VC
About
Media
Startups & Experiments
Writing
Topics
Resources
Contact
Posts
EntrepreneurshipJuly 8, 2026 / 10 min

What Scaling a 6,000-Person Outsourcing Company Taught Me About When to Outsource

After co-founding UnifyCX and scaling it to 6,000+ employees, here's my honest take on outsourcing strategy: when to outsource, when to keep work in-house, and how AI is rewriting the whole model.

VCVik ChadhaFounder • Operator • Investor
What Scaling a 6,000-Person Outsourcing Company Taught Me About When to Outsource
Part of the SaaS Scaling series
Explore topic
First-hand Experience
Co-founded UnifyCX and scaled it to 6,000+ employees globally(Two decades providing contact center, business processing, and technology outsourcing solutions across multiple continents)
Most advice about outsourcing comes from people who have bought it. Mine comes from the other side of the table. I co-founded UnifyCX in Louisville and helped scale it from one location and a handful of employees to an international company with more than 6,000 people providing contact center, business processing, and technology outsourcing solutions. For two decades, my job was to be the thing founders and executives were deciding whether to outsource to. That vantage point teaches you something the buyer's side rarely sees: which outsourcing relationships compound into decade-long partnerships and which ones quietly fail within a year. The difference is almost never the vendor's talent or the hourly rate. It is whether the client understood their own operation well enough to hand part of it to someone else. That insight is the foundation of everything I believe about outsourcing strategy, and it is what I want to unpack here. I have told the story of building UnifyCX in Louisville elsewhere, so I will not retell it. If you want the origin story and why we built there, read Building a Tech & AI Ecosystem in Kentucky and The Midwest Advantage. This essay is about what the business itself taught me.

What Does Running an Outsourcing Company Teach You About Outsourcing?

The single biggest lesson: you cannot outsource a problem you cannot define. The clients who succeeded with us arrived with a process they already understood. They knew their call volumes, their quality standards, their edge cases, and what "good" looked like, because they had done the work in-house first. They were handing us execution, not comprehension. We could take a well-defined operation and run it better and cheaper than they could, because operations at scale was our entire company, not a department they tolerated. The clients who struggled arrived hoping we would figure out their operation for them. They outsourced the mess along with the work, and then were surprised when the mess showed up in the output. No vendor, however good, can supply the judgment about your business that you have not developed yourself. When something went wrong in those relationships, the client could not tell whether the process was broken or the execution was, because they had never separated the two. That pattern repeated so consistently that it became my first principle of outsourcing strategy: do the work yourself at least once before you hand it to anyone else. Not forever, and not at scale, but long enough to write the runbook. If you cannot write the runbook, you are not ready to outsource. You are ready to abdicate, which is a different thing with a different outcome. The second lesson is about relationships. Our first UnifyCX client came from a Louisville networking event, and fifteen years later they were still a client. Relationships like that are not procurement outcomes. They happen when both sides treat the arrangement as a partnership with shared incentives rather than a contract to be managed. The clients who got the most from us shared context generously, told us where their business was heading, and let us invest in understanding their customers. The ones who got the least treated us like a line item and got line-item performance back.

When Should a Founder Outsource?

Strip away the consulting frameworks and the decision is a version of the same question I now ask across the Scalable Ventures portfolio: is this work part of how you win, or part of how you operate? Outsource work when most of these are true:
  1. You have done it in-house and can define it. You know the inputs, the outputs, the quality bar, and the failure modes. You are buying execution capacity, not answers.
  2. It is operationally mature, not strategically live. The process is stable enough that this quarter's learning will not rewrite it. If you are still iterating weekly on how the work should be done, keep it close.
  3. Excellence in this function is someone else's core business. A dedicated operator with real scale will run a mature support, back-office, or technology operation better than your internal team, for the same reason you build your product better than they would. Specialization is real.
  4. Scale is lumpy or uncertain. If your volume swings seasonally, or you do not know whether you will need 10 people or 100 in a year, renting flexibility beats owning fixed cost. This is the same capital-efficiency logic I apply to everything: constraints force discipline.
  5. The economics genuinely work at your scale. Not the rate-card comparison, but the full cost: management attention, quality assurance, transition time, and the overhead of coordinating across organizations. Outsourcing has a real coordination tax, and it only pays when the volume justifies it.
Notice what is not on the list: "we don't understand this area." That is the most common reason companies outsource and the worst one. Confusion does not get cheaper when you export it.

When Should You Keep Work In-House?

Keep work in-house when it is the work you are learning from. Early-stage founders should almost never outsource the functions where the company's core insight is still forming. If you are pre-product-market fit, your customer conversations are not a support cost, they are your research pipeline. Handing them to a vendor, even a great one, cuts the feedback loop that tells you what to build. I have watched founders outsource customer support at exactly the stage when every ticket was a signal, and then wonder why their roadmap drifted from reality. The same logic applies to anything that constitutes your differentiation. At UnifyCX, operations was the product, so we obsessed over hiring systems, training programs, and quality assurance. Those were the last things we would ever have handed to someone else. Your equivalent might be your data pipeline, your onboarding motion, or your pricing engine. Whatever it is, the rule holds: outsource context, never core. If a capability is the reason customers choose you, it belongs inside the building, even when a vendor could technically do it cheaper. There is a third category worth naming: work you plan to outsource eventually but have not yet systematized. Keep it in-house just long enough to make it definable, then hand it off. That sequencing — build, document, then delegate — is the same discipline that separates founders who scale from founders who become the bottleneck, a theme I keep returning to in how we build companies through the studio.

