What immigration paralegal outsourcing actually costs: loaded cost, market pricing and the variables that decide it

Comparing the cost of an in-house paralegal against an outsourced one looks like arithmetic and is mostly definitional. Almost every published comparison — including most produced by outsourcing providers — compares an outsourced monthly fee against an in-house salary. That comparison is not wrong so much as incomplete, and the incompleteness runs consistently in one direction.

This piece sets out what the real numbers are on both sides, which costs get left out, what the market currently charges, and the variables that end up mattering more than the rate.

The comparison almost everyone gets wrong

Here is the comparison you will usually see: an in-house paralegal costs $45,000–$62,000 a year, an outsourced one costs $1,200–$2,600 a month, therefore outsourcing is cheaper. The numbers are accurate. The comparison is still misleading in both directions.

It understates the in-house figure, because salary is not what an employee costs. And it can overstate the outsourced saving, because a monthly fee buys a defined number of hours of a defined kind of work, which may or may not be the work you actually need done.

The unit that makes the comparison honest is cost per hour of administrative capacity actually delivered. Everything below is in service of getting to that number.

What “loaded cost” actually includes

Applied to a US immigration paralegal at $45,000–$62,000 base, loaded cost lands somewhere around $56,250–$86,800 a year. That is roughly $4,688–$7,233 a month, against $3,750–$5,167 on salary alone.

Ramp time is a cost, and it recurs

A new hire is not productive on day one, and in immigration practice they are not productive on day thirty either. They have to learn your matter types, your checklists, your filing conventions, your case management setup, and your escalation habits.

Call it a conservative eight to twelve weeks to reasonable productivity for a paralegal with prior immigration experience, longer without. During that period you are paying full loaded cost for partial output, and consuming supervisory attention that would otherwise be billable.

The point people miss is that this cost recurs. It is not a one-time setup expense amortised across a decade. It is paid again at every departure. If your paralegal tenure is two years, you are paying a ramp penalty every two years, permanently.

Turnover: the cost nobody budgets for

Turnover cost is the sum of several things that are individually easy to dismiss and collectively substantial:

• The vacancy gap. Weeks or months where the work is simply not being done, or is being absorbed by attorneys.
• Recruitment, again. Advertising, screening, interview time, and the opportunity cost of everyone involved.
• Ramp, again. The full eight-to-twelve-week penalty, repeated.
• Institutional knowledge walking out. The undocumented knowledge of which client is difficult, which matter has an unusual history, and where the exceptions live.
• Disruption to matters in flight. The most expensive and least predictable component. A handover mid-matter is where deadlines get missed.

This is where the continuity question becomes a financial question rather than a service-quality one. Ask any provider what happens when your assigned person is unavailable, and whether a briefed backup exists on your account. A provider with no answer is offering you the same single point of failure as an in-house hire, without the control.

What the market charges

Dedicated offshore paralegal support generally runs $1,100–$2,200 per month for generalist work (Virtual Employee, 2026). Immigration-specialised providers price above that band — Legal Soft’s immigration tier starts at $2,922 per month (2026).

Our own pricing sits inside that range:

Engagement models, and what each is good for

Providers sell roughly four shapes. They are not interchangeable, and picking the wrong one is a more common source of disappointment than picking the wrong price.

For most small and mid-sized immigration practices the dedicated monthly model fits best, for a reason that is not about price: the work benefits enormously from someone who knows your matters. Project-based engagement re-buys context every time. That is efficient for genuinely discrete work and wasteful for the recurring administrative load a long queue generates.

The costs that never appear on an invoice

Four costs are routinely absent from comparisons and routinely decisive.

Onboarding your own side. Documenting your checklists, conventions and escalation criteria takes real time from people whose time is expensive. It is worth doing regardless — most firms discover their processes were never written down — but it is not free.

Supervision. Reviewing and approving work takes attorney time. It exists in both models, but a comparison that ignores it will understate the cost of whichever option involves more of it.

Reconciliation. If work happens outside your systems, someone has to bring it back in. This is the cost most likely to consume the savings entirely, and the one least likely to be discussed during a sales process.

