Trigger Warning: Your Gig Platform is About to Get Ghosted

This week, we talk about Protector, an innovative private security app seeking to streamline the security hiring process by replacing traditional agency methods and unclear pricing with a transparent, on-demand marketplace. This platform allows quick access to elite security professionals, but it faces a significant challenge—disintermediation. The trust that Protector builds between clients and security providers could paradoxically lead to the platform's demise. Today, we explore the economic and operational implications of this trend and pose a crucial question: Can Protector maintain its indispensable intermediary role, or will the very trust it fosters prove fatal to its business model?

Imagine stepping out for a high-profile event in Los Angeles, knowing that a former SWAT officer or Navy SEAL is watching your back.

With Protector, a new app blending ridesharing with elite private security, this is becoming a reality.

Founded by Nick Sarath, Protector promises to eliminate the headache of hiring security by replacing outdated agency websites and opaque pricing with an on-demand marketplace, giving users instant access to personal protection from top-tier professionals. Launched in Los Angeles and New York, the app allows users to customize down to the dress code (formal, business casual, or tactical gear) and vehicle options.

The popularity of the app is linked to the killing of Brian Thompson, CEO of UnitedHealthcare, and the realization that many CEOs may need security detail.

But here’s the problem—once clients find a protector they trust, why keep booking through the app?

This is the classic disintermediation dilemma: the very trust that platforms like Protector cultivate between users and providers can lead to their own obsolescence.

Today we examine the economic and operational implications of disintermediation in gig platforms and explore whether Protector can lock down its role as an essential intermediary, or whether trust will ultimately put a bullet in its business model.

Let’s surveil the situation more closely.

Disintermediation in Gig Economy Platforms

Disintermediation refers to users cutting out the intermediary platform after an initial match, completing transactions directly to avoid platform fees.

Research indicates this is a widespread risk that can undermine a platform’s revenue model. Edelman and Hu (2016) note that many marketplaces are inherently vulnerable—once a customer finds a suitable provider (e.g., a reliable cleaner or guard), there’s little incentive to return to the platform for repeat engagements.

A striking example is Homejoy (an early home-cleaning gig platform), which struggled because clients frequently hired the same cleaners offline to avoid fees; this disintermediation was cited as a factor in Homejoy’s shutdown.

Similarly, a client who will find a trustworthy guard through Protector might be inclined to hire them directly thereafter, threatening the platform’s long-term viability.

Empirical studies have quantified the impact of disintermediation: Gu and Zhu (2020) conducted a randomized experiment on a large freelance marketplace and found that enhancing trust and information (through better reputation indicators) increased the hiring of high-quality freelancers but also the rate of disintermediation.

Other research estimates the magnitude of the problem: a Harvard case study documented that roughly 90% of freelance jobs initiated on a platform were ultimately completed off-platform.

Similarly, Upwork doesn’t require clients to transact through the platform after they matched and exchanged contact information, and many clients and freelancers take their business offline. Upwork will do everything possible to make you stay, but they can’t force you.

In China’s ZBJ work platform, providers reported earning ten times more income off-platform than through it. This level of disintermediation is often found to be on the order of 5-10% of transactions in measured studies, and could potentially be much higher in unmonitored settings.

For instance, in a cargo delivery platform, once commissions were introduced, drivers began asking customers to pay offline, effectively “leaking” transactions off the app. Researchers observed that when a 15% commission was imposed, off-platform deals (leakage) roughly doubled. As fees or commissions increase, the temptation to disintermediate also grows.

Additionally, if a platform merely provides an introduction and adds little value thereafter, clients are prone to transfer repeat transactions off the platform. This challenge is especially acute for services that involve ongoing relationships or repeat needs—a common scenario in personal security services. Both sides may perceive greater benefit in a direct arrangement once a relationship is established (e.g., a security consultant may offer a discount to a client for a long-term off-platform contract).

From an operations perspective, disintermediation is notoriously hard to detect and prevent. Since off-platform dealings are unobservable to platform owners, platforms often don’t realize how much business is leaking.

