Your task list does not know what you agreed to.

What is agency management software?

Agency management software runs the operating side of client work: the engagement and the client it belongs to, the brief for the current piece of work, the deliverables that were actually agreed, who inside the agency owns each one, what state each is in, what has gone to the client, what the client decided and against which version, what was delivered, and what happens in the next cycle. It is the record of an engagement after it has been won.

What is agency operations software, and who needs it?

It is the same thing named after the problem rather than the industry — the layer that sits between a signed engagement and the work getting made. Creative, marketing and digital agencies need it once coordination itself becomes the job: several clients, several people, work that has to be agreed before it starts and approved before it ships. A studio of three with one client rarely needs it. An agency with twenty live deliverables across six accounts usually does.

How should agencies manage client briefs?

By turning the brief into two separate things and keeping both. The brief is what the client asked for, in their words, and it belongs to the cycle it arrived in. The scope is what the agency agreed to deliver against it, written as named deliverables with owners. Keeping them apart is what lets you answer, three weeks later, whether something was always part of the job or arrived afterwards.

Project vs retainer — how should the workflow differ?

A project agrees its scope once and closes when that scope is resolved. A retainer agrees scope per cycle, closes the cycle, and keeps the client, the history and the terms running into the next one. The practical difference is what resets: in a retainer the brief and the deliverables reset each cycle while the relationship does not. A system that only understands projects will quietly turn every month into a new unrelated project.

Can recurring retainers run in monthly cycles?

Yes, and that is the shape most retainers actually need: a cycle with its own brief and its own agreed deliverables, closing when those are resolved, with the engagement and its history carrying forward. Whether anything unused carries into the next cycle is a commercial decision your agreement makes — the system applies your rule rather than assuming one.

Can client change requests be recorded?

Yes, and recording them is the point rather than a formality. A request that arrives mid-cycle is attached to the engagement, sits outside the agreed set, and stays there until somebody agrees it in, quotes it, or moves it to the next cycle. The record keeps the decision and who made it, so the question of whether something was included has an answer that does not depend on memory.

How should agencies track agreed scope?

As a named set of deliverables, not as a description. A paragraph in a proposal cannot be checked against later; four named items with owners can. Once that set exists, in scope means membership of it, out of scope means a request against it, and the argument three weeks from now is about a record instead of a recollection.

Feedback vs approval — what is the difference?

Feedback is input: comments, requested changes, a reaction to a version. Approval is a decision that this specific version may move forward. A comment is not an approval, a view is not an approval, and a conditional yes is a request for another version rather than a decision on this one. Keeping them apart in the record is what makes “the client approved it” a checkable statement.

Can clients approve work in a portal?

Where a client portal is in scope, yes — a controlled view of what is ready, a place to comment, and a decision recorded against the version they were shown. The internal delivery record stays the source of truth; the portal is a surface onto the part of it clients are meant to see, which is a deliberate decision rather than a default.

Does agency software handle deliverable versions?

It references them — v1, v2, a revision reference, whatever your agency already calls them — so an approval can be tied to the thing that was actually approved. It is not a design file store, a version-control system or an asset library, and it does not diff artwork. The files stay where your team makes them, and the record knows which version the decision belongs to.

Agency operations vs client portal — which owns what?

Agency operations is the internal source of truth: the agreed scope, ownership, internal states, blockers and the decision record. A client portal is a controlled outside view onto part of that — what is ready, what to comment on, what has been approved. The portal is a surface, not the system, and the useful shape is one record with two audiences rather than two records that disagree.

What happens when a retainer cycle ends?

The cycle closes when every deliverable agreed for it is resolved, and the engagement carries on. The client, the history, the past decisions and the commercial terms stay; the brief, the agreed deliverables and the outstanding decisions reset for the new cycle. Anything unfinished follows your agreement rather than a default — carried, capped or dropped.

Agency operations vs project management software?

A project manager owns tasks, dates, dependencies and who is assigned. Agency operations owns the things around those tasks that are specific to client work: what was actually agreed, whether a new request is inside or outside it, whether something is internally finished or client approved, which version an approval belongs to, and what happens when a cycle ends. This is not a better task board. It is the layer a task board has never held.

