It usually starts on a Tuesday. You open your calendar and there it is — a wall of colour-coded blocks stacked end to end, with a single 30-minute gap at 4:45pm that you already know will get eaten by a “quick sync.” Somewhere between the standup, the client call, the internal review, and the meeting scheduled specifically to plan another meeting, the actual work you were hired to do quietly slides off the edge of the day. You’ll catch up tonight. You always do.
If that sounds familiar, you’re not imagining it, and you’re not weak for struggling with it. The coordination load that sits on top of modern knowledge work has grown into something close to a second full-time job — a job nobody applied for, nobody gets paid extra for, and nobody seems able to switch off. One worker on the engineering forum Blind described spending 16 of every 40 hours in meetings, keeping half an ear open in case a question got thrown their way, dreading the return to the office where they’d have to sit through it all in person. “It’s my nightmare,” they wrote. The replies were not sympathetic so much as competitive: 16 hours sounds like a pleasure cruise compared to my week.
This article is about the gap that opens up when that chaos goes unmanaged — and about a company that has spent 17 years quietly closing it. It’s also about a question that more business owners are asking out loud in 2026: when artificial intelligence can draft your emails, summarise your calls, and schedule your week, why would you hand any of it to a human at all? The answer, it turns out, is more interesting than either the AI evangelists or the skeptics want to admit.
The Hidden Tax Nobody Puts on the Invoice
Let’s name the thing first. The problem isn’t that you have too much to do. The problem is that the coordination of what you have to do has metastasised. Every task now arrives wrapped in a layer of scheduling, status updates, follow-ups, clarifications, and “just looping you in” messages that have to be processed before the task itself can begin.
The research on this is blunt. A survey of 1,500 professionals by Everything DiSC found that almost two-thirds of people who spent more than 15 hours a week in meetings reported severe stress, and that two-thirds of all respondents experienced fatigue specifically from unproductive meetings — the ones where everyone joins the call, looks around, and waits for someone to remember why they’re there. Joe Allen, who runs the Center for Meeting Effectiveness at the University of Utah, has documented something even worse: a compounding effect. “One bad meeting causes three more meetings,” he found, because follow-up meetings become necessary just to work out what the first one decided. Bad meetings breed. They are not a one-time cost; they are an interest-bearing debt.
And the damage doesn’t stop when the call ends. Switching costs are real and they are brutal — it takes the average person around 23 minutes to fully refocus after an interruption, which means a calendar broken into 30-minute fragments isn’t a productive day at all. It’s a series of cold starts. By the time you’ve remembered where you were, the next notification fires.
Bad meetings don’t end when you leave the room. One unproductive 30-minute call can cost you the better part of an hour once you count the time it takes your brain to climb back into focused work.
Some companies have started treating this as the emergency it is. Shopify pointed a bot at its calendars and deleted thousands of recurring meetings, freeing up roughly 322,000 hours of employee time. LinkedIn banned meetings on Wednesdays outright. These are not wellness gestures. They are admissions that the coordination tax had grown large enough to show up in the output of the entire company.
Here’s what most founders and small-business owners miss, though. The Shopifys and LinkedIns of the world can afford to throw engineering teams and policy mandates at the problem. The person running a 12-person agency, a boutique law firm, or a fast-scaling e-commerce brand cannot. They absorb the coordination tax personally — in their evenings, their weekends, and the slow erosion of the strategic thinking that was supposed to be their actual job. That’s the gap. And it’s wider than almost anyone realises.
What the Research Actually Says About Working Remotely
Before going further, it’s worth clearing away a myth, because the entire case for distributed support — virtual assistants included — rests on it being false. The myth is that remote work tanks productivity.
It doesn’t. The most rigorous evidence we have says the opposite, or at minimum, says it’s a wash. Stanford economist Nicholas Bloom, who has studied this since before it was fashionable, ran a randomised controlled trial at the Chinese travel company Trip.com involving 1,612 university-educated engineers, marketers, and finance staff, and published the results in Nature in 2024. The careful finding: hybrid work did not damage performance. What it did do was cut attrition by roughly a third. People stayed. They were happier. And the output held steady.
