An operation that had outgrown its tools.
DON VA Agency runs mass volume email marketing for its clients. Several contractors work in parallel on sending, deliverability, lead generation and replies, across dozens of mailboxes and domains.
The operation had outgrown its tools. Credentials sat in scattered chat messages. Nobody could say which mailbox had contacted which lead. Contractors could see each other's work, and the founder had no single view of what was actually getting done.
The solution, in four pillars
A containment problem, not a visibility problem.
The founder described it as a visibility problem. Underneath it was a containment problem. Mass volume email runs on other people's credentials and other people's domain reputation, and nothing in the setup kept either one contained.
Contractors came and went. Each needed enough access to do the job, and each could see everything. Credentials lived in chat messages, which means they lived forever, in a place nobody controlled, readable by anyone who had ever been in that thread.
And when a domain started failing, nobody could answer the question that actually matters: which leads had that domain already touched. The outreach existed. The mailbox existed. The link between them lived in whoever happened to remember it.
So the system had four jobs, and only one of them was tracking work:
- Give each contractor a locked folder, so access is granted by adding somebody to one place and revoked by taking them out of it.
- Put every mailbox in one database with an owner on every row, so a credential has a home instead of a chat history.
- Link every lead to the mailbox and domain that reached it, so a deliverability problem becomes a filter rather than an investigation.
- Put one dashboard above all of it, so the founder stops asking people what they did today.
Before
- Credentials pasted into chat threads
- Every contractor could see every client
- No link between a lead and the inbox that reached it
- A failing domain meant a manual search
- The founder chased updates in messages
After
- Credentials in one database, behind private fields
- A locked folder per contractor, their work only
- Every lead linked to its mailbox and domain
- A failing domain is one filter away
- One dashboard, refreshed on its own
Why this is not a project management problem
Nothing here was solved by better task tracking. Three of the four jobs are about who can reach what, and the fourth is about keeping a record of which asset did which thing. Build this as a to do list and you get a tidier version of the same exposure.
Every number here is readable in the screenshots.
The roles those folders are built around
Why a folder per contractor and not a folder per client
A client folder would have been the obvious structure, and it would have been wrong. The thing that needs containing here is not the client, it is the person. Contractors join, leave, and handle credentials that are not theirs. Organising by person means access is a property of the folder, so adding somebody is one action and removing them is one action.
The part that is easy to miss
The written procedures live inside each contractor's folder, named so they are read before the work starts, rather than sitting in a drive nobody opens. That is the difference between onboarding somebody in an afternoon and onboarding them over a fortnight of questions.
Central operations dashboard.
One screen for the entire operation. The dashboard pulls live from every contractor workspace: tasks by assignee, monthly workload by status, pending approvals and blocked tasks.
The founder no longer opens six folders or chases updates in messages. Workload imbalance, bottlenecks and progress are visible at a glance.
In this snapshot the system is tracking 1,921 tasks for a single contractor and 642 completed items across the team in one month. That volume breaks a spreadsheet quickly. Every card reads from the underlying workflows, so the numbers are current and nobody assembles them by hand.
What this replaced
Six folders opened one at a time, a message to each contractor asking what they had done this week, and a picture of the operation that was always a day out of date by the time it was assembled. The dashboard is not a report. Nobody writes it.
Private contractor workspaces.
Every contractor works inside a locked folder. They see their own work and nothing else.
The folders follow the same repeatable shape, with each role carrying the workflows that role actually needs:
- Daily Sending Execution: the day's campaign sends for each client, task by task
- Deliverability Monitoring: daily checks, so inbox and domain problems get caught before they burn a client campaign
- Lead Generation and Data Prep: sourcing and cleaning lead data before it enters a sequence
- Replies and Follow ups: every response handled and logged
- Weekly and Monthly Reports: standard reporting per contractor and per function
- Read Before Working procedures: process documentation sitting where the work happens
This protects client data, keeps each contractor on their own work, and lets the agency add headcount without exposing the whole operation to every new hire.
Why this matters most on the day somebody leaves
Because the folder is the unit of access, offboarding is removing one person from one place. Their work stays where it is, the next contractor inherits the same structure, and nobody has to remember which six threads held something that person could still open.
Centralised mailbox and credential management.
Mass volume email runs on dozens of mailboxes across several platforms. Before this build, credentials were scattered across chat messages and personal notes.
That was a security problem and a daily time sink at the same time. Now every mailbox lives in one controlled database, grouped by the contractor responsible for it.
New mailboxes enter through a dedicated intake form. No manual entry, no missed fields, no passwords pasted into chats. Private fields control who sees what, and platform tags plus login page links mean a contractor reaches any mailbox assigned to them in seconds. When a contractor rolls off, revoking access is one change instead of a scavenger hunt.
The security point, stated plainly
A credential in a chat thread is permanent, uncontrolled, and readable by everyone who was ever in that conversation, including people who have since left. A credential in a private field has one owner, one place to change it, and a list of who can see it.
Lead tracking, traceable back to the inbox.
Using space level relationships, every lead record links to the exact mailbox and domain used to contact it.
When a reply arrives, the team knows which inbox it belongs to, who owns that inbox and what happens next. Outreach, response and closing all run through one traceable pipeline.
The lead database also feeds downstream automation, including an n8n connected voice agent workflow flagged per lead, so qualified leads move on their own instead of waiting for a manual handoff. If a domain develops deliverability problems, every lead touched by that domain can be isolated instantly and rerouted through a healthy inbox.
Why the link is the whole point
When a domain burns, the useful question is not which campaign was running. It is which leads have already been contacted from it, because those are the ones that need a different inbox before anyone writes to them again. Without the link that is a search through exports. With it, it is a filter.
From scattered chaos to one structured system.
DON VA Agency went from scattered chaos to one structured system: private repeatable contractor workflows, secure centralised mailbox management, full traceability from inbox to lead, and a single dashboard showing the whole operation.
The system is built to scale. Adding a contractor, a new client campaign, or fifty new mailboxes adds rows, not chaos.
What the founder can answer without asking anyone
- Who is carrying how much work this month, and who is sitting idle.
- Which mailbox contacted a given lead, and who owns that mailbox.
- Which leads are exposed if a domain starts failing.
- What is blocked, and what is waiting on an approval from him.
- Who has access to which credentials, and what changes when somebody leaves.
What I am not claiming
The numbers on this page are volume, not improvement. I can show you what the system handles. I cannot show you what it replaced, because the old way produced no numbers to compare against. That is usually the point: the first thing a system like this gives you is the ability to measure at all.