The template you've already built — your brand kit's fonts and colors applied, your layout settled — is already most of the way to being automatable. The gap between "a template" and "an automated template" is small: name what should change, then wire something to fill it in.
Your existing template already has what it needs
If you built it with your brand kit — logo, fonts, colors — applied, the design half of automation is already done. Nothing about automating a template requires rebuilding it differently; the same template that looks right filled in by hand looks right filled in by a workflow, because it's the identical design either way.
Step 1: identify what actually changes between posts
Look at a few real instances of this format you've made by hand, and note what's genuinely different each time versus what stays fixed. A weekly recap's headline changes; its layout, colors, and logo placement don't. Only the changing parts need placeholders — everything fixed stays exactly as designed.
Step 2: turn those into named placeholders
Type a curly-brace token directly into each text box that should change — {headline}, {stat}, {takeaway} — right in the design, the same way you'd type any other text. This is the entire authoring step; see how template placeholders work for the full mechanics of what happens next.
Step 3: build the automation that feeds it
An Apply Template node pointed at your template automatically finds every placeholder you just added and exposes it as something to bind — a data source (an HTTP Request node, a Custom Agent's output, or both merged) feeds each one. This is the shortest real automation chain: source → Apply Template → Post; see building a workflow from scratch if this is your first one.
What doesn't need to change
Everything about the template's actual design — fonts, colors, logo, layout, spacing — stays exactly as you built it. Automating a template isn't a redesign; it's adding a data-filling step on top of a design that already works. If a template needs a redesign, do that first, by hand, before wiring any automation to it — automating a design you're not happy with just produces the wrong thing faster.
When one template isn't enough
If you find yourself wanting to automate several genuinely different formats, each deserves its own template rather than one template trying to flex into multiple shapes via conditional logic — a weekly recap and a news reaction are different designs, not variations of one. See multi-template fan-out for wiring one data source into several templates when the underlying data genuinely supports more than one format.
The payoff
Once a template is wired this way, "making a post" stops being a design task at all for that format — it's a data-filling task, done automatically, with the design quality of your original template preserved on every single run. That consistency — post #200 looking exactly as considered as post #1 — is the actual value of brand template automation, more than the raw time saved.