A US small business owner with forty employees decides to hire an offshore bookkeeper. Over three weeks, they collect four quotes. A marketplace platform quotes $65 an hour. A venture-backed managed-talent company quotes $9,000 a month with a twelve-month minimum. A mid-market offshore agency quotes $22 an hour with a one-time recruitment fee. A directly-registered India operator quotes $7 an hour, month to month, no fee.
The candidates behind all four quotes live within a few hundred kilometres of each other. Several of them have overlapping employment histories at the same Indian firms. In two of the four cases, the actual person who would be deployed could plausibly be the same individual.
This is the part of the offshore-staffing market that nobody explains to small businesses: the price spread between vendors is not primarily a quality spread. It is a spread in how much apparatus sits between the buyer and the worker, and who is paying for that apparatus. A buyer who does not know how to decompose a quoted rate will read $65 as "premium" and $7 as "risky," when the more accurate reading is often that $65 is enterprise-priced and $7 is directly-priced, and the small business in question is not an enterprise.
I have built and governed offshore-staffing operations in India since 2009 — twelve years on the board of Virtual Employee Private Limited, and the last several running LegelpTech Outsourcing. I should state my interest plainly before going further: LegelpTech operates Zedtreeo, which sits at the low end of the range I am about to describe, publishing rates of $5 to $10 an hour. I am not a neutral party to this question. What I can offer instead of neutrality is the cost structure from the inside — what an hour of Indian delivery actually costs to produce, what each layer above it adds, and which of those layers a small business is genuinely buying versus merely funding.
This essay is a companion to my earlier piece on the hidden costs of offshore engagements, which argued that the headline rate understates true cost. This one argues the other half of the same point: that above a certain rate, the headline is also overstating what the buyer receives. Both distortions are real. They point in opposite directions, and a competent buyer has to correct for both.
The rate stack: what a US buyer's dollar actually buys
Every offshore staffing rate, at every vendor, decomposes into the same categories. The vendors differ in how large each layer is, not in whether the layer exists.
The delivery floor. This is what it costs to put a competent, full-time Indian professional at a desk for an hour: gross salary, statutory employer contributions (provident fund, ESI where applicable, gratuity accrual), equipment, connectivity, workspace, and the HR and payroll function that keeps them employed. For an experienced bookkeeper, marketer, or mid-level developer in Delhi NCR, Pune, or Hyderabad in 2026, this floor lands somewhere in the region of $4 to $8 an hour depending on seniority and city. It is not zero and it is not $2. Any vendor pricing below this floor is either taking it out of the worker's compensation, out of statutory compliance, or out of their own solvency.
Recruitment and replacement. Sourcing, screening, and the cost of the candidates who do not make it. A vendor screening seriously — Zedtreeo's published figure is roughly one in twelve applicants clearing the process — carries the cost of the other eleven. This is a real cost that has to sit somewhere. Vendors either amortise it into the hourly rate, charge it separately as a placement fee, or under-invest in it and pass the cost to the buyer as turnover.
Account management and governance. The named human who owns the engagement cadence, handles escalations, and absorbs coordination overhead the buyer would otherwise carry themselves. I have argued elsewhere that this layer is worth paying for; it is the single largest determinant of whether a buyer's internal manager reclaims their week.
Compliance and certification. An ISO 27001:2022 ISMS is not free. Certification, surveillance audits, an internal audit function, policy maintenance, background-screening infrastructure, access control, and the evidence trail procurement teams ask for. For a single-brand operator this is a meaningful line item. For an operating company running several brands on one certified ISMS, the per-engagement cost falls sharply — the audit does not get more expensive because a second brand uses it.
Distribution. Sales salaries and commissions, marketing spend, paid acquisition, conferences, SDR teams, the Western-country office that makes a buyer comfortable. This is the layer that varies most violently between vendors, and it is the layer buyers almost never see itemised.
Entity and currency overhead. A vendor with a US or UK legal entity carries local incorporation, accounting, employment, insurance, and legal cost, plus FX handling. A single India-registered entity contracting directly in USD carries a fraction of it.
Return expectation. What the vendor's owners require. A founder-operated firm can run on a normal operating margin indefinitely. A venture-funded platform is priced to service a growth expectation and an eventual exit, and that expectation is embedded in the rate card whether or not it appears on the invoice.
