DCO Templates Designed Against Your Feed's Longest String

A dynamic template fails on the ninety-fifth percentile string, not the average one. If your product titles average thirty-four characters and the tail runs past ninety, the layout has to be designed against ninety with a defined behaviour at that length. So this engagement starts with your actual feed export, sorted by character count, before any layout exists. Roughly a third of the time, that sort is what changes the brief. Feed first, layout second: I read the feed you have, not the schema you intended. Which columns drive a visual decision and which are decoration. Which rows are missing an image and what serves in their place. Whether the price column carries a currency symbol or a bare number, and whether the rand and dollar rows agree on the decimal separator. Whether discount is stored as a number or a string. Every one of these has broken a live dynamic campaign for somebody. Each is a five-minute conversation before the build and a full re-export after it. What the template actually contains: Rules, not only artwork. A badge that appears above a defined discount threshold and stays hidden below it. A title that runs two lines and drops to one when the price string is long. An image frame that handles both a transparent cut-out and a full-bleed lifestyle shot without the logo landing on a face. A fallback row that serves when the feed is stale or the endpoint is down, designed to be an acceptable ad rather than an apology. And a defined behaviour for the string that exceeds even the ninety-fifth percentile, because merchandising will eventually ship one. Where it runs: Built on the platform you already hold. Google Studio dynamic content driving DV360, Meta's catalogue and flexible formats, or Celtra, Bannerflow or Smartly if your team owns the seat. I do not sell you a platform and I take nothing from one. Where your existing stack cannot do what the brief asks, I will tell you which part of the brief to drop rather than which vendor to buy. How many axes: Fewer than the tool allows. Six headlines against four images against three calls to action is seventy-two combinations, and your reporting will attribute none of them with confidence at any realistic spend. I cap the axes at what the buy can separate and put the remaining effort into the layout every combination has to pass through, which is where the result comes from in any case. Not part of this: I do not own or maintain your feed. That belongs with the team that owns the catalogue, and a dynamic programme where nobody owns the feed decays into serving discontinued products inside a quarter. No product photography and no image processing pipeline. No media buying, seat management or bid strategy. No copywriting: the strings come from your catalogue and your team. Not for you: A catalogue small enough that authored creative is cheaper. Under a few hundred lines, or where the only variable is the headline, hand-built beats dynamic and I will say so on the first call rather than take the larger fee. A team with nobody able to keep a feed live. A brief that wants the template built before the feed exists, on the promise that the real feed will look roughly like the sample. It never does, and the rebuild costs more than the wait.