What Breaks When Outsourcing Scales?

Growing UnifyCX from a handful of people to 6,000+ across multiple continents taught me that scale does not just make an operation bigger. It changes what the operation is. The same transitions I later watched portfolio companies hit — what works with a team of 10 breaks at 50, and what works at 50 breaks at 200 — hit an outsourcing operation harder, because you are absorbing other companies' growth on top of your own. Three things break predictably: Informal quality breaks first. At small scale, quality lives in individuals: the experienced agent who knows the client's quirks, the manager who catches problems by walking the floor. At thousands of people, quality has to live in systems — hiring profiles, training programs, quality assurance loops — or it does not live anywhere. Operational efficiency at scale was not automatic for us; it required deliberate systems thinking around hiring, training, and quality assurance. Every client relationship that scaled successfully scaled because the system was sound, not because the people were heroic. Communication breaks second. When your outsourced team was eight people, the account lead knew everything. At eight hundred, information passes through layers on both sides, and every layer loses fidelity. The fix is unglamorous: shared metrics both sides trust, a regular operating rhythm, and named owners for every interface between the organizations. The pain of managing distributed teams across countries was real enough that it seeded one of our portfolio companies — HiveDesk came directly from the remote workforce management problems we lived through scaling UnifyCX. Alignment breaks last, and worst. A vendor relationship that starts aligned drifts as both businesses evolve. The client's strategy shifts; the vendor's margins compress; the contract that made sense in year one quietly stops describing reality in year four. The fifteen-year client relationships survive because both sides renegotiate openly as things change. The failed ones let the drift accumulate until it snaps. If you are the buyer, the practical implication is this: your outsourcing strategy needs an owner inside your company, permanently. Not a procurement contact — an operator who understands the work well enough to manage the system, notice the drift, and renegotiate before the snap.

How Is AI Rewriting the Outsourcing Model?

For decades, the outsourcing industry's core offer was labor arbitrage: the same work, done by more affordable people, organized well. AI is ending that era, and I say this as someone whose company employs thousands of people doing exactly this work. I have watched AI compress what used to be six-month implementation cycles at UnifyCX into weeks. AI agents now handle a growing share of customer interactions directly. Summaries that once consumed 60% of our leadership meeting time are generated in minutes. The technology is not hypothetical for us; it is deployed, and it is reshaping the unit economics of the entire BPO category. When software can handle the routine tier of interactions, selling human hours for routine work stops being a durable business. But here is what the "AI kills outsourcing" take misses: the work does not disappear, it moves up the stack. The interactions that reach humans are now the hard ones — ambiguous, emotional, high-stakes — and handling those well requires better-trained, better-supported people than the old volume game ever did. Meanwhile, someone has to design, supervise, and continuously correct the AI layer itself. At UnifyCX we run two-week learning sprints where teams review what the AI got right, what it missed, and what the humans learned from the gap; that structured feedback loop cut our model retraining time by roughly 40%. We embedded AI liaisons in business units to translate between the technology and the operation. That is not labor arbitrage. That is judgment arbitrage, and it is the future of the industry. The human side of this transition is the part leaders underestimate. Early in our AI rollout, one team resisted adoption not because they lacked skill but because they feared that succeeding with AI would eliminate their own jobs. Rebuilding that trust took explicit, repeated commitment to redeployment over displacement. Any founder deploying AI into an outsourced operation — or an in-house one — should expect the same dynamic and address it directly. I wrote more about this in Leadership in the Age of AI. For founders, the strategic consequence is concrete: before you outsource any process today, ask whether AI plus a small internal team could do it instead. Five years ago, the alternative to outsourcing was hiring. Now there is a third option, and for high-volume, rules-based work it is increasingly the right one. The best outsourcing partners will not fight this; they will sell you the AI-plus-human system and charge for outcomes rather than hours. The ones still selling seat counts are selling you the past.

What Does This Mean for Your Outsourcing Strategy?

If I compress twenty years on the vendor side into a checklist for founders, it looks like this:
  1. Do it yourself first. Run the process in-house until you can write the runbook. Outsource execution, never comprehension.
  2. Outsource context, keep core. If it is the reason customers choose you, it stays inside. If it is how you operate rather than how you win, it is a candidate.
  3. Price the coordination tax honestly. Vendor rates are the visible cost. Management attention, QA, and organizational interfaces are the real one.
  4. Check the AI alternative before you sign. For routine, high-volume work, AI plus a small internal team may now beat both hiring and outsourcing.
  5. Assign a permanent internal owner. Systems drift. Someone who understands the work must own the relationship, watch the metrics, and renegotiate early.
  6. Choose partners, not line items. The relationships that compound over a decade are the ones where both sides share context and incentives. If you treat your vendor like a commodity, you will get commodity output.
Outsourcing is neither a virtue nor a vice. It is a tool, and like every tool, it rewards the operator who understands the work before delegating it. The companies that get outsourcing strategy right are not the ones that outsource the most or the least. They are the ones that know, precisely, which work builds their advantage and which work merely supports it — and staff each accordingly.

Related Reading

Making a Build, Hire, or Outsource Decision?

If you're a founder weighing what to keep in-house and what to hand off:
Scalable Ventures logo
Flagship VentureFounder & CEO of Scalable Ventures — The AI-focused studio where I build, fund, and scale capital-efficient B2B SaaS.
Visit Scalable Ventures

Related Articles

Explore more insights on entrepreneurship, AI, and leadership:

Explore More

Dive deeper into related topics and resources:
On this page