The first four categories are things a buyer is genuinely purchasing. The last three are things a buyer is funding. The pricing gap in this market is almost entirely the last three.
Four pricing models, four structural floors
The quotes a US small business collects are not points on a continuum. They come from four distinct business models, each with a different structural floor it cannot price below.
Western marketplaces and premium networks
Toptal is the reference point here, publicly quoted in 2026 in the range of $60 to $200 an hour, with most comparisons landing senior engineers between $80 and $180. Turing sits lower, commonly cited between $30 and $60. Upwork spans $15 to well over $100 depending on who you find and how well you screen.
These platforms sell curation and speed of access to a global pool. That is a real product. Toptal's screening genuinely is stringent, and for a buyer who needs a senior specialist for a bounded project and has no capacity to run their own vetting, the premium can be defensible. But the price includes a large Western sales and marketing organisation, a Western entity, and a platform margin the buyer cannot see — the blended rate is deliberately opaque about the split between what the worker receives and what the platform keeps.
The structural floor here is high because the distribution cost is high. It cannot come down without dismantling the go-to-market motion that makes the platform work.
Venture-funded managed platforms
Andela is the clearest example: managed engagements commonly quoted at $8,000 to $14,000 a month per engineer, historically with twelve-month minimums and dedicated-team framing.
This model sells managed capacity at enterprise scale — a talent cloud, delivery management, and the contractual posture a large company's procurement function expects. For a company hiring twenty engineers and needing a single accountable vendor, it works.
Its structural floor is set by two things at once: a Western cost base, and an investor return expectation that requires growth in revenue per account. The twelve-month minimum is not incidental to the pricing — it is what makes the customer acquisition cost recoverable. A vendor that spends heavily to win an account must lock the account long enough to earn the spend back. The minimum commitment is the acquisition cost, converted into a contract term.
Mid-market offshore agencies
The broad middle. India-based agencies quoting US buyers somewhere between $15 and $35 an hour for the same roles — India-based technical VAs commonly at $8 to $18, agency-supplied offshore VAs at $8 to $15, junior-to-mid developers frequently at $25 to $35, senior architects at $50 to $60.
These are usually genuine India delivery operations with a Western-facing sales layer bolted on: a US or UK front office, a sales team compensated on commission, and often a one-time recruitment or placement fee on top of the rate. The delivery floor underneath is the same $4 to $8. The difference between that floor and the quoted $22 is distribution, front-office entity cost, and margin.
Nothing here is illegitimate. The buyer is paying for a sales relationship and a Western point of contact, and some buyers value that enough to justify the multiple. The question is whether this particular buyer does.
Directly-registered India operators
Vendors contracting from an Indian entity straight to the client, without a Western sales organisation in between. Rates in the $5 to $10 range, which is where Zedtreeo publishes: $5 to $6 for administrative and virtual-assistant work, $6 to $8 for bookkeeping, marketing, design, and HR, $8 to $10 for developers, DevOps, cybersecurity, and legal professionals — roughly $800 to $1,600 a month for a full-time seat.
The structural floor here is the delivery floor plus recruitment, account management, compliance, and a normal operating margin. There is no Western entity, no commissioned sales organisation, and no investor return to service. That is the entire explanation for the gap. It is not a discount, and it is not a subsidy. It is the absence of three cost layers.
Which is the distinction this whole essay turns on, so it is worth naming precisely. Cheap and affordable are not the same thing. A cheap price is one where something has been removed from underneath it — the worker's pay, the statutory contributions, the compliance function, the account manager, the vendor's own solvency. An affordable price is one where the delivery underneath is fully funded and the cost of reaching the buyer is what has been removed. The two look identical on a quote. They behave nothing alike in month nine.
The mispricing is specific to small businesses
Here is the part that matters for positioning, and it is a narrower claim than "the lowest rate wins."
The enterprise-priced models are not overpriced in general. They are priced correctly for the customer they were designed to serve. A 4,000-person company buying offshore engineering has a procurement function that will issue a 90-page security questionnaire, a legal team that will negotiate a bespoke MSA over six weeks, a vendor-management office that requires quarterly business reviews, and an expectation of dedicated customer-success coverage. Serving that customer costs real money, and the rate card reflects it.
A forty-person company has none of that. It has an owner, possibly an operations manager, and a need for one or two competent people starting soon.
