Guides
Virtual assistant agency software and KPI guide
Virtual assistant agency software should be chosen by job, not by brand. Seven jobs a tool set must cover, the access rules, and the measures worth keeping.
What to take away
- Named tools appear below as examples. What a tool is called matters far less than which of the seven jobs below it does, and whether you can get your data out of it.
- Buy one tool per job and refuse the second. Overlapping tools are where small agencies quietly spend their software budget.
- Access control is the part of tooling that carries real risk. You are holding other people's credentials, and the rules for that belong in the tool decision, not after it.
- Six measures are enough. A dashboard with thirty numbers is a dashboard nobody acts on.
This page states no prices, no vendor claims and no benchmark figures. Where a control or an obligation is described, the body publishing it is named.
Seven jobs a tool set has to do
| Job | What it must produce | Who depends on it |
|---|---|---|
| Intake | One place work arrives, with a record of what was asked | Client, assistant, and every scope conversation later |
| Work tracking | Task states, ownership, and what is waiting on whom | Leads and capacity planning |
| Time capture | Hours by client and by service line | Pricing, profitability, and the monthly report |
| Documentation | Working standards, per agency, per service, per client | Training and continuity |
| Credential handling | Where client access lives, and how it is withdrawn | Security, and every exit |
| Communication | Client conversation, separate from internal conversation | Everyone |
| Books | Invoicing, payments, and the financial record | The whole business |
Seven jobs. Most agencies can cover them with fewer tools than they think, and the second tool for any one job is a cost with no return.
Seven jobs a tool set must do
Job
- Intake
- One place work arrives
- Work tracking
- Task states, ownership
- Time capture
- Hours by client, line
- Documentation
- Working standards per agency
- Credential handling
- Where access lives, withdrawn
- Communication
- Client separate from internal
- Books
- Invoicing, payments, record
Must produce
- Intake
- Client, assistant, scope
- Work tracking
- Leads, capacity planning
- Time capture
- Pricing, profitability, report
- Documentation
- Training and continuity
- Credential handling
- Security, every exit
- Communication
- Everyone
- Books
- The whole business
Who depends on it
- Intake
- Work tracking
- Time capture
- Documentation
- Credential handling
- Communication
- Books
Named tools that cover the seven jobs, as examples rather than endorsements. Intake: Google Forms, Jotform, Typeform. Work tracking: Asana, Trello, ClickUp, Monday.com. Time capture: Harvest, Toggl Track, Clockify. Documentation: Notion, Google Docs, Confluence.
Credential handling: 1Password Teams, Bitwarden, Keeper. Communication: Slack, Microsoft Teams, Google Chat. Books: QuickBooks Online, Xero, FreshBooks. A second tool for any one job is a cost with no return.
The two questions that decide a purchase
Can somebody who has never seen it do the job on their second day? Assistants are trained continuously in this business. A tool that needs a week of familiarity is a tool that taxes every hire you ever make.
Two questions that decide a purchase
Can a new person do the job on day two?
Buy it, training cost stays low
Reject it, it taxes every hire
Can you export your data, in full, without asking anyone? Time records, task history and documentation are the raw material of your pricing and your training. A tool that holds them hostage will decide your future decisions for you.
Everything else, the feature comparison, the integrations, the interface, is secondary to those two.
What not to buy
- A second tool that does a job an existing tool already does adequately.
- Anything bought for a client you have not signed.
- A tool chosen because a competitor uses it. Their bench, service mix and documentation are not yours.
- Analytics you will not act on. Every dashboard has a maintenance cost and most produce no decision.
- Per-seat tools for people who need it once a month, when a shared record would do.
The commonest overspend in small agencies is not one expensive purchase. It is four cheap ones with overlapping jobs, all kept because moving the data is tedious.
What not to buy
- A second tool doing an existing job
- Anything for an unsigned client
- A tool because a competitor uses it
- Analytics you will not act on
- Per-seat tools for monthly users
Credential handling deserves its own decision
You hold access to other people's mail, customers and books. That is the part of this business with real consequences if it goes wrong.