Attempts to curb disintermediation include platform design tweaks and policies: information control (e.g., concealing contact details until a booking is made), contractual penalties for off-platform deals, loyalty incentives, or added services that make on-platform transactions more attractive.

Airbnb, for example, withholds direct contact information until after booking to deter guests and hosts from moving the transaction offline. Some studies find that features like instant booking and quality badges can modestly reduce disintermediation by streamlining the user experience or increasing trust in staying on-platform. However, these measures are not foolproof—high trust and strong rapport between users can eclipse platform-provided value.

In summary, most literature highlights disintermediation as a fundamental economic challenge: it directly erodes the platform’s ability to “capture value” from the connections it brokers and balancing trust-building with retaining transactional control is a delicate task for platform operators.

Off-Platform Flows of Goods, Cash, and Information

But to really grasp why disintermediation is such a big risk, we must understand a platform’s main flows.

The number one rule is that platforms cannot monetize the product or service being exchanged unless information and cash flows occur on the platform.

When users disintermediate, the three primary flows—goods (or services), cash, and information—start to occur outside the platform’s ecosystem.

In supply chain theory, these three flows are linked and essential for efficient operations. Effective systems align the flow of the product/service with the flow of payments and information to ensure smooth coordination.

By analogy, a gig platform functions as a coordinated system for these flows: it provides the infrastructure for service delivery (e.g., job listings, scheduling), handles the payment processing, and channels the information (communications, reviews, job details) between parties. If all three flows move off-platform, the result can be fragmentation and inefficiency, as well as trust issues that the platform can no longer mitigate.

Service (Goods) Flow Off-Platform: In gig work, the “goods” are often services or labor. When the actual service provision shifts off-platform (e.g., a security guard privately covering an event without using Protector), the platform loses oversight of service quality and fulfillment. The transaction becomes part of the informal economy.

Information Flow Off-Platform: Information flow includes all exchanged communication and data related to the transaction (job requirements, scheduling details), and also reputational information (ratings, profiles). When users move communications off the platform the platform loses visibility into user interactions. This creates multiple implications such as misunderstandings or errors (operational inefficiency) and misbehavior or safety issues (scams, harassment, etc.) which the platform can neither monitor nor moderate—platforms often monitor chats to detect fraud or to keep records in case of disputes. Additionally, the platform’s reputation system suffers—the platform cannot record the transaction, so the client can’t leave a review for the provider (or vice versa).

This withholding of information can degrade the overall trust environment. A robust flow of information is what “enables effective communication and coordination” in a system; without it, future matching becomes less efficient (since the platform isn’t learning to improve recommendations or verify reliability). In security services, keeping communication on-platform might be critical for safety (e.g., having a log of all interactions). If a client and a bodyguard go off-platform and an incident occurs, the lack of recorded information could make it hard to resolve what happened or assign responsibility—a liability concern that could erode trust not only between the parties but in the platform model as a whole.

Cash Flow Off-Platform: Perhaps the most immediate consequence of disintermediation is that payment happens directly (via cash or other channels) rather than through the platform’s system.

This means revenue loss for the platform (no commission), which threatens its financial sustainability, but also eliminates any secure payment escrow or fraud protections the platform provides.

Economic theory posits that intermediaries exist to reduce transaction costs and risks—one such role is ensuring the buyer’s money is safe until the service is delivered and ensuring the worker gets paid. When clients pay workers “under the table,” both parties lose the neutral guarantor that a platform can serve as. Indeed, platform design experts note that keeping payments on-platform is crucial to user trust, because a vetted payment system provides confidence in secure and timely transactions.

Reevaluating Disintermediation Risk – Evidence from Home-Cleaning Platforms

A study by Astashkina et al. provides a quantitative counterpoint to the prevailing assumption that disintermediation is an inevitable outcome in gig economy platforms. Analyzing data from a home-cleaning platform operating in a post-Soviet city, the researchers tracked the geographic movements of 5,391 cleaners across 95,633 residences over multiple years. Their dataset contained 4.3 million distance snapshots, recording cleaners’ proximity to past clients’ locations both during and after their official platform engagements.