When that forty-person company buys from an enterprise-priced vendor, it pays for the enterprise apparatus and cannot consume it. It funds a customer-success organisation it will interact with quarterly. It funds an MSA negotiation capability it will never invoke, because it will sign the standard paper. It funds a sales cycle that, for a two-seat engagement, cost the vendor more to run than the engagement will return in its first six months — which is precisely why the twelve-month minimum exists.
This is the actual mispricing. Not that premium vendors overcharge, but that their cost structure is calibrated to a buyer profile the small business does not match, and the small business absorbs the mismatch.
The four commercial terms where this bites hardest, none of which are the hourly rate:
Minimum commitments. A twelve-month lock on a two-seat engagement transfers the vendor's acquisition risk onto a buyer far less able to carry it. A forty-person company's requirements can change in a quarter. If the fit is wrong in month two, the enterprise-priced buyer pays for ten more months of it.
Recruitment and placement fees. A separate one-time charge, often several thousand dollars, for the sourcing work. On a large multi-year engagement it amortises to nothing. On a single seat held for eight months it can be the largest single line item — and it is charged before the buyer knows whether the hire works.
Seat minimums and team framing. Vendors organised around dedicated pods and squads price and staff for three-to-five-person units. A business that needs one bookkeeper is either refused or sold a unit it does not need.
Rate opacity. A blended rate that does not disclose the split between worker compensation and vendor margin removes the buyer's ability to sanity-check anything — including whether the person doing their work is paid enough to stay.
Each of these is defensible against an enterprise buyer. Against a small business, each is a transfer of vendor risk onto the party least equipped to absorb it.
Telling an affordable rate from a cheap one
I am aware of how the argument above reads coming from someone whose company prices at the accessible end of the range. So let me make the harder claim, which is the one that actually matters: an affordable rate is only good news if you can identify what structurally produces it. A low rate with no structural explanation is not affordable — it is cheap, and the difference will surface later.
There are exactly four honest sources of an affordable offshore rate.
One: the absence of Western distribution cost. No US entity, no commissioned sales force, no paid-acquisition engine. This is the largest single source and it is entirely legitimate — the buyer simply is not paying for a sales apparatus.
Two: amortised compliance. When one certified operating entity supports several brands, ISO 27001:2022 certification, surveillance audits, the internal audit function, and background-screening infrastructure are paid once and spread across all of them. Zedtreeo's certification sits with LegelpTech Outsourcing, and the audit cost does not increase because a brand shares the ISMS.
Three: recruitment as product rather than cost centre. A vendor running continuous pipelines against recurring role types has already absorbed the sourcing cost when the enquiry arrives — which is what makes a 48-hour shortlist possible, and what makes charging a separate placement fee unnecessary. A vendor recruiting from scratch per request has to charge for it.
Four: retention as the margin. On month-to-month terms with no lock-in, the vendor's entire economics depend on the engagement continuing. There is no contractual mechanism to hold a dissatisfied client, so the only route to profitability is a hire the client keeps. This inverts the incentive that minimum commitments create.
And then there are the sources that make a rate merely cheap. They produce the same number on the quote and a completely different outcome:
Underpaying the worker below the local market. Produces turnover, and turnover is the most expensive thing that can happen to an offshore engagement. The buyer's saving is erased by repeated ramp-up.
Skipping statutory employer obligations. Provident fund, gratuity, proper employment contracts. This is a compliance exposure that eventually reaches the buyer through their own vendor-risk review.
Under-investing in security and compliance while claiming otherwise. A certificate that is not on an IAF-recognised registry, or a scope statement that does not cover staff augmentation, is the tell.
Loss-leading on price to acquire, then repricing. A rate that exists to win the logo and rises at renewal.
The diagnostic question is therefore never "is this rate low?" — a low rate on its own tells you nothing about which of the two you are looking at. It is: "what in your cost structure produces this rate, and what happens to it in month thirteen?" A vendor with a structural answer will give you the layer-by-layer explanation without hesitation, because the explanation is their actual business model. A vendor without one will change the subject to candidate quality.
The corollary is worth stating: ask what the deployed person is paid. Not the exact figure — vendors will reasonably decline — but whether their compensation is at, above, or below the local market for their role, and what the vendor's retention rate is. A vendor whose price is structural can answer both comfortably. A vendor whose price comes out of the worker's pay will not want to.
