Before you hire offshore, write the role down: the recurring tasks, the decision rules, and what a good week looks like. Shopify brands doing $1M to $10M should document process-driven roles like tier one support and order operations first, and leave judgment-heavy creative work until later.
The role has been open for six months because nobody ever wrote down what the job is. Hiring across an ocean does not fix that. It just puts a monthly invoice on it.
I have watched a lot of Shopify brands hit the same wall somewhere between $1M and $10M. Customer support, fulfillment exceptions, and the daily operational cleanup all run through one or two people, and those people quietly become the ceiling. The founder eventually decides to hire. The role gets posted. Four months later it is still open.
The usual explanation is that the local talent market is tight or expensive. Sometimes that is true. More often, in my experience, something else is going on, and it is uncomfortable: the job has never actually been defined. It lives in somebody’s head, assembled over three years of doing it, and the job posting is a list of vibes wearing a bullet list.
That matters enormously the moment you start looking offshore, because distance removes every informal correction mechanism you have been relying on without noticing. Nobody overhears the answer. Nobody reads your face. Whatever is not written down does not exist.
An important role that has stayed open too long usually means the job has never been written down, not that your local talent market failed you. That reframe changes what you do next, and it changes it before you spend a dollar on recruiting.
Nicolas Bivero, co-founder and CEO of Penbrothers, an employer of record that builds remote teams in the Philippines, gave me the cleanest readiness signal I have heard on this topic. Forget revenue thresholds and headcount triggers. In his words, “if you happen to have an open position, and it’s staying open for longer than what you want and hope for, that means that you’re struggling finding the right talent.” We got into the full readiness picture on episode 481 of the podcast, and I would start there if you are weighing the decision itself.
Here is where I want to push the idea one step further than we took it on the recording. An open role is a signal, but it is ambiguous. It can mean the market is tight. It can also mean the role as written is unhirable, because nobody can tell from the posting what success looks like or what the person would actually do on a Tuesday.
There is a fast test for which one you have. Try to write down, in one page, what a person in this role does every day, every week, and every month, and how you would know at the end of thirty days whether they were doing it well. If you can produce that page in an afternoon, your problem is the talent market. If you sit down to write it and discover you cannot, your problem is the documentation, and hiring anyone at all, local or offshore, will expose that rather than solve it.
Most founders at this stage find they cannot write the page. That is not a failure. It is the normal consequence of a role that grew by accretion rather than by design.
Start with roles where success is countable: tier one customer support, order and returns processing, product data entry, and recurring reporting. Leave brand creative, merchandising judgment, and customer-facing copy until you have a manager who can define what good looks like in those areas.
Nico’s rule here is about legibility, not talent. Process-driven roles onboard well because the KPIs and the workflows are explainable across a time zone. Creative roles are harder to hand off early, and his explanation of why is the most useful sentence in the episode: “the person might do a great job, but it’s not what the counterparty wanted.” That is not a knock on the hire. It is a description of what happens when nobody has written down whose taste governs.
Product data is the sleeper on that list. It looks like the most boring work in the building, and it has quietly become one of the highest leverage things a Shopify brand can get right, because product titles, types, and structured fields are what determine whether AI assistants can recommend your catalog at all. I dug into that with Rob Langenback in the episode on product naming and AI discoverability. If you are choosing a first offshore role and you have 500 or more SKUs with inconsistent data, that is a strong candidate.
A usable role document has five parts: the recurring task list with frequency, the decision rules, the escalation path, the access map, and a written definition of a good week. Budget 6 to 10 hours for the first one, and expect the second to take half that.
The recurring task list is the easy part, and it is where most people stop. Do not stop there. Sample four weeks of real volume from your helpdesk and your Shopify admin, and count what actually came in, because the list you write from memory will be missing the awkward twenty percent that generates most of the questions.
The decision rules are where the value is. What is the refund threshold this person can approve without asking? Which tickets get a discount code and which get an apology? What happens when a customer wants an exchange on an item that is now out of stock? Every one of these is a decision you make in four seconds and have never articulated. Write them as if then statements. When you cannot write one, you have found something that needs to stay with you for now.
