Blog
How Digital Influence runs content production, client approvals and reporting
See how Digital Influence manages content ideas, production, client approvals, publishing and reporting through one repeatable agency workflow.
Bogdan Cimpan · 16 September 2026 · 8 min read
An agency rarely starts struggling because it lacks tools. It starts struggling because the information about one piece of content is spread across too many of them.
The ideas live in a document. The raw footage sits in Drive. Feedback arrives on WhatsApp. The approval is somewhere in an email thread. The calendar is a separate app. At the end of the month, someone assembles the report by copying numbers by hand.
None of that is broken, exactly. Each tool does the job it was bought for. What nobody owns is the space between them, and that space is where briefs go stale, versions get confused and approvals quietly expire.
This is how Digital Influence, a Romanian social media agency, organises content production across client accounts, and how that process turned into Cadenus. Digital Influence uses Cadenus every day, so the workflow below reflects actual client work rather than a hypothetical process.
The agency behind the product
Digital Influence produces and manages social content for client accounts: filming days, reels, carousels, captions, scheduling and the reporting that follows. The work repeats every month instead of ending with a single project. Most project-management tools are not designed for that rhythm.
Cadenus grew out of that day-to-day work: content production, approvals and publishing for multiple client accounts, with the same sequence repeating every month.
The order matters. The process was defined first, inside a working agency, and the product was built to carry it. We wrote about the origin separately in why we built Cadenus; this article is about the operational side of the same story.
Where the workflow used to break
These are the failures that showed up often enough to be treated as structural rather than as bad luck.
- Briefs that were not finished. An editor opens a task and still has to ask three questions before starting. The delay gets recorded as a production delay, but it happened before production.
- Several versions of the same video. v2 in Drive, v3 sent on WhatsApp for a quick look, and no single place that says which one is current.
- Feedback arriving in more than one channel. Part of it in a voice note, part in an email, part said on a call. Reconstructing what was actually agreed takes longer than applying it.
- Approvals with no clock. A video sent for review on Tuesday is not late on Thursday, because nobody ever said when it was due.
- No reliable view of workload. The question "what is everyone working on this week" could only be answered by asking everyone.
- Reporting rebuilt by hand every month. The numbers existed, but connecting them back to the specific content that produced them was manual work.
- No clear read on per-client profitability. Revenue was known. The effort actually spent on each account was not.
The complete content workflow
This is the sequence every piece of content passes through. The value is not in the number of stages, it is in the fact that each one has an owner and an explicit condition for moving on.
| Stage | Owner | Cannot move on until | Recorded as |
|---|---|---|---|
| Idea | Content strategist | It names a client, a format and an angle, not just a topic | An item in the client's idea list |
| Brief | Content strategist | An editor could start without asking a question | The brief attached to the item |
| Script | Content strategist | The hook and the call to action are written, not implied | A version on the item |
| Production | Videographer | The raw material exists and is linked to the item | Footage linked, stage stamped |
| Editing | Editor | One named editor owns it, with a due date | Assignee and deadline on the item |
| Internal review | Someone other than the editor | A second person has said it is ready to leave the building | Internal approval on the item |
| Client approval | Client, with follow-up from the account manager | The client approves or writes down the changes | Decision, author and timestamp |
| Scheduling | Account manager | Caption, channel, slot and thumbnail are set | An entry in the calendar |
| Publishing | The system | The scheduled slot arrives | Published post with its live link |
| Reporting | Account manager | The month is closed | Results attached to the published items |
Three of these stages carry most of the risk.
The brief is the real deadline. Most delays attributed to editing started as a brief that was accepted while still incomplete. Using "an editor could start without asking a question" as the exit condition catches the problem while it is still cheap to fix.
Internal review protects the client review. Every typo or wrong logo caught internally prevents an avoidable revision in front of the client.
Client approval needs a date attached. A review request without a deadline is a suggestion. The rules that make this work are the same regardless of tooling, and are set out in the client approval workflow.
Stage age is the number worth watching here. An item that has been sitting in one column longer than usual is either blocked or forgotten, and both need the same intervention. The deeper version of this model is in the video pipeline guide.
