58.com Layoff Claim Tests Technology News: Why the 75% Figure Does Not Add Up

58.com faces a viral claim that it plans layoffs reaching 75%, yet the available technology news evidence does not verify that number.

The allegation surfaced through a viral question that reached a Chinese social platform’s trending list on August 12, 2026. It asks whether 58.com has started another large workforce reduction and whether the rate could reach 75%.

No public company statement, regulatory filing, or independently documented employee count currently confirms a companywide 75% reduction. The most credible recent reports describe different figures from an earlier round.

Reporting published in June 2025 cited employees who estimated that 20% to 30% of staff faced cuts across several divisions. Some teams reportedly lost larger shares, including isolated groups where employees claimed reductions above 60%.

That distinction matters. A severe reduction inside one team does not establish a companywide layoff rate. Repeating the larger number without its denominator can turn a local account into a false corporate total.

The rumor still deserves attention because it fits a longer pattern. 58.com has repeatedly reduced staff while losing ground in recruitment, housing, and local services.

The company also went private in 2020, limiting the financial visibility that once came with its US listing. Outsiders can now see fragments of the business, but not a complete current workforce or audited group performance.

The real story is therefore larger than one disputed percentage. A former classifieds leader is trying to shrink legacy operations while protecting enough product capacity to compete with specialized digital platforms.

The Verified Layoff Story Stops Well Short of 75%

The 75% figure remains an allegation, while documented reporting supports a smaller companywide range and larger cuts inside selected teams.

The freshest widely circulated claim arrived without a detailed headcount, named division, implementation schedule, or company confirmation. Those missing elements prevent a reliable calculation.

A layoff percentage needs two numbers. Readers need the number of affected positions and the employee population used as the denominator.

Neither number appears in the trending question. It is also unclear whether “75%” describes one function, one office, a planned reduction, or accumulated staffing changes over several years.

The strongest accessible account concerns June 2025. According to employee accounts, workers estimated that 20% to 30% of staff were affected across multiple departments.

That report said technology engineering and local services were among the affected areas. It also described reductions involving research, operations, and marketing positions.

One technical employee reportedly said their department lost 40% of its staff. Some smaller groups allegedly experienced reductions above 60%.

Those figures are serious, but they are not interchangeable. Team-level cuts often vary sharply during reorganizations because companies close projects rather than reducing every unit equally.

The report also said employees received an N+1 compensation arrangement. In this formula, affected workers receive statutory compensation plus one additional month, subject to the applicable agreement.

58.com did not publish a complete workforce reconciliation alongside those reports. Its private status makes outside verification harder because regular public disclosures are no longer available.

A separate 2023 episode reportedly affected 30% to 50% of workers across several business lines. 58.com did not release a final companywide count for that round either.

That history creates fertile ground for exaggerated totals. People may combine several rounds, apply one team’s reduction to the whole company, or treat a planned target as completed action.

The 75% number also appears in unrelated parts of 58.com’s history. Shareholder approval for its privatization exceeded 75%, for example, but that figure had nothing to do with layoffs.

This does not prove how the new rumor originated. It does show why reporters should trace every percentage to its original subject and denominator.

The correct conclusion is narrow. 58.com has faced recurring, substantial workforce reductions, but a new companywide 75% cut has not been independently verified.

That conclusion can change if the company releases a statement or employees provide consistent documentation across divisions. Until then, headlines should separate the confirmed pattern from the viral number.

Why 58.com Keeps Returning to Technology News

58.com’s repeated restructuring reflects the erosion of a broad classifieds model that once benefited from concentrated web traffic.

58.com began as a digital marketplace connecting consumers with jobs, housing, used goods, vehicles, and local service providers. Its scale made it a default destination for many urban needs.

That position was valuable during the desktop and early mobile internet eras. A single platform could aggregate fragmented listings, attract consumers, and sell visibility to merchants.

The company listed in New York in 2013. It later expanded through acquisitions, including major deals involving housing and competing classifieds properties.

By 2019, 58.com operated a collection of multi-category and vertical listing platforms. Its final full-year public filing described a business connecting local companies and consumers across China.