The escalation path names a person, a channel, and a time expectation. Not “ask if unsure.” Something closer to: post in the operations channel, tag me, expect an answer within four hours during overlap, and if it is a chargeback or a legal threat, escalate immediately regardless of hour.
The access map is the part people do in a rush and regret. Scope permissions to the job rather than handing over an administrator seat, and take twenty minutes to read how store level permissions work in the Shopify admin before you invite anyone. Do the same in your helpdesk, and if you have not settled that stack yet, our roundup of customer service apps and what they cost per agent is a reasonable starting point.
Record rather than write wherever you can. A fifteen minute screen recording of you working through eight real tickets, narrating why you did what you did, is worth more than three pages of prose and takes a fraction of the time. If you want a starting structure for the posting itself, the job description template in our guide to hiring a virtual assistant covers the basics, and the role document is what you attach behind it.
Run your own 180 day plan in three phases: weekly one to ones and live shadowing through day 60, biweekly reviews against the written definition of a good week through day 120, then normal team cadence. The first six months are where offshore hires are won or lost, and you do not need a vendor program to run this.
Penbrothers built a 180 day onboarding framework it calls Hypercare around exactly this window. What I found most useful in Nico’s description was not the schedule but the design principle: both sides get onboarded, the client as well as the new hire, because the thing that fails is frequently the environment rather than the person. He was careful not to attach a retention number to it on the show, and I am not going to invent one.
The portable version of that idea is not complicated. Nico’s point about overlap is worth building around: aim for two to four hours where you and your new hire are genuinely awake at the same time, and protect those hours for questions rather than filling them with meetings.
The rest of it I learned firsthand rather than from a framework. When I was on the Merchant Success team at Shopify, I carried a book of merchants and had a team lead in a role we called People Accelerators. That person did no client facing work at all. Their entire job was making sure the people on their team had what they needed. We had weekly standups, ongoing training, and short async video updates dropped into Slack when the week was too busy for a live call. We had our own team channel and shared channels, so I could see what engineering and marketing were doing without asking anyone.
None of that was designed for offshore hires. It was designed for a distributed company, and it worked across Europe and the APAC region for the same reason it works across a twelve hour gap: somebody’s actual job was the health of the team, and information was ambient rather than requested. If your new hire has to ask permission to find something out, you have built a worse version of your own company for them to work in.
The all in cost of an offshore hire is compensation plus statutory contributions plus any management fee plus equipment, software seats, and your own management time. Nico is right that the management fee is the wrong number to fixate on, and the episode does not break down what the right number contains, so here it is.
Statutory contributions are not optional and they are not small. In the Philippines, employers must register and remit monthly to three separate government programs covering social security, health insurance, and housing, with penalties for late or missed remittance. This is the single most common thing brands paying freelancers directly get wrong, usually without realizing it.
Nico told a story on the recording that I have not stopped thinking about. A company came to Penbrothers in 2016 with five or six freelancers they were paying directly. One of them became pregnant, went to the doctor, and found out she could not draw on social security or health insurance because nobody had ever paid in. She raised it with the person she believed was her employer. That founder looked at the arrangement, decided it was not sustainable at any real scale, and restructured. That company now has more than 100 people employed through Penbrothers.
Then there are the line items nobody budgets. A laptop and a decent internet stipend. Seats in your helpdesk, your Shopify admin, and your project tool, which for most stacks runs $50 to $150 per person per month depending on how many tools you are carrying. And the one that actually costs the most: your own time. Plan on five to eight hours a week for the first two months, and if you cannot find those hours, you are not ready to hire yet regardless of what your revenue says.
One more, because it is the kind of thing that quietly ends a good working relationship. Nico flagged that wage inflation in the Philippines is real and that salaries need annual review. His framing stuck with me: if you do not adjust, “people will assume that you don’t like them and then they will find another job.” Put a compensation review on the calendar at the twelve month mark before you make the offer, not after somebody resigns.