Credential handling rules
- Credentials never shared in a message
- Every grant has a named holder
- Every grant has a recorded removal route
- Access withdrawn the day somebody leaves
Size the controls to what you can genuinely maintain. The NIST small business quick-start guides describe practices built for operations this small, and the CISA material for small and medium businesses covers the same ground from the threat side. A short policy that is followed every time protects more than a long one nobody reads.
The books, and what they have to support
Your financial tooling has two jobs beyond compliance: it has to let you attribute revenue to service lines, and it has to let you see cash rather than invoices.
What the books must support
- Attribute revenue to service lines
- See cash rather than invoices
- Retention and retrievability as baseline
- A service-line code on every invoice
Retention and retrievability are the baseline. The IRS guidance on which business records to keep sets out what has to be kept and for how long. Whatever tool you choose has to make that straightforward rather than a year-end reconstruction.
Add one thing the accounting package will not give you: a consistent service-line code on every invoice. Without it, no profitability question can ever be answered.
Six measures worth keeping
More numbers do not produce better decisions. These six each change an action.
- Utilization per person.Delivered hours divided by paid hours. The earliest warning of a pricing or scheduling problem.
- Reserved against used, per client.The number that makes a retainer legible to both sides.
- Repeat questions per account.A direct measure of documentation quality.
- Time in clarifying.How long work sits waiting on a client answer. Rising means vague requests or a thin working standard.
- Accounts with only one trained person.A count, not a rate. Every one of them is a client at risk.
- Cash collected against cash committed.Not revenue. What arrived against what you owe.
Named tools that can produce these measures: Harvest and Toggl Track for utilization and time in clarifying, Clockify for reserved against used, Asana and ClickUp dashboards for repeat questions and single-person accounts, QuickBooks Online and Xero for cash collected against cash committed, and Google Looker Studio or Databox for a monthly dashboard.
Report utilization as a percentage, reserved against used in hours or dollars, repeat questions as a count, time in clarifying in hours, single-person accounts as a count, and cash in dollars.
Read them monthly and per account rather than in aggregate. Averages hide the account about to become a complaint. What each of these means in practice, and how to compute it, feeds directly into the reasoning in the pricing and profit guide.
What the measures must not become
Three failure modes are worth naming.
A measure used to rank people rather than to find a system problem. Repeat questions per account is about documents, not about the person asking.
A measure with no owner. If nobody is responsible for acting when it moves, it is decoration.
A measure the tooling cannot actually produce. Define the formula and the data source before you promise to report it, or you will spend month end reconstructing something approximate.
Tooling follows process, not the other way round
The order matters. Decide how work will arrive, how it will be tracked and what has to be documented. Then choose tools that do those things.
Agencies that reverse the order end up with a process shaped by a product's assumptions, which is fine until they change tools and discover the process has to change too.
The tools serve a process defined in three guides.
The onboarding workflow for new clients covers account setup, the sales and proposal guide covers what was promised, and the hiring and training guide covers how documentation gets built.
Who is in the systems, and on what basis
One question sits underneath both tooling and measurement: who employs the people holding client credentials, and who directs their work day to day.
It is an access question and a status question at once.
The Labor Department's misclassification myths page says the economic reality of the working relationship settles worker status, not the wording of an agreement. An agency that cannot answer plainly will struggle with the sophisticated buyers who ask.
Record the answer in the account file alongside the access register. Clients ask, and the answer should not vary by who is asked.
Reviewing the stack
Once a year, list every tool, the job it does, what it costs annually, and who would notice if it disappeared. Anything without an answer to the last column is a cancellation.
Then check the measures against decisions actually taken in the past year. A number that never changed anything should stop being reported.
Where the stack genuinely serves retention, the renewal and reporting playbook describes the practices. Your public claims about tooling and capability face the same evidence standard as everything else in the marketing and growth guide.
Common questions
Do I need dedicated agency software?
Not early. Seven jobs, covered adequately, matter more than a product built for the category. Specialized tooling earns its place when coordination across many accounts becomes the constraint.
How many tools is too many?
More than one per job. That is the whole test.
Should clients get access to my task board?
Only if it is genuinely readable to them and you are prepared for every state change to be watched. A monthly summary usually serves them better.
What is the most commonly missing tool?
Time capture by client and service line. Its absence is why so many agencies cannot answer basic questions about which work earns.