Go back to Bloom’s earlier work at the call centre operator Ctrip and the numbers are even friendlier to remote arrangements: a 13% performance increase among home-based workers and a 50% drop in attrition, strong enough that the company rolled the policy out across the business. A 2024 Nature Human Behaviour analysis of 60,000 Microsoft employees and a Bureau of Labor Statistics study spanning 61 industries both point in a consistent direction: for focused, individual work, distance is not the enemy people feared.
There’s a sharp caveat, and honest writing has to include it. A 2023 study in the Journal of Political Economy Microeconomics by Gibbs, Mengel, and Siemroth tracked more than 10,000 skilled professionals at an IT services firm and found that total hours worked rose by about 30%, with an 18% jump in after-hours work. In other words, output per hour can quietly fall when people are left to coordinate themselves — they end up working more to produce the same. Microsoft’s 2025 Work Trend Index found the same fault line from a different angle: cross-team collaboration scores dropped 17% in fully remote settings versus structured hybrid ones.
Read those two sets of findings together and a pattern emerges that almost nobody states plainly. Remote work doesn’t fail on the work. It fails on the coordination. The output is fine when someone is focused. The slippage happens in the connective tissue — the handoffs, the follow-ups, the “wait, who’s doing this?” The thing that breaks is exactly the thing the coordination tax describes. Which is why the answer was never “force everyone back into the building.” The answer was to put a capable human in charge of the connective tissue. That insight, as it happens, occurred to a South African entrepreneur a full six years before the pandemic made it obvious to everyone else.
How a Quit-Your-Job Decision Became Africa’s Largest VA Agency
In 2008, Karen Wessels had had enough of her boss. So she quit, and started a small remote admin company in Cape Town called Lime Tree Consulting Solutions. Nobody was calling it “virtual assistance” yet. The term barely existed. Her first client was German, and by her own account the discipline and structure of that corporate relationship left a permanent mark on how the company would later operate — a kind of cultural DNA that stuck around long after the name changed.
For six years, Lime Tree did what most freelance support outfits do: it found clients, completed tasks, and tried to keep the wheels turning. But Wessels had built and run several businesses before this one, and she’d lived the founder’s problem from the inside — the suffocating reality of trying to grow a company while personally managing every operational thread. She saw something that the gig-economy marketplaces flooding the market in the early 2010s either missed or ignored: business owners didn’t actually want access to talent. Access to talent is what a marketplace sells. Thousands of profiles, millions of gigs, infinite choice, and zero accountability for whether any of it works.
What owners wanted was the right person, matched to their specific situation, working reliably from day one — and someone other than themselves to make sure it kept working.
So in 2014, Lime Tree became VA Connect, and Wessels did something that sounds small but was structurally radical. She pioneered what the company calls the Managed Virtual Assistant model. Instead of being a hiring board, VA Connect would own the entire relationship: vetting candidates, assessing cultural fit, matching personalities to working styles, training its people, and standing behind the result. If a VA didn’t perform, that became VA Connect’s problem to fix, not the client’s.
“I don’t want to be the biggest VA company. I want to be the one where nobody leaves — not the clients, and not the VAs.” — Karen Wessels, Founder, VAConnect
That line is the whole philosophy in one sentence. The company grew into Africa’s largest managed VA agency, built a free training platform called VAVarsity to keep its people sharp, and now serves clients across nearly every continent. After 17 years and placements across four continents, Wessels says she kept watching the same cycle repeat — founders burned by unreliable freelancers, brilliant South African professionals overlooked by the global market, and agencies treating both sides as disposable transactions. VA Connect was her attempt to break it.
Back in 2015, in an interview with Lionesses of Africa, she summed up the ambition with characteristic directness: “We want to own the African space when it comes to outsourced assistance. After that, there’s the rest of the world!” A decade on, that no longer reads as bravado. It reads as a roadmap she mostly executed.
