The most useful post in a portfolio is the weakest one they left in
Every agency portfolio shows the same three or four posts, and every one of them is good. That is not a coincidence. It is the entire point of a portfolio, and it means you have learned almost nothing about what the agency does most weeks. There is a faster way to find that out, and it does not involve the portfolio page.
The portfolio is a highlight reel by design, not by accident
A portfolio is chosen, not sampled. Someone looked at a much larger body of work and pulled out the pieces that made the agency look best. That is not a flaw in how portfolios get built, it is the job. But it means a portfolio cannot show the average, because selection is the one thing designed to avoid it.
The hero post proves something real: the agency is capable of this level of work under the best conditions it has had. A generous timeline, a client who signed off fast, an idea everyone was excited about. That is a legitimate claim, and a narrow one. It shows the ceiling exists, not how often the agency reaches it.
This is a different question from the general inventory of what a portfolio structurally omits, which is its own subject. This piece is about one specific technique: finding the post the agency did not bother to hide, and reading it as the baseline.
Where the weak post actually lives
Not on the portfolio page. On the live feed, because the feed contains everything, not just what got pulled into a case study or a portfolio tile.
Scroll past whatever is pinned or most recent, often still curated, and keep going until you reach a stretch with nothing attached to it. No campaign, no launch, no seasonal hook. An ordinary week is where the agency’s default behavior lives, because nobody was watching closely enough to make it exceptional.
You will know it is uncurated rather than an outlier by the shape of the day around it. Engagement should look ordinary for that account, not unusually low, and nothing suggests a special constraint like a holiday or a crisis. A quietly bad post during a crisis is not evidence of the baseline. A quietly ordinary post on a day when nothing was going on is exactly that.
What to look at once you have found it
- Whether the caption reads like it was written for this post, or dropped in from a template used across many posts. A caption fitting three unrelated photos was not written for any of them.
- Whether the visual still follows the account’s stated system, grid logic, color and typography, or whether that system quietly stops applying when nobody is watching closely.
- Whether the post has a point at all, a reason it exists beyond filling the calendar slot, versus being visibly there to keep the cadence going.
What a strong floor looks like versus a weak one
Picture two accounts, neither named, because the pattern matters more than the example. On the first, the ordinary post is smaller in ambition than the hero post. It was not built to impress anyone, but it keeps the same visual discipline as the rest of the grid, the caption still sounds written for this photo, and there is still a small, real reason the post exists.
On the second, the ordinary post breaks the visual system the agency claims to follow elsewhere. The caption is generic enough to run under any similar photo, the tell it was never written for this one. The post has no reason to exist beyond occupying a date on the calendar.
The useful measurement is not the hero post at all, but the gap between the hero post and the ordinary one. A small gap is a better sign than a spectacular hero post sitting far above a mediocre baseline. A wide gap means the ceiling was a special occasion, not a standard.
The failure mode of judging by the hero post alone
A reader who stops at the portfolio has seen the ceiling and nothing else, so no information about consistency. Consistency, not the best single thing an agency has made, is what determines what an account actually looks like most weeks, because most weeks are ordinary.
The incentive here is not symmetrical. The agency has every reason to make its ceiling visible, in a pitch, a case study, the portfolio page. It has no equivalent reason to volunteer its floor. The floor has to be found, not offered.
None of this is about whether to sign a contract or bring an agency on, a different question entirely. This is only about whether the work is genuinely good, and an ordinary post the agency did not think to hide is some of the clearest evidence available for that.
Try it yourself
You do not need a portfolio to run this, only a live feed to scroll. See what a portfolio leaves out for the structural side, one excellent campaign does not tell you much on small samples, and whether an agency built a system or just one good post for the checklist step. Or nominate an agency whose ordinary weeks hold up.
A short version of the method
Three steps, in order: find an uncurated week, isolate the least remarkable post in it, check that post against the account’s own stated system, not against the hero post.
The target is not the worst post that could ever exist, which may be an anomaly. It is the ordinary post the agency did not bother to hide, the baseline, and the baseline is what a client actually receives most weeks.
FAQ
Is it fair to judge an agency by its worst post?
Not quite the method. The single worst post risks being an anomaly, a bad day that says little on its own. The useful target is the ordinary, uncurated post from a normal week, a fair read of the baseline because it is what the account looks like most weeks.
What if an agency’s feed has no obviously weak posts at all?
That is informative on its own. A small gap between the best post and the ordinary post signals more consistent craft than a portfolio full of highlights sitting on top of an uneven feed.
Where should I look if the account only has a handful of posts to judge from?
Be honest about the limit. This method needs a body of ordinary output, so a new or sparse account cannot be judged this way yet. That absence of a track record is worth noting, not filled in with a guess.