That model depended on a reinforcing loop. More merchants created more listings, more listings attracted users, and greater traffic encouraged merchants to purchase promotional services.

The loop weakened as user behavior fragmented. Specialized apps began owning individual transactions instead of directing users through one general marketplace.

Recruitment shifted toward products designed around direct communication between candidates and employers. Housing platforms added agent networks, transaction services, and standardized property data.

Short-video and social platforms also became discovery engines. Consumers can now find cleaners, repair providers, apartments, and used products inside personalized content feeds.

These competitors do not simply offer another listing page. They control a different part of the customer journey.

A recruitment app can manage conversations and candidate matching. A housing platform can connect listings with agents, financing, and closing services.

A social platform can create demand before a user begins an explicit search. That ability challenges a classifieds service built around users arriving with a defined need.

58.com recognized the transition before the latest rumor. Management previously described a shift from traffic-based revenue toward service-based revenue.

The strategic logic was sensible. Listing visibility becomes less defensible when competing apps can generate traffic through recommendation algorithms and closed transaction systems.

Execution is harder. A broad platform must upgrade several verticals simultaneously while focused competitors concentrate investment on one customer journey.

That creates duplicate technology, sales, support, and management structures. Each business needs specialized expertise, but the group must also control costs across the portfolio.

Repeated layoffs can therefore reflect more than a weak quarter. They can signal that the company has not found a stable operating shape for its collection of mature businesses.

Cost reduction buys time, but it does not rebuild user preference. The company must still demonstrate that its services solve problems better than dedicated alternatives.

This is why the story qualifies as technology news rather than a routine employment report. It concerns the decline of one internet distribution model and the rise of several specialized replacements.

Specialized Rivals Turned Scale Into a Liability

58.com’s breadth once lowered customer acquisition costs, but focused rivals now use deeper workflows to keep users inside their products.

The pressure is clearest in recruitment. 58.com historically served a wide range of employers and workers, including many local and blue-collar roles.

BOSS Zhipin built its experience around direct exchanges between job seekers and hiring representatives. That interaction model makes the product feel closer to messaging than conventional classified advertising.

The rival’s recent results show that online hiring still supports growth. In the first quarter of 2026, BOSS Zhipin reported 60.9 million average monthly active users.

Its paid enterprise customers reached 7.1 million during the preceding 12 months. Revenue increased 7.6% from the comparable quarter, according to its first-quarter results.

Those figures do not reveal 58.com’s current performance. They do demonstrate that recruitment demand has not disappeared from every online platform.

The competitive question is where that demand goes. A service with stronger communication, matching, and employer tools can capture spending that once flowed to promoted listings.

Housing presents a similar problem. 58.com’s Anjuke business built a large property information marketplace, but Beike developed a more integrated online and offline transaction network.

Beike reported full-year 2025 gross transaction value of RMB3.18 trillion. Its housing results also showed weaker fourth-quarter existing-home activity, confirming that the market itself remained difficult.

That comparison needs care. Beike and Anjuke use different operating models, and transaction value does not equal revenue.

Still, Beike’s model illustrates the challenge facing a listing-led platform. It connects digital discovery with agents, stores, and transaction infrastructure.

A broad classifieds company must decide whether to match that operational depth or remain a lighter information layer. Either choice carries costs.

Building transaction services requires capital, compliance systems, and local execution. Remaining an information layer leaves the platform exposed to services that control the final purchase.

Local services create a third front. Consumers increasingly begin searches on large delivery, mapping, social, or short-video platforms.

These products already know a user’s location and preferences. They can rank providers, show video evidence, manage communication, and sometimes complete payment.

58.com still benefits from a recognizable brand and extensive merchant relationships. However, brand memory does not guarantee repeated product use.

The portfolio’s size can also complicate restructuring. Shared functions may serve several divisions, making a central cut appear efficient while weakening multiple products.

Conversely, keeping separate teams for every vertical can preserve expertise but increase overhead. Management must find a balance without current public-market scrutiny.