When the direct model is the wrong choice
The direct model is not correct for every buyer, and pretending otherwise would undermine the rest of this essay.
If you need on-site presence, or a vendor whose staff can attend your office, none of this applies.
If you are hiring a genuinely scarce senior specialist — a distributed-systems architect with a specific domain history, a security engineer with particular certifications — the deep global search that premium networks run is worth its premium. The direct model is priced for competent professionals in well-supplied categories, and it is honest to say that the top one percent of a niche is not usually found there.
If your own procurement function requires a same-country contracting entity, a US-signable MSA, and a domestic legal remedy, that requirement is a real constraint and it costs what it costs.
If you have no capacity whatsoever to manage remote staff — no defined work, no review cadence, no one who will answer questions — no price fixes that. My earlier essay on hidden costs covers this at length: the governance overhead is the buyer's to carry regardless of rate, and a lower rate does not reduce it.
If you need to scale to thirty people in a quarter, the enterprise-priced managed platforms are built for exactly that and a smaller operator may not be.
What is left after those exclusions is, in my experience, the substantial majority of what US small and mid-sized businesses actually need offshore: bookkeeping and finance operations, marketing and content execution, design, HR and recruitment support, virtual assistance, healthcare revenue-cycle work, drafting, and mid-level software and DevOps. Well-supplied categories, competent professionals, one to five seats, requirements that will change within the year. For that buyer, at that scale, paying an enterprise rate card is paying for an apparatus they will never touch.
Six questions that read any quoted rate
Carry these into every vendor conversation, at any price point. They work equally well against a $65 quote and a $7 one.
- "What does the person doing my work actually receive, relative to their local market — and what is your twelve-month retention rate?" Tests whether the price comes out of the worker.
- "Which country's entity issues my contract, and what does that entity cost you to maintain?" Surfaces the entity and distribution layers directly.
- "Is there a recruitment, placement, or onboarding fee, and is it refundable if the hire does not survive ninety days?" Non-refundable fees charged before fit is established are the term most likely to hurt a small buyer.
- "What is the minimum commitment, and what specifically does that minimum protect you against?" An honest vendor will tell you it protects their acquisition cost. That is a legitimate answer — and it tells you their acquisition cost is embedded in your rate.
- "What is your ISO 27001 scope statement, and is your certificate on an IAF-recognised registry?" Scope matters more than the logo. A certificate scoped to something other than staff augmentation does not cover your engagement.
- "What happens to this rate at renewal, and is that in writing?" Distinguishes a structural price from an acquisition price.
A vendor at any price point who answers all six cleanly is one you can transact with. The rate then becomes a straightforward commercial decision rather than a signal you are trying to decode.
The gap closes when buyers learn to read the stack
The five-to-tenfold spread in this market is not sustainable on the merits. It persists because the buyers most exposed to it — small businesses without procurement functions — have no framework for decomposing a quoted rate, and so fall back on the only heuristic available, which is that price signals quality. In a market where price mostly signals distribution cost, that heuristic transfers a great deal of money for nothing.
What I would want a US small business to take from this is not that the lowest number wins — an unexplained low number is the one thing in this market you should genuinely be afraid of. It is that the question "why is this the price?" has a real, answerable, layer-by-layer answer at every vendor in this market, and that a vendor unwilling to give it is telling you something. The premium vendors have a good answer, and for enterprise buyers it is the right one. The direct operators have a good answer, and for a forty-person company buying two seats it is usually the better one. The vendors in trouble are the ones whose answer is a brand story.
Ask for the stack. The vendors worth hiring will walk you through it.
Disclosure: Zedtreeo is a remote-staffing brand operated by LegelpTech Outsourcing Private Limited, of which the author is a Director. Published rates and terms cited above are Zedtreeo's own, current as of July 2026. Competitor rates are drawn from publicly published 2026 market comparisons and are indicative ranges, not quotes.
Ashu Mishra is Director, LegelpTech Outsourcing Private Limited (CIN U82990DL2025PTC446352). He served on the board of Virtual Employee Private Limited (CIN U74900UP2010PTC041120) from 2009 to 2021. Career began at HCL Technologies on the British Telecom account. Reachable at ashu@legelp.com.