What Digital Influence standardized
Each of the earlier failures was answered with a rule first. The product support came after the rule, not instead of it.
| Problem | Operational rule | How Cadenus supports it |
|---|---|---|
| Ideas and briefs scattered | One place per client for ideas and briefs | Ideas and production items live on the client, and an idea is promoted into production rather than retyped |
| Unclear state of work | A fixed set of stages, each with an exit condition | The pipeline board, with the stage stamped on every move |
| No named responsible person | Every item has exactly one owner | Assignee per item, visible on the board and on the person's own view |
| Invisible deadlines | Every item carries a due date | Deadlines on items, surfaced per person and per client |
| Verbal approvals | Approval is an event with an author and a time | Client portal decisions written to the item's history |
| Lost revision history | Every version and every comment stays attached | Versions and comments on the item, not in a thread |
| Calendar detached from production | Only approved work reaches the calendar | Approved items move into the composer and the calendar |
| Reporting rebuilt by hand | Results stay attached to the content that produced them | Published posts carry their own results into the client report |
Launch reel · v20:32The client side stays deliberately small: one link, the current version, approve or request changes. Notice that the decision and the comment land in the record underneath. That record is what ends the "we never approved this" conversation, in both directions.
Why separate tools stopped being enough
This is not an argument against Notion, Trello, Drive, Slack or WhatsApp. Each of them is good at what it does, and the agency still uses some of them for what they are good at.
The problem was not that any single tool failed. The problem was that the workflow between them depended on people remembering what to update.
When a video is approved in one place, someone has to move it to a second place, then copy the caption into a third, then note the result in a fourth. Each handoff works most of the time. The failures are rare per handoff and constant in aggregate, and they always surface at the worst moment: the wrong cut published, a stale caption, a client asking about a post that was approved but never scheduled.
Adding another tool makes that worse, because it adds another connection to maintain by hand. The durable fix is to remove unnecessary handoffs.
The calendar is the clearest example. When it is a separate app, it is a second list that has to be kept honest. When it is fed by the production pipeline, unapproved work simply cannot appear in it.
What changed after using one workflow
These are the changes the team can actually point to.
- Fewer places to look. The item itself shows the current status, so nobody has to reconstruct it from three different threads.
- Clearer ownership. Every item names one person, so "who has this" stopped being a question asked in a group chat.
- A shorter path to approval. The old path asked the client to open a storage account, or to approve somewhere in Slack or WhatsApp, and a review could sit for days. Now they open a link, watch, and either approve or leave the comments in the same place. That is the single change the team felt most.
- Work that can be found later. Versions, briefs and comments stay attached to the item, including months after publishing.
- A usable read on production volume. These are recurring accounts, not one-off projects: every month, each client gets at least 15 pieces of content, and hundreds of items move through the pipeline. What is in progress, for whom, and by when is answerable without asking anyone.
- Reporting tied to real activity. The month's report is built from the posts that actually went out, not reassembled by hand.
The main change was simple: work became findable and decisions became traceable. That made ownership, approvals and reporting easier to manage.
Lessons for other social media agencies
Most of this transfers regardless of which tools you run.
- Define the stages before you choose the tool. A tool applied to an undefined process produces a tidier version of the same confusion.
- Give every piece of content one owner. Not a team, not a channel. One name, visible to everyone.
- Treat approval as a recorded event. Who approved, when, and with what comment. If it only exists in a conversation, it did not happen.
- Do not keep the calendar separate from production. A calendar that anyone can fill independently will eventually contain work that was never approved.
- Decide how you will report before you publish. Reporting is painful mostly when the link between a published post and its result has to be rebuilt afterwards.
- Do not automate a process you have not defined. Automation applied to an unclear workflow makes the unclear part run faster.
About Digital Influence
Digital Influence is a Romanian social media agency working with service businesses, personal brands and growing companies. The team handles social media strategy and production, publishing and paid distribution, and runs its own client work on Cadenus.
The process came first
Cadenus was not built to add another tool to an agency stack. It was built to remove the work of holding a stack together: the copying, the reminding, and the reconstructing that sat between tools that each worked perfectly well on their own.
The workflow itself is portable. An agency can adopt most of it with the tools it already uses. Cadenus keeps the process intact when the team gets busy.
If the sequence looks like yours, see how it runs in the product, or start a trial and bring your current chaos with you.