On the structural choice behind the agency, and why slow craft beats glossy pitch decks.
Here is what I learned watching performance creative agencies fail their clients for five years before I started Maple Media. The work that actually wins isn’t pretty. It’s structured.
Most agencies sell you the pretty version. The pitch deck shows Cannes Lion winners, lifestyle campaigns, beautifully produced hero pieces. Six months later you are staring at a pipeline that produces four polished ads a month, two of which never see meaningful spend, and you cannot figure out why your CAC keeps climbing.
The disconnect is not talent. It’s how the agency is built.
Traditional performance creative shops still operate like ad agencies from a previous era. The creatives sit in one room and defend concept integrity. The media buyers sit in another room and defend spend efficiency. The two teams meet on a weekly status call to disagree politely and produce work that satisfies neither metric particularly well.
One pod, one number
Maple Media collapsed those silos on day one. Each pod has a creative strategist, an editor, and a media buyer working off the same weekly numbers. Same Slack channel. Same planning doc. Same definition of winning. A pod doesn’t produce ads. It produces profitable ads, then iterates on the winners until the algorithm stops responding.
That single structural choice shaped almost everything else.
Velocity beats volume
Volume is the agency model that promises forty creatives a month and ships thirty-eight mediocre ones. Velocity is producing fewer concepts and iterating on them faster. Killing losers without sentiment. Pouring variations into winners while the algorithm still rewards them.
Three concepts iterated eight times will almost always outperform thirty concepts shipped once. That has held across every client Maple Media has worked with, regardless of category.
Voice-of-customer or nothing
Every concept starts in customer research. Not workshop personas. Not internal assumptions about who the buyer might be. Actual Amazon reviews, Reddit threads, YouTube comments, support transcripts, post-purchase surveys. The exact phrases buyers use when they are being honest about the problem. Then we mirror that language back in the hook.
It’s slow. It’s unglamorous. Most agency teams skip it because sitting in a doc with two hundred customer reviews doesn’t feel like creative work. The slowness is exactly why it works. The agencies skipping it are the same agencies producing ads that read like brand statements and convert like cold direct mail.
Compliance built into the script, not bolted on
A meaningful share of our work sits in health, wellness, and supplements, where one wrong claim can flag an entire ad account. We bake compliance into the script stage rather than the QC stage. Our strategists know the language constraints before they are three hours into a draft. Our editors know which footage survives platform review. The first draft slows by ten minutes. The next three drafts speed up by hours.
Building a memory
Five years from now, I want Maple Media known for the kind of work that compounds. Avatar research from Q1 informing the angle library in Q2, feeding the hook bank in Q3, hardening into the full creative system by year-end. Most agencies start fresh every quarter. We are building a memory.
None of what I have described is glamorous. There is no agency-of-the-year award for predictable returns over twenty-four months. But if you are a DTC founder, you don’t actually need a viral ad. You need a creative engine that pays for itself, then funds the next round, then funds the round after that.
Compound that for two years and you have a category leader. That is the agency we are building. Slowly, deliberately, and with the stubbornness that comes from watching the alternative fail too many times.
