Most digital strategies for campaigns get built in the wrong order. Channels first, audience second, message last. That's why so many of them feel generic. Here's the order we use, and why each step has to come before the next one.
1. Define the Audience Before You Touch a Platform
The single biggest mistake we see is starting with "what platforms should we use." That's the last question, not the first. The first question is who you actually need to reach.
Start with the voter file. Build three universes:
- Base: your high-propensity supporters. They'll vote, but they need reinforcement and turnout reminders.
- Persuadable: mid-propensity, mixed history. This is where the race is decided.
- Soft supporters: voters who lean your direction but don't always show up. The GOTV target.
For each universe, know the count, the geography, the rough age distribution, and the issues that polled best with them. If you don't have polling, use precinct-level past results and field intel.
2. Build the Message Before You Build Creative
Message comes from the audience, not from the candidate's preferences. For each universe, you should be able to answer in one sentence:
- What does this voter currently believe about the candidate (or not believe)?
- What do they need to believe to vote for the candidate?
- What's the bridge between the two?
The bridge is the message. Everything you build (video, static, mail, doors) should ladder back to it. If you can't write the message in one sentence, you don't have one yet. Don't move to creative until you do.
3. Pick the Channel Mix Based on the Audience
Now you can pick channels. Not before. The mix should follow from where the audience actually is.
- Older base voters: Facebook, CTV, direct mail.
- Persuadables under 50: Meta (Reels and feed), YouTube, programmatic CTV.
- Younger soft supporters: Reels, TikTok if available, YouTube Shorts, peer-to-peer text.
Budget tier matters too. Under about $11K total digital, two platforms (Meta plus Google) is enough, spreading thinner kills learning velocity. Above $11K, three platforms (Meta, Google, programmatic) becomes the default.
Then phase the budget backward from Election Day. Foundation 18%, persuasion 37%, GOTV 45%. Front-loading is the most common mistake; it leaves you out of money during the only weeks that matter.
4. Build Creative That Fits the Placement
One creative idea, multiple cuts. Build the right ratio for each placement. Vertical first, square second, horizontal only when you need it. Captions burned in. Disclaimer inside the safe zone. No auto-cropped reuse.
For each phase, keep two to three concepts in rotation and refresh them every two to three weeks. If a concept underperforms by week two, kill it. The temptation to "let it run because the candidate likes it" wastes real money.
5. Measure What Actually Matters
Click-through rate is a vanity metric in political. Watch the metrics that map to outcomes:
- Reach into the persuasion universe: what percent of your target audience has actually seen the ad, and how many times
- Completion rate on video: for CTV and YouTube, anything under 85% means the creative isn't holding
- Brand search volume: Google Trends and search query data on the candidate's name. Real campaigns move it.
- Cost per identified supporter: for campaigns running ID programs, the only cost metric that matters
- Early-vote returns by precinct: the closest thing to a real-time outcome signal
Build a weekly check-in. Look at the data, decide what to cut, decide where to shift budget. Don't sit on a plan that isn't working because you wrote it down.
Bottom Line
A digital strategy is just five questions in the right order. Who are we talking to. What do we need them to believe. Where do we reach them. What do we say. How do we know it's working. Get those right, in that order, and the rest is execution.