Exit. What happens to your data, your documented processes and your matters in flight if the arrangement ends. A short notice period is worth less than it appears if there is no orderly handover.

Where the work happens changes the price

This deserves separating out, because it is the single largest hidden variable.

Support that operates inside your case management system works on the live matter. The record updates as work happens. Your audit trail stays intact. Access is governed by permissions you control and can revoke.

Support that operates outside it receives documents, works elsewhere, and returns output. Every cycle needs reconciling: filing the returned work, updating the record, checking nothing was superseded while it was out. That reconciliation is unbilled work performed by your staff, it scales linearly with volume, and it introduces a category of error — the stale copy — that does not otherwise exist.

Small-firm operations data consistently identifies information scattered across disconnected tools as the largest single source of friction. Buying support that adds another disconnected location is buying more of the problem you were trying to solve. Ask this question early; the answer materially changes the effective price.

Compliance overhead is a real line item

Delegating work that touches confidential client information carries obligations: disclosure and client consent where non-lawyers outside the firm receive protected information, supervision under Model Rules 5.1 and 5.3, and competence under Rule 1.1. Several states impose more.

Those obligations do not disappear if the provider ignores them. They land on you. Concretely, if a provider cannot supply client-consent language, a written supervision protocol, and a data processing agreement, then your firm is drafting all three — which means counsel time, at counsel rates, before a single matter is assigned.

Price that in. A provider $300 a month cheaper that leaves you to build the compliance scaffolding yourself is not cheaper. Our full treatment of what is required is in the compliance questions immigration firms ask before delegating.

Three worked scenarios

Illustrative constructions, not descriptions of actual firms. The point is the method rather than the numbers.

Scenario one: solo practitioner, 60 open matters

At roughly 1.2 administrative hours per matter per month, demand is about 72 hours — well under a full-time equivalent. Hiring a full-time paralegal at $56,250+ loaded would buy roughly double the capacity actually needed, and the excess is not recoverable.

Part-time support at $1,200 a month for 20 hours a week supplies about 80 hours monthly, which fits. The relevant comparison is not against a full-time hire at all — it is against continuing to absorb 72 hours a month of administrative work personally, at whatever an hour of the principal’s time is worth.

Scenario two: four-attorney practice, 200 open matters, one paralegal

Demand at 1.3 hours per matter is roughly 260 hours a month. One paralegal at 60% administrative time supplies about 96. The gap is roughly 164 hours — about one full-time equivalent.

Here the comparison is direct: a second in-house hire at $4,688–$7,233 a month loaded, against full-time outsourced support at $2,600. The second is materially cheaper, but the decision should still turn on ramp time, continuity cover, and whether the work can be performed inside your systems — not on the monthly delta alone.

Scenario three: twelve attorneys, 520 open matters, four paralegals

With a court-weighted caseload at 1.8 hours per matter, demand is about 936 hours. Four paralegals at 55% administrative time supply 352. The gap is roughly 584 hours — around 3.6 full-time equivalents.

At this scale a managed team makes sense, and the arithmetic stops being the interesting part. The binding questions become whether the provider can supply consistent quality across several people, whether continuity survives individual absence, and whether the compliance scaffolding scales. A firm this size absorbing 584 hours a month of administrative work is losing far more in attorney time than any plausible price difference between providers.

Run your own figures with the case capacity calculator.

Break-even: when does this stop making sense?

Honest answer: several situations, and a provider who claims otherwise is selling rather than advising.

• Very low volume. Below roughly 40 open matters the administrative load may not justify any dedicated arrangement. The overhead of managing the relationship can exceed what it saves.
• Highly bespoke work with no repeatable pattern. If every matter is genuinely unique, there is less delegable procedural work than the model assumes.
• No documented process at all. Delegation makes an undocumented process worse before it makes it better. Firms in this position benefit more from a month spent writing things down.
• Jurisdictional constraints you cannot satisfy. If your state’s rules or a particular client’s requirements make offshore delegation impractical, price is irrelevant.
• Systems that cannot support external access. If your case management platform cannot grant scoped external access, the reconciliation cost may swallow the benefit.