This creates the article’s main tension: specialization versus scale. 58.com accumulated categories, brands, and organizational layers when aggregation generated an advantage.

Now those layers can slow product decisions and obscure accountability. A specialist can measure one funnel, while 58.com must manage several distinct markets.

A 75% companywide reduction would make that problem harder, not easier. Such a cut would raise questions about whether the group could maintain safety, product quality, and merchant support.

That is another reason the rumor requires stronger proof. A reduction of that scale would represent a fundamental operating transformation, not ordinary expense control.

The Privatization Reversal Changed What Outsiders Can See

Going private gave 58.com more freedom to restructure, but it also removed the disclosures needed to evaluate today’s viral claims.

58.com completed its take-private transaction in September 2020. Public investors stopped receiving the same recurring operational and financial information available during its US listing.

The deal remained legally relevant years later. In May 2026, a Cayman Islands court rejected an effort by dissenting investors to secure a higher valuation.

The court concluded that the transaction price was fair, according to the buyout ruling. That decision addressed the 2020 deal, not the company’s present workforce.

Privatization can help management pursue changes without quarterly market pressure. It can also make credible reporting harder when employees post partial information.

Before privatization, investors could compare revenue, margins, business risks, and employee numbers across annual reports. Those figures established a consistent baseline.

Today, observers must assemble information from affiliated-company filings, competitor reports, employee posts, corporate announcements, and media investigations.

Each source covers a different perimeter. An affiliate’s financial statements cannot automatically represent the entire group.

Likewise, a worker may accurately describe one division but lack knowledge about every office. Anonymous reports can reveal important events without supplying a reliable companywide denominator.

This information gap helps explain why conflicting figures circulate. The public knows that reductions occurred but cannot easily measure their accumulated effect.

The gap also makes simple narratives tempting. One version says 58.com is merely removing bureaucracy after years of expansion.

Another says the platform has entered an irreversible decline. Neither conclusion is proven by a single layoff percentage.

Repeated cuts can improve near-term cash flow. They can eliminate overlapping roles, close weak products, and shorten reporting structures.

However, sustained reductions can produce hidden costs. Experienced staff leave, product maintenance slows, and remaining employees inherit systems they did not build.

Trust and safety can suffer when platforms reduce review teams. That risk matters for classifieds services, where fraudulent listings can damage both users and the brand.

Sales cuts can also reduce merchant revenue before automation replaces those relationships. Local businesses often need onboarding and support, especially when advertising products are complex.

Technology teams face their own constraints. Retiring an old system safely takes engineering work, even when management plans to consolidate products.

A company can therefore become trapped between two expensive states. It cannot afford the old organization, yet it must invest to reach the new one.

Private ownership does not remove that tradeoff. It changes who sees the numbers and how quickly outsiders learn about the consequences.

The current rumor should be read through that lens. The uncertainty is not evidence that nothing happened, but it is also not permission to publish the largest available number.

Reliable technology news must mark the boundary between documented reductions, plausible interpretation, and unsupported extrapolation.

AI Cannot Solve a Marketplace Trust Problem by Itself

Automation can reduce operating costs, but 58.com still needs users and merchants to trust the information moving through its marketplaces.

58.com has presented artificial intelligence as part of its next operating model. The company can apply language models to customer support, listing creation, matching, and internal workflows.

These uses have a clear efficiency case. AI can summarize job requirements, classify property listings, detect repeated text, and route customer questions.

It may also help merchants produce clearer advertisements. Search systems can interpret conversational queries instead of depending entirely on fixed categories.

None of those capabilities guarantees marketplace quality. A model can rewrite a misleading listing just as easily as a legitimate one.

Fraud detection requires verified identities, behavioral signals, payment information, human escalation, and enforcement. Language generation handles only part of that system.

The same limitation applies to recruitment. Better matching can reduce search time, but it cannot establish that every employer offers a legitimate position.

Housing information creates another verification challenge. Automated tools can detect anomalies, yet local market knowledge remains necessary when addresses, prices, or availability conflict.

This is where aggressive workforce reduction can clash with an AI strategy. Automation projects need clean data, process knowledge, evaluation, and ongoing oversight.

