What we refuse
to ship
Most design agencies define themselves by what they can do. We define ourselves by what we won't.
Standards aren't a feature of execution. They're a filter - the boundary between work worth shipping and work that exists only to mark a task as "done." The more you refuse, the stronger the output.
After fifteen years of working with multi-location brands, we have our own list. Six things we simply don't deliver, regardless of pressure, deadline, or client request. Each refusal has an operational reason.
01 - Briefs without decision ownership
Signal: "We have 5 people involved in this project. All must approve."
A brief without a clear decision-owner isn't a brief - it's a negotiation. Five people with veto rights guarantees five rounds, each pulling in a different direction. Result: production stalls, time slips, brand tracker drifts.
Before starting any workstream, we require one thing: the name of the person who says yes or no. Everyone else - input, not veto. Without it, we don't start. Clients sometimes push back. After the first delivery, they always understand.
02 - Aesthetics without operational context
Signal: "Make us something fresh, original, with wow factor."
A brief focused only on visual effect, without specifics about channel, location, brand system integration, or production capital - is a recipe for "beautiful but useless." A designer can make a stunning layout that breaks the first time it's translated to another format.
Every project starts with context: channel, audience, constraints, integration with the existing system. Aesthetics are the last layer, not the first.
- ● Channels: exact list, not "social media."
- ● Format constraints: real specs, not "responsive."
- ● Integration: what templates exist, what cannot be touched.
03 - Deliveries without QA
Signal: "Urgent, can you send it today at 4pm?"
Work without quality control isn't work - it's hope. Every file that leaves goes through a 5-7 point checklist: spec compliance, brand consistency, copywriting, technical export, accessibility, naming convention, version control.
If QA shows an issue - we return to production. Always. "We'll ship and fix later" are words that have cost too many brands. You won't fix it later. The version that goes into circulation stays for good.
Standards are filters,
not aspirations.
What you refuse is what you are.
04 - Revisions "just a little"
Signal: "Can you change just one tiny detail? Just this color."
Small revisions don't exist. Every change after sign-off means: reopen file, reload context, new QA pass, new export, new version control, new distribution. Aggregate 40-90 minutes per "tiny thing."
We have a rule: after sign-off, changes cost. Not to earn - to make price a signal. Client starts thinking before requesting. The number of post-sign-off revisions drops 80% in the first three months of partnership.
- ● Sign-off is final. What changes after goes into new scope.
- ● 3 rounds of revision in production = enough. Fourth = brief problem, not execution.
- ● Every change has an owner, deadline, and documented rationale.
05 - Work without system integration
Signal: "Just make us one piece, one-off event, not tied to the brand."
Every one-off creates a precedent. "Just this once" becomes a reference point six months later: "you made us this unorthodox thing already, now do something similar." Brand drifts through the accumulation of "exceptions."
We only work on things that integrate with the client's brand system. Either we strengthen the existing system or expand it with a new branch (channel, format, sub-brand). We never make orphan assets.
06 - Work in urgency mode as default
Signal: "Everything is urgent. No time for planning."
Urgent as default is a transparent signal: the sign of an organization operating in reactive, not proactive mode. We work in a steady weekly cadence - briefs planned 2 weeks ahead, output on Friday, retro on Monday.
One "urgent" sprint per quarter we accept - requires justification and agreement to deprioritize another workstream. Four urgent ones per quarter - we refuse. Which means: maybe the client isn't right for us, or we for them. Better to recognize that early than later.
Each of these refusals has its place in our 8-step methodology r3loop. Standards aren't opinion - they're part of the governance layer (step 6), brief layer (step 1), and QA layer (step 7). The client who wants to work with us is buying the system that enforces these standards.
What refusal buys
Every "no" builds a "yes." The more clearly we define the boundary, the easier it is for the client to understand what we deliver. Standards don't reduce scope - they increase the value of work within the boundary.
A client who accepts these six refusals gets a partner. A client who doesn't accept them gets a more suitable agency for their work style - and doesn't lose a quarter to operational conflict.
Standards defend
quality.
Saying "yes" to everything
doesn't build it.
If you work in a multi-location organization where operational chaos eats energy - a system that says "no" at the right moments costs less than an endless wave of revisions.