What you are actually buying

The framing that makes decisions cleaner: you are not buying a person. You are buying coverage of a specific administrative gap.

Headcount is a proxy for capacity, and a poor one. Two paralegals who spend 40% of their time on substantive work supply less administrative capacity than one who spends 90% on it. The question is not how many people you have. It is how many hours of the specific work you need actually get done, and what each of those hours costs once everything is counted.

Answering that requires knowing your gap, which is why the arithmetic comes before the shopping. Firms that skip it end up comparing monthly prices for capacity they have not sized, which is how you buy the wrong amount of the wrong thing at a good rate.

Questions worth asking any provider

1. Who is the contracting entity, and where is it registered?
2. Where are personnel located, and is that disclosed to my clients?
3. Do you provide client-consent language and written supervision terms?
4. Is there a data processing agreement, and what does it say about sub-processors and breach notification timelines?
5. Will work happen inside my case management system, or yours?
6. Is client data copied out of my systems at any point?
7. What happens when my assigned person is unavailable?
8. Is there a briefed backup on my account, and how are they briefed?
9. What is the ramp period, and is there a trial before full handoff?
10. What is the notice period, and what happens to my data and matters on exit?

Questions three, four and six are the ones that separate providers. A vendor who answers them immediately has done this before. A vendor who needs to come back to you is one whose compliance work you will end up doing yourself.

Red flags

• No named contracting entity. If you cannot identify and verify who you are contracting with, nothing else on this list matters.
• Vagueness about location. Where personnel sit is not a detail; in several states it determines your disclosure obligation.
• “Our AI handles that.” Modern case management platforms already auto-populate large numbers of forms. A provider whose differentiator is form-filling is selling you something you probably already license.
• Guarantees about privilege. Nobody can guarantee how a particular court will rule. A provider claiming certainty is overselling something they do not control.
• Reluctance on the trial period. A provider confident in quality has no reason to resist being evaluated on real work.
• Pricing that only makes sense at scale you do not have. Team pricing sold to a solo practitioner is a mismatch, not a bargain.

The four alternatives, costed

Outsourcing is one of four things a firm can do about an administrative gap. Comparing it only against hiring skips two of them, and one of those two is what most firms are actually doing by default.

The first row is the one worth sitting with. “Absorb it” appears free because it never generates an invoice, but it converts the most expensive hours in the firm into administrative work. A firm losing 150 attorney hours a month to administration is spending more than any option below it on the table — it simply never sees the number.

This is also why the software-versus-people framing is usually a false choice. They address different work. Software handles the repeatable transcription; the residual is unstructured, judgement-adjacent work that still needs a person. Most firms end up needing both, and the useful question is what remains after the software has done what it can.

Contract terms worth negotiating

Price is the most negotiated term and rarely the most valuable one. Five others are worth more attention:

1. Trial period on live work. A defined paid trial on lower-risk matters before full handoff, with an exit at the end. This converts an assessment problem into an observation problem and costs the provider nothing if they are confident.
2. Named backup, briefed in advance. Not “we have cover” but a specific person with context on your account. Continuity is a financial term disguised as a service term.
3. Right to remove an individual without cause. No explanation required, no penalty. Fit problems surface fast and should be cheap to fix.
4. Notice period and exit handover. Short notice is worth little without an orderly handover — data returned or deleted, matters in flight documented, access revoked on a defined timeline.
5. Scope change mechanism. How work is added or removed, and what happens to price. Without this, scope creeps in one direction and price only ever moves in the other.

Terms on data handling, sub-processors and breach notification belong in a data processing agreement rather than the commercial contract — see the compliance questions immigration firms ask before delegating for what that should cover.

How to tell afterwards whether it worked

Most firms never establish whether an arrangement delivered, because they never captured a baseline. Three numbers, measured before and again at ninety days, settle it.

Non-billable administrative hours by role. The headline measure. If attorney administrative hours have not fallen, the arrangement has added capacity somewhere that was not the bottleneck.

Deadline near-misses. Count items completed within a day or two of their deadline. This should fall. Missed deadlines are too rare to be a useful metric and too serious to wait for.