Employees often hold the operational knowledge needed to define edge cases. Removing them before encoding that knowledge can make automation less dependable.

AI adoption can still support a smaller organization. The sequencing matters more than the slogan.

A company must first identify stable processes, measure current performance, and define acceptable error rates. It then needs evidence that automation maintains those standards.

Without those steps, headcount becomes a misleading proxy for productivity. Fewer employees may reduce spending while increasing unresolved complaints and merchant churn.

The skeptical question is therefore not whether 58.com uses AI. It is whether AI improves the specific workflows where specialized competitors already offer a stronger experience.

Recruitment users care about relevant employers and quick responses. Housing users care about authentic inventory and dependable agents.

Local-service customers care about availability, quality, and accountability. None of these outcomes follows automatically from lower staffing.

The company’s broad dataset remains a potential advantage. Years of local marketplace activity can support classification, risk detection, and matching.

That advantage depends on data quality and permission. Old or duplicated information can weaken recommendations rather than improve them.

The 75% rumor risks encouraging an overly simple automation narrative. It invites readers to assume that AI has suddenly made most roles unnecessary.

There is no verified evidence supporting that conclusion. Recent documented cuts are better explained by business contraction, organizational overlap, and pressure from focused competitors.

AI belongs in the analysis as a restructuring tool, not a confirmed cause. Until 58.com publishes measurable outcomes, claims about an AI-led transformation remain management ambitions.

That standard should apply across technology news. Companies must connect automation claims to service quality, revenue, retention, and verified productivity results.

Three Signals Will Show Whether the Restructuring Works

The next stage will be decided by verified staffing data, competitive operating metrics, and evidence that automation improves marketplace outcomes.

The first signal is a direct company response to the 75% claim. 58.com should define the affected business perimeter, timeline, and workforce denominator.

A denial without figures would narrow the story but not resolve it. A detailed statement would allow reporters to distinguish a departmental closure from a groupwide program.

Consistent employee documentation could also clarify the scope. Multiple accounts across locations and functions would carry more weight than one unattributed percentage.

If evidence confirms a companywide reduction near 75%, the restructuring would represent a far deeper transformation than previously reported. It would weaken the argument that this is routine optimization.

If the number proves to describe one team, the broader conclusion changes. The company would still face repeated cuts, but the viral framing would be materially misleading.

The second signal is competitive performance in recruitment and housing. BOSS Zhipin’s user and enterprise growth provides a clear benchmark for online hiring.

Beike’s transaction network provides another benchmark, even during a weak housing cycle. Future results will show whether focused platforms keep gaining engagement and merchant spending.

58.com does not need to copy either rival completely. It does need evidence that its remaining products retain users, improve conversion, or generate stronger merchant value.

New hiring activity would offer another clue. Open roles in engineering, risk control, or AI evaluation would suggest selective reinvestment after organizational cuts.

The absence of hiring would not prove failure. However, continued broad reductions without visible product investment would strengthen the contraction interpretation.

The third signal is measurable automation performance. Watch for disclosed changes in response times, listing quality, fraud rates, customer retention, or support resolution.

Token usage and internal adoption figures do not establish business value. They measure activity rather than outcomes.

A credible AI transition would connect automation with improved service quality and lower operating costs. It would also explain how the company protects users when models make errors.

If 58.com publishes those results, its restructuring narrative becomes stronger. If it offers only broad AI language while complaints rise, the narrative weakens.

For employees and job seekers, the immediate lesson is simpler. Treat the 75% figure as unverified, but do not dismiss the underlying pattern of repeated reductions.

For technology buyers and investors, the case offers a broader warning. A large historical user base cannot protect a marketplace when specialized products own more of each transaction.

For reporters, the standard should remain strict. Percentages require denominators, team-level claims require labels, and old events should not be presented as new announcements.

The question now is not whether 58.com has experienced layoffs. That is well documented.

The question is whether its next reduction creates a focused technology company or leaves a thinner version of the same challenged platform. Watch the staffing disclosure, rival metrics, and product outcomes.