When an agency’s different clients all look the same

You open an agency’s portfolio and every piece in it looks good. But one strong post cannot tell you whether the agency builds work around a brand, because a strong post can also come from a strong template that landed on a suitable client. What you need to know is whether this agency thinks about each account or reaches for the same system every time. Two clients, side by side, will tell you that.

The comparison test portfolios don’t invite you to run

A portfolio is assembled one client at a time, each represented by its best moment. The peak of account A sits next to the peak of account B, which is exactly the arrangement in which a reused system becomes invisible. Each piece is asked to prove it is good, never that it was made for this brand rather than assembled from what was already lying around.

So change the question. Not whether a post is good, but whether this same post, structurally, could have been made for a completely different brand with another logo and another noun in the headline. You need nothing from the agency to run it. Two public client accounts, publicly attributed to the same agency, are the whole requirement, and agencies usually publish that attribution themselves on a roster or case study page.

Set up the comparison

Pick two clients from clearly different categories. Say the roster holds a professional services firm selling a long, cautious, high consideration purchase, and a consumer lifestyle brand selling something cheap and impulsive. If the work is built for each, the execution should look and sound close to unrelated, so the further apart you pick, the louder any resemblance becomes.

Open the last ten to fifteen posts from each account in two windows and read them as two bodies of work, not two highlights. One post from each proves nothing. Fifteen show habits. Hold three things constant. Composition and template: grid layout, colour blocking, crop ratios, type treatment. Caption cadence: how the first line is built, how long the sentences run, where the call to action sits. Trend response: which formats each account reached for, how recently, and whether the format arrived adapted or lifted intact.

What a reused visual system looks like on screen

• The same crop ratio, colour grade, or filter on two brands whose visual identities have nothing in common. A warm, lifted, low contrast grade on both a law firm and a snack brand is a preset, not a choice.
• The same title card or carousel cover layout on both accounts, with only the logo and the headline swapped.
• The same carousel skeleton every time. Hook slide, three point list slide, call to action slide, whatever the brand actually sells.
• The same icon set, overlay style, and transition on every short form cut. Transitions are cheap to vary, and nobody varies what they are not thinking about.

What a reused caption cadence sounds like

Voice survives reskinning, which makes captions the more reliable tell. Listen for shared skeleton, not shared quality. The same opening move turns up on both accounts: a question opener, a confession style first line, a flat one word sentence. Then the same rhythm underneath, three short sentences and a question, whether or not the brand has reason to talk that way. Then the same call to action phrasing, word for word.

Distinct work fails the swap test immediately. Move a caption from one account onto the other and it reads as obviously wrong. A stodgy B2B account and a playful consumer account should read like two different people wrote them, even when the same team did. Our reading of what separates a written caption from a filled in one goes further into that.

Consistent quality is not the same as a flattened voice

Be careful what you penalise. A recognisable production bar across a roster is a good sign. Tight editing, competent colour, and a trend pipeline fast enough to ship a format while it still means something are capabilities, and an agency with them shows them everywhere. The problem named here is narrower: a brand’s voice and audience set aside in favour of whatever the agency already had built.

So do not ask whether every account looks polished. Ask whether this one looks and sounds like it was designed around this specific brand, or could have been produced for almost any other client on the roster with a search and replace.

The same line separates good trend work from lazy trend work. An agency that adapts a format to each brand’s voice is doing craft under time pressure. One that drops the identical format onto several accounts in the same week regardless of fit is doing distribution. That boundary is where using a format ends and copying an idea begins.

A short checklist for your own browsing

1. Pick two clients from different categories on the agency’s public roster or case study page.
2. Open the last ten to fifteen posts from each account side by side.
3. Note whether the visual template, the caption opener, and the call to action phrasing repeat across both, despite the brands having different audiences and different reasons to sound different.
4. Check whether the same trend format appears on both accounts in the same week, and if so, whether it was adapted to each brand or dropped in unchanged.
5. If most answers point to reuse, weigh that above any showcase piece: it describes the process, not the best day. That gap is what a portfolio leaves out.

Try it on an agency you are assessing

Run the comparison next time a roster page impresses you. For the agencies chosen in this year’s edition, see this year’s selections. If you know an agency whose accounts genuinely do not resemble each other, nominate an agency.

FAQ

Is a consistent visual style across an agency’s clients always a red flag?

No. A shared production standard is reasonable, and a consistent bar of craft reads as competence rather than laziness. The problem is the other kind of sameness: a voice and a template copied across brands with different audiences and different reasons to sound different. Judge the fit to each brand, not the polish.

How many clients do I need to compare to spot this pattern?

Two is usually enough, as long as they come from clearly different categories, for example a B2B brand and a consumer lifestyle brand. The more different the brands are, the less their execution should resemble each other if the work was built around each one.