The South African Advantage Most UK Businesses Are Still Sleeping On
Here’s where the story gets genuinely surprising, and where the efficiency gap stops being abstract.
When you outsource support work, geography is destiny — not because of cost alone, but because of overlap. A virtual assistant who is asleep when you’re awake is not a partner; they’re a relay station. You send instructions into the void at 5pm and find out tomorrow whether they landed. The whole point of having help — the real-time, “can you sort this before the 2pm call” kind of help — evaporates across a 10-hour time difference.
This is the trap that the traditional offshoring destinations fall into for UK and European businesses. India and the Philippines offer enormous talent pools and rock-bottom costs, but for a London founder they mean late-night handoffs, next-day turnarounds, and the constant low-grade friction of working with someone on the opposite side of the clock.
South Africa sits in a different position entirely. On GMT+2, it offers near-complete business-hour overlap with the UK and Western Europe — a one-to-two-hour difference that means a South African VA is online, reachable, and working in real time during your actual workday. When you message at 10am, you get an answer at 10am. That single fact changes the relationship from outsourcing to something that feels a lot more like having a colleague who happens to work remotely.
Then there’s language. South Africa ranks among the strongest English-speaking nations outside the native-English world, and crucially, the accent is widely rated as neutral and easy to understand for British and European ears — without the perception friction that some other outsourcing accents can trigger on a sales or service call. For relationship-heavy, communication-heavy work — the exact category where research shows offshore arrangements usually struggle — this is a structural advantage that India and the Philippines simply cannot replicate for a UK-facing role.
Cultural alignment compounds it. South Africa’s business norms, professional etiquette, and frames of reference map closely onto Western markets. As one recruitment analysis put it, the country lets firms skip the late-night meetings and “cultural overcorrection” that other offshore models demand. Things just translate.
And none of this comes at a premium. Cost savings of 50 to 65% against UK and European salaries are achievable across most knowledge-work roles. So the trade-off that businesses think they’re making — cheaper but harder to work with — inverts. You get the time-zone overlap, the language fit, and the cultural alignment of a nearshore partner, at offshore prices.
Time-zone overlap of one to two hours. Cost savings of 50–65% against UK salaries. Among the most neutral, intelligible accents in the outsourcing world. South Africa isn’t a compromise between cheap and good — it’s a refusal to choose.
The market has noticed, even if individual businesses haven’t caught up. The 2026 Global Outsourcing Talent Index, produced by Ataraxis and evaluating all 193 UN-recognised countries, placed South Africa 5th in the world — the highest-ranked African nation, with its English-speaking workforce and digital infrastructure singled out as decisive. The UK is now the single largest source of outsourced work flowing into South Africa’s growing global business services sector, a corridor that accelerated through 2025 as British wage and employment costs climbed. The South African BPO market, valued at around $1.85 billion in 2023, is growing at roughly 10% a year. McKinsey ranked the country among the most attractive offshoring destinations on the planet years ago. This is not an emerging story. It’s an arbitrage that’s been sitting in plain sight, and Karen Wessels positioned VA Connect on top of it in 2014, before the rest of the world finished doing the maths.
The Human in the Loop: Why a VA Beats Pure Automation
Now to the question hanging over all of this in 2026. If AI can do the admin, why pay a person?
It’s a fair question, and the lazy answer — “AI can’t really do it” — is wrong and getting wronger by the month. Modern AI genuinely can draft a competent email, transcribe and summarise a meeting, sort a calendar, and generate a first pass at almost any document you need. For raw, repeatable, well-specified tasks, automation is often the correct tool, and any honest VA agency should tell you so.
But look closely at where the coordination tax actually lives, and you’ll notice something: almost none of it is well-specified. The exhausting part of your week isn’t writing the email. It’s knowing which email matters, reading the room on a touchy client thread, deciding that the angry message from your biggest account needs a phone call rather than a reply, remembering that this particular supplier always needs a gentle reminder two days early, and catching the typo in the contract that would have cost you £4,000. That work is judgement. It is context. It is the thousand tiny acts of caring about your business the way you care about it — and AI, for all its fluency, does not care. It has no skin in the game and no memory of why last month’s near-miss mattered.