Under $500K, use a contractor on Upwork or Fiverr to find out whether the work can be done remotely at all. Between $1M and $10M, document the role and hire properly. Above $10M, what you usually need is a manager, not another individual contributor.
I want to be fair to the gig economy here, because Nico was, and he had every commercial incentive not to be. He said plainly that when you are small and do not need a full time person, Fiverr and Upwork make total sense. His question comes later: once somebody becomes essential to your business, “do you really want to have that person on the platform, on a gig economy kind of setup that you have zero control over?”
I have lived the answer to that. Sending annual independent contractor agreements to people who are functionally on your team, asking them to confirm they handle their own taxes, opens a can of worms around classification, data access, and equipment that gets harder to close the longer you leave it.
For Emerging brands under $500K, the honest guidance is to automate before you outsource and to test with a contractor before you commit. For Growth brands at $1M to $5M, document one role, hire one person, and resist the urge to hire three. The failure mode I have watched most often at this stage is premature complexity: too many people, too many tools, and too many channels layered on before the fundamentals are written down. For Scaling brands at $5M to $10M, the constraint shifts from documentation to management capacity, and the fourth offshore hire is usually the one that breaks the founder who is still running every check in personally. For established operators above $10M, this is a hiring plan question rather than a first hire question, and the answer usually starts with a team lead.
Whichever stage you are at, the sequence does not change. Write the role. Then hire the person. Doing it the other way around is how you conclude that offshore hiring does not work, when what did not work was hiring somebody into a job that was never defined.
Document a process by recording yourself doing it rather than writing it from memory. Use screen recording software to capture fifteen to twenty minutes of real work, narrating each decision as you make it, then convert that into a written task list with frequencies, a set of if then decision rules, and an escalation path that names a person and a response time. Sample four weeks of actual volume from your helpdesk and store admin so the awkward edge cases make it in. Expect the first role document to take 6 to 10 hours across a week. The test of whether it is finished is simple: could someone who has never worked for you follow it without asking you a question in the first three days?
Outsource process driven roles first: tier one customer support, order and returns processing, product data entry, and recurring reporting. These transfer well because success is countable and the workflows can be explained across a time zone. Hold back roles where quality is subjective, including brand creative, merchandising decisions, and customer facing copy, until you have a manager who can define what good looks like and review against it. Product data entry is often the best first choice for catalogs above 500 SKUs, because inconsistent product titles and missing structured fields directly affect whether AI shopping assistants can recommend your products, and the standard for correct is easy to write down.
Plan for six months to full productivity, with the first sixty days requiring the heaviest investment from you. A workable structure is weekly one to ones plus live shadowing through day 60, biweekly reviews against a written definition of a good week through day 120, then your normal team cadence. Employer of record providers build formal programs around this window, and Penbrothers runs a 180 day framework it calls Hypercare, but you can run the same rhythm yourself with a calendar and a document. Budget five to eight hours per week of your own time for the first two months. If those hours do not exist in your week, delay the hire rather than compressing the onboarding.
An offshore hire costs compensation plus statutory contributions plus any management fee plus equipment, software seats, and your management time. In the Philippines, employers must remit monthly to three separate government programs covering social security, health insurance, and housing, so the loaded cost sits meaningfully above the salary figure. Software seats across a helpdesk, store admin, and project tool typically run $50 to $150 per person per month. Ask any provider for the all in monthly and annual figure rather than the management fee alone, request the breakdown line by line, and budget for an annual compensation review, because wage inflation in the Philippines is real and a salary you never revisit reads as a signal to your employee.
Use a freelancer to test whether work can be done remotely, and an employer of record once the person becomes essential to your operation. Freelance platforms like Upwork and Fiverr are genuinely well suited to experimentation when you do not need a full time person and are still learning what the role requires. The calculus changes when that person holds customer data, uses your systems daily, and would be difficult to replace. At that point worker classification, statutory contributions, data privacy, equipment control, and retention all become your exposure rather than theirs. The practical trigger is straightforward: if losing this person would disrupt your operations for more than a week, the freelance arrangement has outgrown itself.
