The Democratic Party
A national campaign already has a brand. It has a color palette, a logo with a very famous E in it, a typeface, and a style guide that thousands of people are working from simultaneously. What it doesn't have is a version of itself that a specific person, in a specific community, scrolling a specific site at eleven at night, feels was made for them.
That gap is the work. Twice, for the Democratic Party, across the 2020 and 2024 elections, through Bravely.
It was familiar ground. I started my design career on political work, at a small agency in Jacksonville, building full campaigns for city council seats, judicial races, school board, and the mayor's race. Local politics is a good teacher. Budgets are small, turnarounds are short, and the audience is close enough that you hear about it directly when the work misses. Years later the scale changed and the question didn't.
The brief both times was some version of: reach young, urban, digital audiences. That sounds like a media problem and it is partly a media problem, but underneath it is a design problem, and the design problem is harder.
A national campaign speaks nationally, by necessity. It has to be legible to everyone, which means it can't be specific to anyone. Run that creative unchanged at a community that has heard national political messaging its whole life and been let down by a good deal of it, and the work reads as what it is: a thing made for a general audience, pointed at you.
So the question underneath every asset was not what does the campaign want to say. It was who is standing here, what have they already heard, and what would make this feel addressed rather than broadcast.
Before any layout, the question was who we were actually talking to and where they'd encounter this. A demographic isn't a targeting parameter, it's a set of people with a history of being marketed at. The visual language had to earn a second of attention from someone with every reason to scroll past.
Each cycle got a visual system customized for the community it was speaking to, working inside the campaign's brand rather than around it. The distinction that matters: not a national ad resized, but a coherent language that happens to share DNA with the national one.
Individual graphics were designed to work together, so a site takeover, a banner, and a social unit encountered in sequence read as one presence rather than three separate asks. Political digital is a volume game with a lot of surfaces. Cohesion is what stops volume from becoming noise.
The 2020 work landed against the Kamala Harris VP announcement, which is not a date you negotiate. Video and design had to be built, approved, and live while the moment was still the moment. Campaign work is the only place I've done where being right next week is the same as being wrong.
A national campaign has to be legible to everyone, which is the same thing that stops it feeling made for anyone.
The brand arrives finished: the palette, the typeface, the logo with a very famous E in it. What it cannot arrive with is the sense that it was built for the specific person scrolling at eleven at night. Closing that distance was the assignment, in 2020 and again in 2024.
A four-month engagement timed to Kamala Harris joining the ticket. Website takeovers, digital ad sets built out across every standard unit size, site theming, and a short-form video against current events. Everything carried the campaign's official disclaimer, which is its own constraint: there is no such thing as a casual asset when it says paid for by at the bottom.
Top row square, bottom row vertical, and two of these are the same message rebuilt, not resized. That's the constraint that shapes political digital: a media buy wants every standard unit, and the argument has to survive each one intact.
Four years later, the same client contact came back with the same shape of brief, pointed at a new candidate. That is the part worth stating plainly, because it is the outcome: the 2020 engagement is the reason the 2024 one happened. Campaigns assemble their bench fresh every cycle and there is nothing automatic about being called twice, so a second phone call from the same person is about as direct a verdict on the first round as this kind of work ever gets.
The scope tracked closely. Site takeovers again, across the same kind of hip, urban-leaning properties. Units built for the same communities, under the same disclaimer discipline. What moved was the center of gravity: 2024 leaned further into video. Four years had changed where these audiences were and what actually stopped them mid-scroll, and a static unit that carried its weight in 2020 was carrying less of it by 2024. Same question as before, answered in a format the room had shifted toward.
Not just the units, the placement. The cost-of-living creative shown in context on one of the hip, urban-leaning properties these ads were built for. The takeover ran a video billboard, an in-article unit, and a sticky anchor off the same set.
The 2024 cost-of-living argument, built across formats for the same hip, urban-leaning properties. The center of gravity moved toward video this cycle; these static units carried the message everywhere else.
One argument, every standard unit size. The same message rebuilt to survive a 900x500, a 970x250 billboard, and a 320x50 leaderboard, each carrying the Paid for by Harris for President disclaimer.
Presidential cycles, four years apart.
Ad sets, site takeovers, theming, and video, designed for specific communities rather than resized for placements.
The same message had to land differently depending on who was reading it, so the visual language was defined per audience and then held across every unit size the buy demanded.
That specificity is a form of respect, and audiences can tell instantly when it's missing.
The failure mode in this category isn't ugly work. It's competent work that was clearly made for everyone and then aimed at you. People have a very fast, very accurate detector for that, and once it fires you've lost them for the whole cycle, not just the impression. The whole job was building something that survives that detector.
The other lesson was about deadlines that don't move. A conference date can slip. A VP announcement cannot. Working against a news cycle taught me more about what's genuinely essential in a design process than any amount of unhurried work, because you find out very quickly which steps were load-bearing and which were habit.
ADCOLOR
Lead digital designer since 2021. One voice carried from a social post to a projected animation on a venue wall.
Read the case study →
Johnson & Johnson Vision
They sell lenses to your optometrist, not to you. Two audiences, two arguments, one brand.
Read the case study →