What to build before you hire a VA

Most first virtual-assistant hires fail for a boring reason: the business wasn't ready, not the person. The book behind this kit opens with exactly that story: a new VA who spends her first week sending twenty clarifying emails because nobody wrote down how the business actually works, and an owner who ends up more stressed than before he hired anyone.

The instinct afterward is to blame the hire and try again with someone else, which usually produces the same result, because the actual gap was preparation, not talent. A VA relationship runs on documentation that doesn't exist yet in most small businesses: not because owners are disorganized, but because the steps they do automatically were never written down for anyone else to follow. The book's framing is blunt: preparation determines most of whether a VA succeeds, and it happens before you post the job.

The task inventory, and what actually belongs on a VA's plate

The book's method for figuring out what to delegate starts with a full week of tracking, not guessing. Write down everything you do in the business for seven days, including the small stuff: checking email, scheduling calls, updating a social post, answering a routine customer question. Most owners are surprised by how long the list gets once it's actually written out.

Once the list exists, sort every item into three groups. The first is work only you can do: the calls that require your specific relationship with a client, the decisions with real consequences. The second is work you should keep doing because you're specifically good at it, like strategy or the creative direction of the business. The third is everything else: tasks anyone could do with clear instructions. That third group is the VA task list, and the book's own observation is that most entrepreneurs are startled by how much of their week actually falls into it.

From there, the book has you pick the highest-impact, lowest-complexity items from that third group for the first round of delegation: not the hardest thing on the list, the easiest one that still matters. Each of those gets a written procedure: plain steps, screenshots where they help, and real examples of what a finished version looks like, so a new hire has a target rather than a guess.

The book calls this the "McDonald's method": the point isn't hiring a perfect person, it's building a system specific enough that a reasonably capable person can follow it and get a consistent result. Alongside the procedures, it has you write a role definition specific enough to be useful: not "manage email" but "respond to customer inquiries within four hours using the provided templates." And it has you set decision boundaries up front (which calls the VA can make alone, and which get escalated) so neither of you is guessing where that line sits three weeks in.

The last piece is measurement, and the book's rule here is worth keeping: track results, not hours logged. A VA who finishes in four hours and nails it beats one who logs eight and produces something you have to redo, and a tracking system built around hours instead of outcomes rewards the wrong thing.

Where people go wrong

The most common mistake is hiring before any of this exists: posting a job, doing a quick interview, and expecting the new hire to figure out the business from context. That's the exact failure the book opens with, and it's rarely about the person's skill; it's that nobody handed them a map. A second is documenting everything at once before hiring anyone, which delays the actual hire for weeks over a hundred-page manual nobody asked for. Three clear procedures for the highest-impact tasks are enough to start.

The third failure shows up after the hire: measuring the relationship by hours worked instead of results delivered, which produces exactly the wrong incentive and tells you nothing useful when something goes wrong. Without a results-based metric, it's hard to tell whether a slow week is a bad fit or just a harder task.

Deciding what to hand off to a person versus what to automate with software is the same underlying decision, and we cover that trade-off across the rest of the stack in the AI & Automation guide.