Compared with a traditional agency
An agency rents you output. This leaves you an asset.
Agencies are good at things this is not good at. The comparison worth making is narrower: for recurring content in a founder-led company, what do you have on the day the relationship ends.
The problem
The retainer produces work. It does not produce capability.
A content retainer buys a monthly quantity of deliverables. That is a real service and for a lot of companies it is the right one. The structural issue shows up at the end rather than during: after two years and a substantial spend, you have the published work and nothing else. The understanding of how your brand sounds lives in the agency's account team, and it leaves when they do.
The second issue is the layer. You buy senior thinking in the pitch and get junior execution in the delivery, with a review process in between that is meant to close the gap. Sometimes it does. When it does not, the correction happens on your side, in your review, on your time.
Neither of these makes agencies bad. They make agencies a rental, and rentals are the wrong shape for something you intend to do every week for years.
What it costs
What you are actually paying for.
- A coordination layer
- Account management exists because the person doing the work is not the person who understood the brief. It is a real cost and it does not appear as a line item.
- Re-briefing on turnover
- Agency staff turnover is high. Each change means re-explaining the brand to someone new, and the quality dip afterwards is yours to absorb.
- Review time on your side
- Catching off-brand language in a final review is a permanent tax, because the review catches instances and never removes the cause.
- Zero terminal value
- Stop paying and output stops the same month. Nothing was accumulating.
Why the usual answer does not hold
The model is built to keep the knowledge on their side.
This is not cynicism. If an agency wrote down everything it knows about your brand in a form you could hand to anyone, it would have converted a retainer into a one-off project. The incentive runs the other way, so what you get is deliverables and a relationship, and the understanding stays where it is commercially useful.
The alternative is not a cheaper agency. It is inverting what the engagement produces: the specification, the term lists, the templates, and the publishing config are built as your assets from day one, and the ongoing work is operating them. If you stop, you keep the machine and you have to run it. If they stop, you keep nothing and have to start again.
Side by side
The differences that matter after month twelve.
| Retained content agency | Build and Run | |
|---|---|---|
| What you own at the end | The published deliverables. Working files sometimes, on request. | The specification, the term lists, the templates, the publishing config, and every source file, in your repository from day one. |
| Who does the work | Senior in the pitch, junior in the delivery, account management in between. | One accountable operator, with the standard written down so it does not depend on them being reachable. |
| How consistency is held | A style guide plus review. Catches instances, does not remove causes. | A written specification plus a blocked-term check at the pipeline. A draft containing a blocked term never reaches review. |
| How capacity scales | By adding people, each of whom re-learns the brand. | By adding formats to an existing pipeline. The specification does not need re-teaching. |
| Where brand knowledge lives | In the account team. It leaves when they do. | In a versioned file in your repository. |
| What a correction does | Fixes this piece. The same note gets given again next quarter. | Goes into the specification, so the class of error stops recurring. |
| Approval | Varies. Often a proof round with a deadline attached. | A standing gate. Nothing publishes without passing it, and there is no path around it. |
| If you stop | Output stops that month. Nothing transfers. | You keep operating from the runbook. Access is revoked from your own settings. |
| Where an agency is the better call | Campaigns, brand identity work, paid media buying, and anything needing a room full of people for six weeks. | Not offered. This is recurring operation, not campaign work. |
What gets installed
What exists at the end of Build that a retainer never produces.
- The specification
- A written file describing how the brand sounds, specific enough that two different people produce recognisably the same work from it.
- The blocked-term lists
- Enforced before a draft becomes a candidate for review, rather than caught during one.
- The production templates
- Code, in your repository. A new format is a change to a template rather than a new brief.
- The publishing config and the runbook
- Written on the assumption that you might operate it yourself, because you might.
Control and ownership
What each side is actually responsible for.
Stays human
- Positioning and every claim made on your behalf
- Approval before anything publishes
- Which topics matter this quarter
- Whether the specification is right, which you sign off
Yours to keep
- The specification and the term lists
- The templates and the publishing config
- Every source file and export
- Every account and credential
How it gets done
Moving from a retainer without a gap in publishing.
- 01
Keep the retainer running
Build takes weeks and does not need the current arrangement to stop. There is no dark period.
- 02
Build the specification from the agency's best work
Their strongest output is evidence of what right looks like. It goes into the specification rather than being discarded.
- 03
Run both for one cycle
Same brief, both sources, your review. You are comparing on your own standard rather than on a pitch.
- 04
Move channels over as you are ready
One at a time. Nothing requires switching everything at once.
- 05
Give notice when the comparison is settled
Not before. If the comparison goes the other way, you have a specification you keep and can hand to the agency.
Proof
The founder stopped being the bottleneck, and kept the storefront.
He makes the creative calls and approves the drops. The publishing runs underneath him on a daily cadence, and the storefront, the tracking, and the drop calendar are in his accounts.
- 10
- Days from nothing to shippingStorefront, brand, and the daily publishing railsProduction records · January 4, 2026
- 6 months
- Cadence held without a gapFirst upload 2026-01-04, most recent 2026-07-26YouTube Data API · July 27, 2026
- 71
- Videos publishedPublic uploads on the channelYouTube Data API · July 27, 2026
- $0
- Paid spend on the channelEvery figure below is organicProduction records · July 27, 2026
The full account, including reach figures with their capture date and paid spend stated, is on the NVUS Hearts case study.
How to decide
Which one to pick.
This is the right call when
- Content is a permanent function rather than a campaign
- You want the brand knowledge to be an asset you hold
- Consistency has already been a problem
- You would rather have one accountable person than a team to manage
It is the wrong call when
- You need a campaign with a start and an end date
- You need brand identity, naming, or design system work
- You need paid media buying
- You want a team on site or in the room
Objections
The reasonable worries.
- One person is more risk than an agency
- It is a different risk, and the honest answer is on the how we work page. An agency's continuity comes from bench depth; this comes from everything being documented in your repository, so the work does not stop and does not depend on us being reachable. Neither is risk-free.
- Agencies bring outside perspective
- Some do, and it is genuinely valuable. If that is what you are buying, buy it. This is production and operation, and it is a poor substitute for strategic work you actually need.
- We already have a good agency
- Then the useful question is not whether to replace them. It is whether you would still have a working content operation if they resigned the account tomorrow.
- This sounds like it needs technical people on our side
- It does not. The repository exists so the work is portable, not so you have to open it. If you never look at it, everything still functions.
Questions
Asked before the call.
Is this cheaper than an agency?
Not always, and cost is the wrong axis for the comparison. The difference is what exists at the end. Scope and figures come out of the call, and neither appears on this site.
Can this work alongside our agency?
Often the cleanest arrangement. They keep strategy and campaigns; recurring production runs on the specification. The agency page covers that arrangement from their side.
What if we want to bring it fully in-house later?
That is the intended end state for some clients. The handover package exists from the end of Build precisely so that is a decision rather than a project.
How is this different from a freelancer?
A freelancer produces work. This produces the specification and the pipeline first, then operates them. A freelancer working from your specification is a perfectly good arrangement, and some clients end up there.
Read next
Where this connects.
- Compared with hiring in-house is the other half of this decision for most companies.
- For agencies is the same comparison from the agency's side of the table.
- Build and Run sets out the phases, the scope drivers, and the notice terms.
Run the comparison on your own work.
Thirty minutes. We read what your current arrangement is producing and say what a specification would have to capture to reproduce the good parts.