The study’s key finding is striking: disintermediation was nearly nonexistent. Based on statistical modeling, the authors estimate that at most 1 in 83 cleaner-client relationships (1.2%) resulted in off-platform transactions—a rate far lower than expected given previous research on gig platform leakage.

Several factors contributed to this low disintermediation rate:

Low Relationship Retention: 90% of cleaner-client pairs only engaged once, 5.8% completed two cleanings, and only 0.5% reached ten or more. The low frequency of repeat interactions naturally limited opportunities for disintermediation.

Platform Stickiness Due to Security: Cleaners entered private homes, making platform-provided background checks, secure payments, and dispute resolution valuable to clients. This aligns with broader economic findings that platforms offering legal and financial protections experience lower disintermediation rates.

Lack of Cost Sensitivity: The average cleaning fee was $31.58, a price point deemed relatively insignificant for affluent users. Research suggests that disintermediation is more prevalent when commissions create substantial savings incentives—typically at rates of 15–30% of transaction value.

Behavioral Data on Cleaner Mobility: GPS-based tracking showed that, in the post-engagement period, cleaners rarely returned to past clients’ locations. The statistical distribution of cleaner movements closely mirrored that of non-working days rather than working days, indicating an absence of off-platform work.

This empirical evidence suggests that disintermediation risk is not uniform across all gig economy platforms. While prior research has documented severe leakage in freelance and delivery marketplaces, the home-cleaning case indicates that factors such as security concerns, relationship frequency, and platform-provided protections can substantially mitigate transaction migration.

For platforms like Protector, where security and trust are paramount, these findings highlight potential structural defenses against disintermediation. Security services inherently involve higher personal risk than home cleaning, so if a low-risk service like home cleaning retains transactions at a 98.8% rate, gig security marketplaces may experience even lower leakage—provided they emphasize liability coverage, real-time incident support, and institutional trust mechanisms that are difficult to replicate off-platform.

Experiential Learning and Disintermediation: A Dangerous Cocktail for Gig Platforms

Understanding why disintermediation occurs also requires examining how experiential learning—the direct interactions between users—can amplify a platform’s vulnerability.

A key challenge in gig economy platforms is how trust formation and experiential learning influence disintermediation. As buyers and service providers engage repeatedly, they accumulate valuable private knowledge about each other—knowledge that platforms struggle to capture.

The findings of Moon et al. shed new light on this phenomenon: platforms that fail to balance repeat engagements with new exploration risk reinforcing their own demise.

The study highlights that 87% of the variation in hiring decisions in online labor platforms is explained not by visible ratings but by experiential learning—buyers’ firsthand knowledge of worker quality. In other words, workers are evaluated more through direct engagement than platform reputation systems.

The question is once this trust is established, what’s stopping them from cutting a private deal?

A recurring issue in gig platforms is whether to prioritize known, trusted matches (which promotes efficiency) or push users to try new workers (which prevents disintermediation). Platforms that only encourage repeat hires may inadvertently train users to disintermediate

Moon et al. found that platforms that prioritize new match exploration over repeat hires increase buyer welfare by up to 45-47% of platform revenue. In contrast, platforms that over-prioritize repeat work suffer from under-exploration, which leads to stagnant growth and a vicious cycle of high-value workers exiting the platform.

The paradox of trust and disintermediation is a double-barrel threat: the more trust the platform helps build, the more it risks being cut out of the deal. If Protector wants to avoid becoming an “introducer” rather than an essential intermediary, it must rethink how it structures long-term relationships. Otherwise, the best talent in the gig security market may simply lock and load—then walk away from the platform for good.

Will Protector Be the Next Uber or the Next Handy?

