Ask to See the Worst Quarter, Not the Best Month
Every agency you talk to will show you the best month an account ever had. That month tells you almost nothing about the other eleven, which is what you are actually paying for.
The portfolio you’re shown is a highlight reel by design
A pitch deck or a case section on an agency’s site defaults to peak windows for a plain reason: those are the weeks where a product launched, a season landed, or the account caught a trend it did not have to invent. None of that requires the agency to have done anything dishonest. A launch gives a team something to build around, a holiday gives a reason to post something people already want to see, a trend supplies its own attention before a single caption is written.
Showing that work first is normal practice, not a red flag by itself. Nobody leads a pitch with a Tuesday in February. The problem is not that the best month gets shown, it is what happens if the reader mistakes it for the whole story.
Most of the calendar year has no tailwind
Strip out the launches, the holidays, and the moments an account got lucky enough to catch a trend, and what is left is most of the calendar. A generic quarter with no product news, no seasonal event, and nothing viral running through it is not the exception in a client relationship, it is closer to the rule: routine posts, no external reason to be interesting, no built in audience already paying attention.
That stretch, not the best month, is what a client is actually buying across a year. A retainer pays for the ordinary weeks between the launches, where there is nothing to hide behind. Judging an agency only on the weeks it had help from the calendar is judging the wrong thing.
What to actually ask for
- Ask for a full quarter of posts from the account’s public feed, not a curated set: three consecutive months, pulled straight from the profile in order.
- Specify that the quarter contains no product launch, no seasonal event, and no trending moment the account happened to catch. If any land inside the window, ask for a different quarter.
- Do not let the agency choose which quarter counts as unremarkable. Check a public launch calendar or major dates against the account’s posting history, and pick the window yourself.
- If the account is new or the history is thin, ask for the longest available quiet stretch rather than accepting a curated substitute.
What a genuinely unremarkable stretch looks like when the craft is real
Picture a mid sized retailer’s account in a plain March, nothing on the marketing calendar. A genuinely well run quiet quarter still has posts with a clear point: a product feature explained in one clean idea, a customer question turned into content, a small observation about the category that did not need a campaign concept to justify it. Filler exists because a slot needed filling and says the same nothing on any account.
The visual system holds without a concept propping it up: fonts, color use, photo treatment and layout stay consistent with no unifying campaign idea to lean on. Small human touches keep showing up too, a caption that answers the actual question a customer might have, a reply that responds to what was specifically said rather than a copy pasted thank you. Attention in the quiet weeks is the tell, since there is no spike in traffic rewarding the agency for giving it.
What a weak floor looks like, even from an agency with a strong best month
The same agency that produced an excellent launch campaign can still have a weak floor, and the quiet quarter is where it shows. Watch for posts that exist only to fill the calendar slot: generic quote graphics with no connection to the brand, reposted content with no point of view, captions that restate the image instead of adding to it.
Compare the visual quality against the campaign driven months in the same portfolio. A noticeable drop, thinner design, less considered photography, dashed off captions, signals the effort was reserved for moments that would get noticed. Watch too for a stretch where the account goes quiet or loses consistency because there was no external event to organize the work around. An agency that needs a launch to stay disciplined is telling you something about the rest of the year.
Why this matters more than it seems
A year of client work is mostly the quiet quarters. The floor, not the peak, is the more honest predictor of what the relationship will feel like month to month, because the peak happens a handful of times a year and the ordinary months make up the rest.
This is different from judging one weak post in isolation. Every account, even a well run one, will have a single post that does not land. The floor test is about a sustained stretch: does the quality hold for weeks with nothing to lean on. One bad post is noise. A whole flat quarter is a pattern, and why one excellent campaign tells you almost nothing works the same argument from the other direction: neither a single high point nor a single low point is the real measure, the sustained middle is. The clearest way to apply any of this is to look at one account directly rather than take a pitch deck’s word for it.
See this year’s selections for accounts reviewed against exactly this kind of stretch, or if you run or know an agency worth this kind of look, nominate an agency.
FAQ
How do I find an unremarkable quarter without the agency picking it for me?
Cross reference the account’s posting history against a public calendar of major dates and any known product or seasonal moments for that brand or category, then choose a three month window where none overlap. Doing this yourself keeps the choice out of the agency’s hands rather than accepting whichever quarter it offers as representative.
Is it fair to judge an agency by its quietest stretch?
Yes, because that stretch makes up the majority of any real client relationship, not an exception to it. A single excellent campaign shows what an agency can do under ideal conditions, a narrower and rarer question than what it does the rest of the time, the same reasoning behind why one excellent campaign tells you almost nothing on its own: a peak and a floor are different measurements, and the floor is the one that predicts the year.