito changes routing, terms or volume expectations abruptly, partners can experience existential risk. Either way, transaction infrastructure becomes relational capital, not just procurement.

A platform that wants more completed orders needs enough supplier discipline to protect users and enough supplier fairness to keep capable partners willing to invest. That balance is hard and will not be visible in top-line revenue until something breaks.

The best current judgment, then, is conditional confidence. KEH eCommerce has already shown that Avito's technology and demand footprint can produce substantial profit. The company is not merely preserving infrastructure. It is using infrastructure to price attention, reduce search costs, broaden delivery radius, sell seller tools, improve reliability and prepare financial-service layers. But the next ruble of profit is less certain than the last one. The early classifieds model captured attention cheaply.

The next model captures transaction value, and transaction value comes with parcel loss, refund disputes, fraud controls, partner conflicts, financial regulation and seller pushback. If KEH can absorb those costs while keeping margins high, Avito becomes a stronger commerce infrastructure company. If not, the technology footprint remains large but starts to look like a defensive cost of maintaining a dominant listing venue.

For now, KEH eCommerce should not be written off as infrastructure without commerce economics. The evidence shows a profitable operating company attached to a deep marketplace, a measurable network surface, a serious internal technology stack and a growing transaction perimeter. Nor should it be treated as a fully integrated marketplace with unassailable control.

The honest view is narrower and more useful: KEH has built or assembled enough technology and trust infrastructure to make Avito's demand pool monetisable at high margin, but the next stage of value depends on converting more completed orders, financial leads and delivery flows without turning user trust into the hidden cost of growth.

Sources