This is the part that gets glossed over in the rush to automate. A generic AI tool will happily produce a confident, polished, wrong answer and hand it to your customer with the same cheerful tone it uses for everything. It cannot tell when “technically correct” is relationally catastrophic. It will write copy in your brand voice only if someone who actually understands your brand voice is steering it — and steering AI well is itself a skill that takes time most founders don’t have.
This is precisely the seam where the managed-VA model earns its keep. When a VA Connect client says they need someone who can “understand our brand voice and manage high-stakes client communication,” they aren’t hoping a freelancer interprets that correctly. They’re matched with a vetted professional who passed skill assessments, personality testing, and interviews built to find exactly that capability — and who then gets better at it over months of working inside one business.
The smart configuration isn’t human or machine. It’s a human in the loop with the machine. Let the VA use AI to crush the volume — the drafts, the summaries, the data entry, the first 80% — and then apply human judgement to the 20% that decides whether your client feels valued or processed. The VA becomes a force multiplier on the automation, catching the hallucinations, supplying the context, and humanising the output before it reaches a real person who can tell the difference. Pure automation gives you speed. A human in the loop gives you speed plus trust. In a market where every competitor now has access to the same AI models, trust is the only thing left that’s hard to copy.
AI can write the email in seconds. It cannot know that this particular client needs a phone call instead. The whole value of a great virtual assistant lives in that gap — and the gap is getting more valuable, not less, as automation commoditises everything around it.
The Managed Model vs. The Marketplace Free-For-All
So if the work needs a capable human applying judgement, why not just hire a freelancer off a marketplace and skip the agency fee?
Because the marketplace sells you the start of the relationship and abandons you for the rest of it. This is the difference Wessels built her company around, and it’s not a matter of degree. It’s categorical.
A marketplace optimises for breadth. It shows you a thousand profiles, lets you compare hourly rates, and hands you the entire burden of vetting, onboarding, managing, and quality-controlling a stranger. If they vanish, ghost you, or simply turn out to be mediocre, that’s your problem — and your sunk cost. You eat the onboarding time you invested, and you start the search over. For a business owner already drowning in coordination, you’ve just added the single most coordination-heavy task there is: line-managing a remote contractor you don’t know.
A managed agency optimises for depth. VA Connect assesses not just skills but personality compatibility, working styles, and values alignment — pairing a fast-moving startup founder with someone who thrives in chaos, and a detail-obsessed financial advisor with someone who naturally gravitates toward structure. The training is targeted through VAVarsity rather than left to chance. And the accountability runs the other way: if the match isn’t working, VA Connect replaces the VA at no additional cost and manages the full transition, so the client never loses their onboarding investment. The risk that a marketplace dumps on you, a managed agency absorbs itself.
Think about what that does to the coordination tax we started with. The marketplace freelancer adds to it — now you’re managing them on top of everything else. The managed VA subtracts from it — they take ownership of the connective tissue, the follow-ups, the scheduling, the handoffs, precisely the layer that the productivity research identified as the failure point of distributed work. One model hands you another job. The other takes one off your plate.
This is also why the founder’s own framing — do what you do best, and delegate the rest — is more than a tidy slogan. Delegation only works if the thing you delegate to is more reliable than doing it yourself. A flaky freelancer fails that test. A managed, accountable, time-zone-aligned professional passes it. The model is the product.
The Gap Is Wider Than It Looks — And It’s Still Growing
Step back and assemble the pieces, and the scale of the divergence is genuinely startling.
On one side, you have a business owner doing it themselves or stitching together gig-economy help. They’re paying the full coordination tax personally. They’re spending 16 hours of a 40-hour week in meetings and the evenings catching up. They’re losing 23 minutes of focus to every interruption. Their freelancers add management overhead rather than removing it. When help disappears, they restart from zero. And because their attention is consumed by operations, the strategic work that actually grows the business — the thing only the owner can do — keeps getting postponed to a someday that never arrives. The productivity research is clear about what happens here: hours go up, output per hour drifts down, and people quietly burn out while their metrics still look fine.