Protector enters the market with a compelling premise: seamless access to elite security personnel through an app. But will they revolutionize private security, like Uber transformed transportation, or will they struggle with scale and sustainability, as seen with Handy in the home services sector? The answer depends on several critical factors: market fit, scalability, unit economics, and platform stickiness—with disintermediation emerging as a decisive challenge.

The Case for Success: Why Protector Could Be the Uber of Security

1. High-Trust, High-Stakes Industry

Security is a premium service where trust is paramount. Protector’s model addresses a fragmented market by eliminating opaque agency pricing and inefficient booking processes, much like Uber streamlined black car services. Given that safety is non-negotiable, clients may prefer a structured platform that guarantees vetting, insurance, and accountability, reducing the likelihood of widespread disintermediation.

2. Scarcity and Prestige as a Feature

Unlike Uber, which thrives on an abundant labor supply, Protector operates in an exclusive, high-barrier industry—its guards are ex-SWAT, SEALs, or Air Force PJs. This creates scarcity-driven demand, positioning Protector closer to a luxury brand like Soho House than a mass-market gig platform. Research on platform economics suggests that exclusivity can increase customer retention and willingness to pay, as users value continued access to a select pool of high-quality providers.

3. Strong Unit Economics at a Premium Price Point

Protector’s pricing model—charging upwards of $1,000 per night for an armored escort—places it in a high-margin, low-volume category. Unlike Handy, which struggled due to low-margin transactions and infrequent usage, Protector’s premium pricing means it doesn’t need millions of users to be viable. Research on platform sustainability indicates that luxury services with infrequent but high-value transactions tend to have stronger retention than low-cost, high-frequency platforms.

4. Network Effects and Social Proof

If influencers and high-profile clients normalize on-demand security, demand could scale rapidly. Much like Uber Black became a status symbol, being seen with a Protector escort could serve as a luxury flex, broadening the Total Addressable Market beyond celebrities to business travelers, VIP nightlife goers, and tech executives. This “social proof” effect has historically fueled growth in high-end service platforms.

The Case for Failure: Why Protector Could Struggle Like Handy

1. The Disintermediation Paradox

While trust-building is essential for user adoption, too much trust can undermine the platform itself. Gig economy research shows that when repeat engagements are high, disintermediation follows. A corporate client booking the same security professional multiple times may prefer to arrange future services off-platform to save on commissions, while a protector with repeat clients may view the platform’s fees as unnecessary overhead. If Protector does not implement strong stickiness mechanisms, it risks bleeding transactions.

2. Limited Scalability of Supply

Uber scaled by onboarding virtually anyone with a car, but elite security professionals are a finite resource. If demand spikes, Protector cannot simply onboard thousands of new protectors overnight, leading to bottlenecks, surge pricing, and dissatisfied clients. The platform must balance exclusivity with scalability—possibly by expanding into lower-tier security roles or investing in its own certification pipeline to grow the supply of vetted professionals.

3. High Customer Acquisition Costs

Luxury services require expensive marketing to niche audiences. Unlike Uber, which could attract everyday users with low-cost rides, Protector must target ultra-high-net-worth individuals, executives, and corporations, requiring specialized sales channels. If the CAC is too high relative to Lifetime Value (LTV), the platform will struggle to achieve sustainable unit economics.

4. Trust and Liability Risks

Protector’s brand relies on the reputation of its workforce. However, gig platforms struggle with liability. A single high-profile failure could destroy trust and trigger regulatory scrutiny. The platform must proactively mitigate these risks through stringent vetting, real-time incident reporting, and liability coverage to maintain credibility.

Final Prediction: Protector Has Uber-Like Upside, But Handy-Like Risks

Protector holds significant potential in the luxury security market, but its success hinges on its ability to retain transactions on-platform. According to Astashkina et al., platform stickiness can be stronger than previously assumed, but only if the service offers value beyond matchmaking.

Carving out a profitable, defensible niche (similar to Uber Black) is within reach if Protector can first mitigate disintermediation, scale supply while maintaining quality , and broaden its customer base without losing exclusivity.

In short, Protector can win—but only if it prevents its best protectors from going rogue.