What does a copywriting agency actually deliver?
A copywriting agency delivers written product content as a project: a batch of descriptions, sometimes titles, occasionally category copy, briefed by the store and delivered on a date. Good agencies bring craft and a voice the store may not have found on its own. The deliverable is text; what happens to it in the catalog, the feed and the schema afterwards is the store's job.
The scope stops at the batch. New SKUs after delivery, supplier data that changes, feed disapprovals, duplicated copy that reappears on an import: none of these are the agency's, unless the store buys another project.
What does a Product Catalog FTE deliver instead?
A Product Catalog FTE on ftes.ai delivers ownership of the catalog as a job function. It audits every SKU daily, proposes rewrites from the store's own product data with a claim-to-field map, fixes feed disapprovals at the source, completes schema and alt text, and reports completeness and search movement it measured. It is reviewed on targets, and every change waits in Approvals until the owner approves it.
The scope does not stop. A SKU added tomorrow is audited tomorrow. A supplier sheet that changes a weight produces a proposed correction. The function runs every day, which is the difference between a deliverable and an employee.
Where does the agency win?
The agency wins where the work is invention rather than application. A store that has no brand voice yet, a launch collection that needs copy with real craft, or a flagship product page that deserves a writer's hours: those are agency jobs, and a Digital FTE does not replace them. The Product Catalog FTE writes in a voice it was given; it does not create one from nothing.
The agency also wins on the small catalog. A store with forty SKUs that change twice a year does not need a daily function; it needs a good batch and a checklist.
Where does the Product Catalog FTE win?
The Product Catalog FTE wins on scale, accuracy and continuity. On scale: thousands of SKUs cannot be written by hand at a price a store can pay, and they change constantly. On accuracy: the fact-gate refuses to write a claim that no field or sheet supports, which is a stronger guarantee than a brief, because a writer under deadline will fill a gap. On continuity: feed disapprovals, schema and alt text are checked daily, which no project does.
It also wins on the record. Every change is logged with who approved it and what evidence was attached, so a store can answer why a page says what it says a year later.
How do the costs compare honestly?
The costs compare on different units, and an honest comparison says so. An agency is priced per deliverable, per word or per project, and the store pays again for each new batch. A Product Catalog FTE on ftes.ai is metered in tokens from one organization wallet shared by every role, with a balance always visible and a hard stop at zero. The store pays for work done, not for a retainer, and can see the spend per rewrite.
Which is cheaper depends on catalog size and change rate. For a large, changing catalog the daily function costs less per SKU kept complete; for a small, static one the batch does. Neither vendor should claim otherwise without the store's numbers.
What is the strongest pattern for a store?
The strongest pattern for most stores is both, in sequence. The agency, or the store's own writer, defines the voice and writes the flagship pages that deserve craft. That copy becomes approved copy in the Product Catalog FTE's remit, one of the three sources it may draw from. The FTE then applies the voice across the catalog, keeps every SKU complete and feed-clean, and flags the pages that deserve a writer's attention next.
This pattern keeps judgment with people and application with the Digital FTE, which is the division the whole ftes.ai model is built on.