Most companies evaluate a creative agency on the pitch, the portfolio, the chemistry of the first call. Almost nobody asks what the first 90 days actually look like day to day, which is the part that determines whether the engagement works. Here’s a realistic breakdown of what should happen in that window, and the specific points where projects quietly go off track.
Weeks 1 to 2: Discovery, and the first real warning sign
A serious engagement starts with structured discovery, not a kickoff call that jumps straight to mood boards. That means working sessions with leadership to surface positioning, competitive set, and what’s actually not working about the current brand or site, documented in writing, not just discussed on a call and left to memory.
The first real warning sign shows up here: if the agency is already sharing visual direction or logo concepts in week one, before the strategic groundwork is written down and agreed on, that’s a tell. It usually means the agency is designing against assumptions instead of research, and you’ll pay for it later when the direction has to be unwound.
Weeks 3 to 4: The strategic foundation gets locked
By the end of week four, you should have a written positioning document and, for a brand engagement, a messaging framework, not final copy, but the governing logic for how the company talks about itself to its actual buyers. This is the document every design decision gets measured against for the rest of the project.
This phase should include a real point of friction: a good strategist pushes back on at least one assumption you walked in with. If every strategy conversation in this window is pure agreement, nobody’s actually doing strategic work, they’re just reflecting back what you already believed.
Weeks 5 to 7: First creative direction, not first final design
This is where visual or design exploration actually starts, grounded in the strategy locked in weeks 3 and 4. A well-run process presents two or three genuinely distinct directions, not minor variations of one safe idea, each one defensible against the positioning document, not just “which one do you like.”
The useful question to ask in this window: can the agency explain why each direction exists, in terms of the strategy, not in terms of taste? If the answer is some version of “we just thought it looked good,” the strategic phase didn’t actually connect to the design phase, they ran in parallel instead of in sequence.
Weeks 8 to 10: Refinement, and where scope creep actually starts
Once a direction is selected, this phase builds it out: full identity system, or website architecture and page-level design, depending on scope. This is also where most scope creep begins, not from the agency, but from the client. New stakeholders get looped in, new opinions surface, and the brief starts expanding without anyone formally re-scoping it.
The fix isn’t to avoid feedback, it’s to make sure every round of revisions is happening against the fixed scope agreed on at kickoff, with a clear, written process for what happens when a request falls outside it. Agencies that let scope drift silently are being easy to work with in the short term and expensive in the long term.
Weeks 11 to 13: Build, QA, and the handoff that actually matters
For a website, this is development and quality assurance across devices and browsers. For a brand identity, this is finalizing the full guidelines document and production-ready files. The deliverable that gets skipped most often here, and causes the most pain six months later, is a real, usable brand or design system document, not a PDF that gets opened once and archived, but something your team, your printer, and your next vendor can actually reference without you re-explaining it.
A 90-day engagement that ends without this handoff done properly hasn’t actually finished, it’s just stopped. The work only compounds in value if someone besides the agency can pick it up and use it correctly a year later.
The pattern across all 90 days
Every phase above follows the same structure: strategy before design, a written document before visual work, and a clear point where you could stop and still have something usable. Engagements that skip steps to move faster usually end up slower overall, because the unresolved strategic questions resurface during design, or the unresolved design questions resurface during build, each time costing more to fix than it would have cost to settle upfront.
If you’re currently 90 days into an engagement and none of this sounds familiar, that’s worth naming directly with whoever’s running your account, not as a complaint, but as a specific question: which of these steps happened, and which got skipped?