On the other side, you have a business that has put a capable human in charge of the connective tissue. The coordination tax is being paid by someone whose entire job is to pay it — efficiently, in real time, in your time zone, in clear English, using AI to handle volume and judgement to handle nuance. The owner gets their evenings back and their strategic attention back. Attrition risk is covered by a replacement guarantee. The relationship deepens and compounds over time instead of resetting.
The distance between those two businesses isn’t 10% or 20%. Over a year, it’s the difference between a founder who spent the year running operations and a founder who spent the year building the company. One of them has a coordination problem. The other has a competitive advantage that is almost invisible to outsiders, which is exactly what makes it dangerous to compete against.
And the gap is widening, not narrowing, for a reason that should make the do-it-yourself crowd nervous. As AI commoditises the easy tasks, the premium shifts entirely onto judgement, context, and human trust — the precise things a great managed VA supplies and a generic tool cannot. The businesses that figured out how to put a skilled human in the loop, in an aligned time zone, at a sustainable cost, are pulling away while the others are still asking whether they can just automate their way out of it. They can’t. Not all of it. And the part they can’t automate is the part that matters most.
The Bottom Line
Karen Wessels quit a job she’d had enough of in 2008 and, almost by accident, spent the next 17 years solving a problem most businesses still haven’t named. The coordination chaos that eats your week is not a personal failing or a scheduling glitch you’ll fix with a better app. It’s a structural cost of modern work, and it lands hardest on exactly the people least able to absorb it — the founders, the owners, the small teams growing faster than their systems can handle.
The companies pulling ahead aren’t the ones with the most AI subscriptions or the largest freelancer rosters. They’re the ones who put an accountable, capable, well-matched human in charge of the connective tissue, in a time zone that overlaps their own, speaking their language, supported by an agency that owns the outcome rather than just the introduction. South Africa happens to be the best place on earth to find that human for a UK or European business, and VAConnect spent 17 years building the model to deliver them. The gap between the businesses that have figured this out and the ones still struggling alone is no longer a rounding error. It’s the whole game.
DIY Coordination vs. Generic Freelancers vs. VAConnect
| Factor | DIY Coordination | Generic Freelancers | VAConnect (Managed VA) |
|---|---|---|---|
| Who pays the coordination tax | You do — personally, in evenings and weekends | You do — plus you now manage the freelancer too | The VA does, as their core job |
| Time-zone overlap (UK/EU) | N/A — it’s all on you | Random; often 8–10 hr gaps | 1–2 hr gap (GMT+2), real-time same-day work |
| Vetting & matching | None | Self-managed; you screen strangers | Skills, personality & values assessment by the agency |
| Onboarding risk | All yours | All yours; restart from zero if they quit | Covered — free replacement, managed transition |
| Quality accountability | You | You | The agency stands behind the outcome |
| Brand voice & nuance | Stretched thin across everything | Hit or miss; no continuity | Trained, context-aware, improves over time |
| AI used well (human in the loop) | Rarely — no time to steer it | Inconsistent | Yes — VA drives AI on volume, applies judgement on nuance |
| Cost vs. UK in-house salary | Hidden cost of your own lost hours | Low rate, high management overhead | 50–65% saving with managed support included |
| Effect on your week | Adds load | Adds load | Removes load |
| Trajectory over 12 months | Running operations | Repeatedly re-hiring | Building the business |
Sources referenced: Bloom et al., randomised controlled trial published in Nature (2024); Gibbs, Mengel & Siemroth, Journal of Political Economy Microeconomics (2023); Everything DiSC meeting-fatigue survey of 1,500 professionals; worker sentiment via the Blind professional forum; the 2026 Ataraxis Global Outsourcing Talent Index; Grand View Research / Alpha BPO on the South African market; Sourcefit on time-zone, accent and cost data; the Lionesses of Africa founder interview with Karen Wessels (2015); and VAConnect’s own published company history and service data (vaconnect.co.za / vaconnect.co.uk).