Median matter age. Slow to move and worth tracking anyway. Administrative capacity does not speed up adjudication, but it does stop matters stalling for reasons within the firm’s control.

A fourth, softer signal: how often clients contact the firm asking for an update. Scheduled proactive updates should reduce unscheduled inbound contact noticeably within a couple of months, and that reduction is itself a recovered cost.

Transition cost, in detail

The first sixty days of any arrangement cost more than the steady state, on both sides of the comparison. Firms that budget only the monthly fee are surprised, conclude the arrangement is not working, and sometimes abandon it before it reaches the point of paying back.

What the transition actually consumes:

• Process documentation. Writing down checklists, conventions and escalation criteria that currently live in people’s heads. Expect several days of senior time. This is genuinely valuable independent of the arrangement — most firms discover their processes were never written down and that two people were doing the same task differently — but it is real cost.
• Access provisioning. Setting up scoped permissions, which is quick if your system supports granular roles and slow if it does not.
• Close review during the trial. Checking output far more thoroughly than you will later. This is supervisory time at attorney or senior-paralegal rates, and it is the largest single transition cost.
• Rework. Early output will need correcting. Budget for it rather than treating it as evidence of failure — the point of a trial is to surface the gaps in your own instructions, which are always larger than expected.
• Compliance setup. Consent, supervision protocol and data processing agreement, reviewed by counsel. One-time, but not small.

A reasonable planning assumption is that months one and two deliver perhaps half the steady-state benefit while costing additional internal time, and that the arrangement reaches its real economics somewhere in month three. Judge it at ninety days, not thirty.

Common pricing mistakes

Four that recur:

1. Buying to the average when the load is spiky. A firm with hearing-driven peaks that resources to its mean will be comfortable most of the time and short exactly when it matters.
2. Comparing monthly prices for unsized capacity. Doing the capacity arithmetic after choosing a provider is how firms end up buying the wrong amount of the wrong thing at a good rate.
3. Counting soft costs on one side only. Supervision, onboarding and management exist in both models. Loading them entirely onto the in-house column is exactly as misleading as the salary-versus-fee comparison.
4. Ignoring the do-nothing baseline. The status quo has a cost — attorney hours diverted to administration — and it is usually the highest of all the options. It just never generates an invoice.

The economics differ by practice type

Administrative load per matter is not a constant. It varies enough by practice mix that a single per-matter figure will mislead any firm whose caseload is concentrated. Four broad patterns:

Two practical consequences follow. First, if your caseload is heavily weighted toward one of these, use a per-matter figure derived from that pattern rather than a blended average — a removal-defense-heavy practice that plans on a family-practice number will under-resource badly.

Second, the shape of the load should influence the engagement model. Steady, high-volume, procedural work suits a dedicated monthly arrangement. Spiky, deadline-driven work concentrated around hearings argues either for more capacity than the average implies, or for an arrangement that can flex. A firm buying to its average will be comfortably resourced most of the time and short at exactly the moments that matter.

Asylum practices deserve a specific note. The preparation burden is front loaded and heavy, then the matter goes dormant for a very long time, then it needs substantial refresh before the hearing. Firms consistently under-budget the third phase, because the first one is memorable and the second one is silent. Evidence gathered two years before a hearing is rarely usable without work.

What good looks like at six months

A useful way to sanity-check any arrangement is to describe the state you are buying, rather than the price you are paying. At six months, a working arrangement generally shows five characteristics.

Attorney administrative hours have fallen measurably. Not marginally, and not by impression — against the baseline you captured before starting. If this has not moved, capacity was added somewhere that was not the constraint.

Deadline near-misses have fallen. Work is completing with margin rather than at the wire.

Escalations are routine and unremarkable. Things get raised to an attorney regularly and without drama. An arrangement with no escalations is not running smoothly; it is one where someone is making judgement calls they should not be.

Continuity has been tested at least once. Somebody has been away, cover worked, and nothing slipped. Until this has happened, continuity is a claim rather than a fact.

The cost per hour of delivered capacity is calculable. You can state what you are paying for administrative capacity actually received — which means you can compare it against alternatives on an informed basis rather than a headline one.

Glossary

Frequently asked questions