What separates a written caption from a filled in one
You are three screens into an agency’s portfolio account and something in the writing is bothering you. The images are clean, the edits are tight, and every caption reads like it was assembled from the same parts. Asked to name the problem, you would say the copy is generic. That is not a judgment, it is a shrug. Genericness leaves marks on a caption, the same way a bad edit does.
The half of the post nobody is trained to read
Most people reviewing social work have vocabulary for the visual. They can say the composition is static, the grade fights the product, the cut lands a beat late. Copy gets none of that scrutiny. It is read on autopilot, absorbed as tone rather than examined as craft.
On a portfolio review that is a mistake. Caption text is the most checkable thing on the post, there in full. You can read the words and ask whether anyone made a decision. Apply the standard behind what actually separates good social media work from average and accounts that look strong at a glance come apart.
What a template sounds like once the blanks are filled in
Filler copy fails in a few repeatable ways, each visible in the text itself.
- It is interchangeable. Swap the product name for a competitor’s and the line stays true and publishable. Nothing anchors it to this post.
- Stock connectives do the work. “Designed with you in mind”, “because you deserve it”. They sit where a specific observation belongs.
- Abstract nouns stand in for details. Quality, craftsmanship, our process. Each is a placeholder for something nameable: the material, the stitch, the hour it was shot.
- The call to action is the only working sentence. If everything above “shop now” could be deleted without changing what the post communicates, the caption is a button with a preamble.
Generic copy against specific copy, dimension by dimension
| Dimension | Generic looks like | Specific looks like |
|---|---|---|
| Word choice | Category words any rival could use that day. | A word that would be wrong under anything else. |
| Sentence rhythm | The same length every post, whatever the subject. | Length varies, and the variation tracks the content. |
| Specificity of detail | Evaluative adjectives. Claims, not facts. | A name, a texture, a time, a concession. Something checkable. |
| What the CTA is doing | Bolted on, in a different voice from the lines above. | Continuous. It follows from what was just said. |
Reading two captions the same way you’d read two frames
Both are invented for this piece. Picture the same photograph above each, a canvas tote on a counter.
The bag that goes everywhere with you. Designed with you in mind, built to last, and ready for whatever the day brings. Which colour are you going for? Link in bio.
The handles are stitched twice because the first run tore at the seam. It holds a laptop and a loaf of bread, which is roughly what we tested it with. It creases. That is canvas.
“Stitched twice” is a manufacturing fact, not an adjective. “The first run tore at the seam” is an admission, and admissions cannot be templated because a template has nothing to admit. “A laptop and a loaf of bread” replaces the abstract noun capacity with two objects you can picture. “It creases. That is canvas.” pre-empts a complaint instead of denying it. The first caption is not bad for being short, it is barely shorter than the second. It is bad because no clause in it could only apply to this bag.
Rhythm is a decision, not a byproduct
A caption can be full of concrete nouns and still read as machine assembled if every sentence carries the same weight. Deliberate rhythm is audible: a longer sentence holding the detail, then a short one or a fragment landing right after it to change pace. Scroll a grid and count clauses. If every post is two sentences and a question whether it announces a product or marks a closure, nobody was calling pace. The same reading applies to what the replies under a post reveal.
The interchangeability test, stated as a single question
If you swap out the product or account name and the caption still reads as true and complete, it was filled in rather than written.
The test only works across several posts on one account. Any single caption can be legitimately generic, because some posts have nothing specific to say. That is reading strategy from what an agency actually published, at the level of the sentence.
Where this way of reading breaks down
Restraint is not emptiness. A deliberately spare account can run captions of a few words and still be written, because the words chosen are unusual or precise in a way that holds from post to post. A considered short caption contains a choice you can point at. An empty one takes the flattest phrasing available.
One excellent caption establishes nothing, the same way one strong campaign does not. Craft is a pattern held across a grid, under posts nobody found interesting to write.
Read a few for yourself
See this year’s selections and run the swap test on the captions rather than the visuals, then do the same on an account you are unsure about. If an agency’s writing survives that reading, nominate an agency.
FAQ
Isn’t a short caption just efficient rather than generic?
Length and genericness are separate measurements. A short caption can be highly specific if those words could not sit under any other post, and a long one can be padding. Swap the account name and see whether it still stands. Efficiency says one exact thing briefly. Genericness says nothing only it could say.
Can a caption use slang or a trending phrase and still be well written?
Yes. Register is not the marker. What matters is whether the phrase fits this exact post, or was dropped in because it was available that week. A borrowed phrase bent to say something only this post could say is writing. The same phrase left as it arrived is a blank being filled. Whether the account should have joined the trend at all is a separate judgment.
Does the caption matter if the visual is doing the work?
It matters as a separate signal. On a visual led platform the image can carry the post entirely, so filler copy underneath costs very little there. When strong visual craft sits above copy nobody bothered to write, you are seeing where attention stopped inside the team.
