Campaign Social Media Toolkits and Why Volunteers Skip Them
A congressional campaign I talked to in October had a 41 megabyte folder on Google Drive called Supporter Toolkit. It held 60 graphics in nine sizes, three cover photos, a set of profile frames, and a PDF of suggested captions.

A congressional campaign I talked to in October had a 41 megabyte folder on Google Drive called Supporter Toolkit. It held 60 graphics in nine sizes, three cover photos, a set of profile frames, and a PDF of suggested captions. Six hundred people had the link. In the following two weeks, the campaign's monitoring tool found 23 posts using any asset from that folder, and 11 of those came from paid staff. The communications director blamed apathy. I think she built the wrong object.
I have been watching this pattern since 2008, when the thing you mailed supporters was a web banner and a signature file for their email. The format changed, the failure did not. Campaigns build toolkits for designers and then hand them to people standing in a supermarket queue with 40 seconds and a cracked phone screen. A social media toolkit that volunteers use looks almost nothing like the one that looks good in a staff meeting.
What volunteers post when nobody gives them anything
Before you fix a toolkit, look at what supporters produce on their own. In the races I have followed closely, the unprompted posts fall into four shapes, and only one of them involves a graphic.
The most common is a photo of the candidate taken by the volunteer, usually badly lit, usually at an event, with a caption about why they showed up. The second is a screenshot of a news story with two sentences of commentary. The third is a reply or a quote post arguing with someone, which is the most engaged-with and the least controllable thing in the whole ecosystem. The fourth, far behind, is a shared campaign asset.
That order should tell you where the effort goes. A supporter's own photo outperforms your designed graphic because the algorithm on every major platform treats original media from a personal account more generously than a reshared branded image, and because the volunteer's aunt can tell the difference. A campaign in Michigan ran an unscientific test last spring: the same message as a designed square and as a staffer's phone photo, posted by 14 volunteers split between the two. The phone photos got roughly four times the comments. The sample is tiny and the campaign never wrote it up, so treat it as a hint rather than a finding.
Why the polished folder loses
Three frictions kill a toolkit, and all three are logistical rather than motivational.
The first is the download. Opening a shared drive folder on a phone, finding the right size, downloading it, then switching apps to post it is seven or eight taps and two app launches. People give up at tap four. The campaigns that get traction send assets inside the app where the volunteer already is, as a message in a group chat or a pinned post in a server, already cropped for the one platform that matters to that person.
The second is the caption problem. If you supply a caption, most volunteers either paste it word for word, which produces 30 identical posts that platforms demote as coordinated content, or they read it, decide it does not sound like them, and close the tab. Neither outcome is what you wanted. The fix is to supply the facts and let them write the sentence: a figure, a date, a link, and a line saying what the post is for.
The third is permission anxiety. Volunteers are more worried than staff realise about posting something slightly wrong. A toolkit that does not say plainly what is approved, what is off limits, and who to ask will produce silence from exactly the careful people you want posting. One line solves it. The person running the campaign's accounts should be named, with a direct way to reach them, and the answer should arrive the same day.
Build it in the order a volunteer uses it
Here is the sequence I would use for a race below the statewide level, where one person is doing all of this alongside other work. It takes a day, not a week.
- Pick two platforms, not six. For most local races that is Facebook and one of Instagram or the short video app your volunteers already use. A toolkit covering six platforms is a toolkit covering none of them well.
- Write the one page brief first. Candidate name spelled the way it goes on the ballot, the three message points, the two things never to post about, the disclaimer text if your state requires one on supporter content, and the name and phone of the person who answers questions.
- Shoot original media on a phone. Thirty vertical photos and six clips under 20 seconds, taken at real events, beats a designer's render. Keep the raw files, not just the exported ones.
- Produce assets in exactly two sizes, vertical and square, and put the text inside the image only where it is unavoidable. Text baked into a graphic cannot be corrected after 200 people have it.
- Deliver through a channel, not a folder. A broadcast channel, a group chat, or a volunteer server with its own assets room all work, because the asset lands next to the conversation about using it.
- Give every drop a reason and a window. "Here is the clip from Tuesday's forum, useful through Sunday" gets used. "Here are our evergreen graphics" does not.
- Ask for a reply, not a repost. The highest value thing a supporter does is answer a neighbour's question in a local group with their own name attached, and that needs facts rather than images.
Nothing in that list requires a budget over a few hundred dollars. The 2008 campaigns that got this right were working with worse tools and a slower web, and their advantage was that they treated supporters as people who would speak rather than as distribution for finished messages.
The day of action is usually a mistake
The standard advice says to coordinate: pick a day, give everyone the same graphic and the same hashtag, and post together to manufacture visible momentum. I have come to think that advice is backwards on every platform built after about 2016.
Identical text and identical media posted in a narrow window is the exact signature that integrity systems are built to catch. You are not gaming the feed, you are raising a flag. At best the posts get limited distribution. At worst a platform labels the effort inauthentic, a reporter notices the label, and the story becomes your methods rather than your message. We watched a version of this happen to a mayoral campaign in 2023, where a coordinated push produced a local news segment about astroturfing and a candidate spending two days explaining that the supporters were real.
The version that works is staggered and varied. Same week, same facts, different words, different pictures, spread across six days. It feels less satisfying in the war room because you cannot screenshot a wall of your own posts. It reaches more people. The one exception I would make is a genuine live moment, a debate or a rally, where simultaneous posting is what people do anyway and the platforms expect it.
A test worth running this month
Take your 20 most active supporters and ask each of them to send you a screenshot of the last political thing they posted. Not a survey, a screenshot. Twenty replies will tell you more about what your toolkit should contain than any analytics dashboard, because you will see the platform, the format, the length, and the voice they actually use.
Then rebuild one asset to match the most common shape you got back, send it to those 20 people only, and count what happens over four days. If the number is higher than 23 out of 600, you have learned something. More of this thinking sits in the rest of the social media archive.
FAQ
Should a supporter toolkit include suggested captions?
Include facts and a link, not finished sentences. Identical captions across many accounts read as coordinated to both platforms and people. Give volunteers the figure and the date and let them phrase it.
Do supporter posts need a paid-for disclaimer?
Organic posts by genuine volunteers usually do not, but rules differ by state and country, and anything the campaign pays for or scripts closely can change the answer. Ask your compliance counsel before the first drop, not after.
How many assets does a local campaign actually need?
Fewer than you think. Two sizes, one message point per drop, and roughly one drop a week through the final eight weeks. Sixty files in a folder is a sign the work went into production rather than distribution.
What is the single biggest fix?
Move delivery out of a shared drive and into the app your volunteers open anyway. Download friction, not enthusiasm, is what the 23 posts out of 600 was measuring.
More in Social Media and Politics
Social Media and Politics
Campaign Discord Servers and How Volunteers Organize
A state house campaign I followed last spring set up a Discord server on a Tuesday, posted the invite link on the candidate's...
Social Media and Politics
Campaign Reddit AMAs and What Candidates Get Wrong
A candidate agrees to a Reddit AMA, the staff post a verification photo, and within twenty minutes the thread has two thousand...
Social Media and Politics
Political WhatsApp Groups and How Campaigns Organize in Them
A campaign can post the perfect video on every public platform and still lose the argument inside a family chat it never sees.