Agency operations vs CRM?

A CRM owns everything up to the moment the client says yes — the prospect, the pitch, the proposal, the commercial relationship. Agency operations owns everything after it. They meet at the handover: a won deal becomes an engagement with a brief and an agreed scope, and an agency running both usually wants that handover to be a link rather than a retype.

Does agency operations replace HRMS or accounting software?

No, and it should not try. HRMS owns the employee record, attendance and the employment lifecycle; this record only knows who owns a deliverable. Accounting owns the invoice, the ledger and the payment; this record can carry a billing reference and mark a milestone ready, and stops at the point money starts.

Does agency management software include time tracking?

It can, and it should only where the agency genuinely needs it. Effort estimates, actuals, timers and billable designation are ordinary parts of some agency builds and pure overhead in others. What is never included is monitoring — no screenshots, no activity scoring, no surveillance of how somebody spends an afternoon. Time is recorded because it informs delivery and billing, or it is left out.

Can agency software show team workload and capacity?

It can show what is genuinely knowable: who owns which deliverables, what is already assigned, what is blocked and what could realistically start next. That is built from real assignments rather than a model. What it will not do is compute a utilisation percentage, forecast capacity or match people to work by skill score — those need inputs and definitions most agencies do not have, and inventing them produces confident numbers nobody can defend.

Can agency software calculate project profitability?

Only where the inputs genuinely exist. Profitability needs the fee, the real cost of the people who worked on it, how their time was allocated and an agreed accounting treatment — and if any of those is missing, the number is a guess wearing a decimal point. Where an agency has all four, it can be built and defined explicitly. Where it does not, this record shows what was agreed, delivered and approved, and leaves margin to the books.

What determines the scope of an agency operations project?

Mostly four things: how many engagements run at once, whether retainers or projects dominate, how a scope change actually gets decided, and whether clients will make decisions inside the system. Feature lists predict an agency build badly; those four predict it well.

Who owns the system and the data after a custom build?

You do. The code, the database and the hosting account are yours, and every client, engagement, deliverable, comment and decision in it is yours. There is no per-seat licence, nothing meters how many clients you run, and nothing stops working if a relationship ends.

Can current clients and projects migrate from spreadsheets or existing tools?

Usually, and the first job is finding out what each current tool can export rather than assuming. Clients, live engagements and open deliverables normally move cleanly once duplicates have been decided. Three years of task comments, file attachments and custom fields depend entirely on what shape they are in and whether that history is worth the work — which is a decision made with you.

Can it connect to the CRM and project tools we already use?

That is often the right shape rather than replacing everything. A won deal in a CRM becoming an engagement here, or an existing project tool keeping the task detail while this record holds scope and approvals, are both normal. Each connection is scoped against what the other tool can actually expose — we check that first, because a connection nobody verified is a promise nobody can keep.

What should an agency digitise first?

Engagements and agreed scope, then approvals. A clean record of what you owe each client is what everything else hangs off, and an approval you can point at is what makes delivery defensible. Digitising tasks or time before those exist just moves the confusion into a nicer tool.

Custom agency software, or off-the-shelf?

Off-the-shelf is the right answer more often than agencies admit. A capable project tool plus a CRM and a shared drive covers a great many agencies, and there are established agency platforms that cover more. A custom system earns its place when the agency-specific half stays scattered around those tools — scope that has to be agreed and defended, approvals that must name a version, retainer cycles that are not projects, or several systems that have to stay in step rather than be replaced.

Does every agency need a custom agency management system?

No. A small studio with a handful of clients runs perfectly well on a project tool, a CRM and good habits, and building something would be overhead it never recovers. It starts paying when coordination itself becomes the job: enough live engagements that nobody holds them all in their head, scope arguments that need a record, or approvals that have to be findable a year later.

Do agency teams need a mobile app for this?

Usually not. Agency operations is desk work — scope, review, approval and delivery all happen where the work happens, and a responsive web system on a laptop covers it. An app earns its place when there is a genuine recurring mobile action, device notifications are central to how the team works, or people are regularly on location. It is a decision made from how the agency actually operates rather than a default.