A voter education and turnout program for GIPL Action Fund's first cycle.
$500+
added to the average Georgia Power bill in three years.
6
rate hikes approved.
2
PSC seats on the 2026 ballot.
01 What we propose
VoteShift is an LGBTQ+-owned Democratic media, buying, and consulting firm based in Birmingham, Alabama. Founded by Daniel Deriso and Griff Gray, two organizers turned strategists. We produce media for television, radio, CTV/OTT, and digital, we buy across every major channel in-house, and we build the digital infrastructure organizations need to communicate, organize, and win. Reed and Pollie award winners.
Digital, SMS, radio, and paid phones, run off one voter universe with one creative system. Built to the budget lines already in your plan and flighted across a September education window and an October ballot-completion window.
Logo and brand system in 72 hours, site live in four to seven days after approval. The website carries the PSC voter guide landing page that every channel drives to.
Coached self-shoot video for volume, plus two in-person shoot days for the leaders whose delivery carries the program. The same reads become radio spots and text scripts.
A single monthly retainer covers strategy, creative production, program management, and reporting across every channel we run. Media, messages, and dials are billed at cost.
02 The problem this program has to solve
Ballot completion is the number that decides this race.
In 2025 the PSC seats were the only statewide races on the ballot. Voters who showed up were there for the PSC, and roughly a million of them voted for the clean energy candidates. In 2026 the PSC sits underneath a governor's race and a U.S. Senate race, on a ballot with dozens of state and local contests. Millions more Georgians will vote, and the large majority of them have never heard of the Public Service Commission. Turnout goes up and the risk goes up with it, because a voter who arrives for the top of the ticket and quits three races down is a voter your program paid for and lost.
A voter who sees the ad once learns that an election is happening. A voter who sees it nine or ten times learns to look for the PSC race on the ballot and fill it in. That is why the plan is built on floors of 9x on connected TV and 10x to 12x on Meta rather than on maximum impressions. We would rather reach 325,000 voters ten times than a million voters three times, because three impressions does not change ballot behavior on a race nobody knows exists.
325,000 voters × 10 impressions
Changes ballot behavior
Ten touches teach a voter to look for the race and fill it in.
1,000,000 voters × 3 impressions
Tells a voter an election is happening
Broader reach, and no measurable effect on a race nobody knows exists.
Both buys cost the same money. Only one of them finishes the ballot.
Six rate hikes in three years and more than $500 a year added to the average Georgia Power bill is the most portable political fact in the state. The data center buildout is the other one. Your polling with the Energy Foundation showed opposition to data centers crossing party lines, and we are seeing the same thing nationally: Republicans running against data centers, Democrats running against data centers, and voters in both parties treating it as a cost-of-living issue rather than an environmental one. For creative, that means affordability leads and clean energy follows.
Your plan launches outreach in August and the digital media plan we built flights October 13 to November 3. That October window is the right place for the majority of the money, and running only inside it asks the creative to do two incompatible jobs at once.
03 Recommendation: split the flight
Carve roughly 25 percent of the digital budget into a September education flight and put the other 75 percent behind the October 13 to November 3 ballot-completion window.
Select a phase to see what runs in it. Dollar figures follow the $130,000 tier selected in the media plan.
Voters have to learn that this office exists before a ballot-completion ask means anything to them. Two weeks is enough time to remind somebody of something. Teaching them something new and then asking them to act on it takes longer.
At $70,000 we would compress the education flight rather than drop it, because the non-metro counties in your plan have almost no organic exposure to PSC coverage.
SMS, radio, and paid phones flight inside the same two windows and carry the same two asks. September teaches the office and drives to the voter guide. October finds the race and fills it in. Sequencing them off one universe is what lets the frequency numbers in section 06 actually land, because a voter who hears the radio spot, gets the text, and sees the connected TV ad has been touched three times by three formats saying one thing.
04 Targeting
That constraint is manageable, and it is most of what we do for labor and advocacy clients operating in the independent expenditure lane.
| Layer | Source | What it does |
|---|---|---|
| Voter file base | Georgia statewide file, accessed through GIPL or an allied partner, or purchased on your behalf | Vote history, geography, age, turnout propensity. Isolates the medium-propensity voters who need touches. |
| Energy and environment interest | Third-party interest segments | Finds the affordability and clean-energy audience without touching party registration or primary history. |
| Cost-of-living stress | Third-party household and income data | Finds the voters actually feeling the Georgia Power increase. This is targetable. |
| Congregational geography | GIPL green team and congregation addresses | Radius and address-level match around each participating house of worship. |
| America Votes overlay | AV data team | Gap analysis, priority county refinement, deduplication against allied programs. |
The asset nobody else has
Twenty congregations raised their hand for the 2025 C3 program. That list, mapped against the voter file, lets us serve a specific faith leader's video to the specific community that leader serves. A pastor in Decatur talks to Decatur. That is a one-to-one , not a lookalike.
It also solves a reporting problem: when a pastor asks whether the video they recorded reached anybody, we can answer with a number for their community.
These two blocks have different media economics and should not be bought the same way. Metro Atlanta connected TV inventory will be expensive and heavily contested in a governor's and Senate year. The non-metro hubs are materially cheaper per impression, less crowded, and Georgia Power bills there are just as high. We would overweight the non-metro hubs relative to their share of the universe, and use Meta and digital audio to carry more of the metro Atlanta load where CTV pricing gets irrational.
The universe above is built once and drives digital, SMS, radio targeting, and the phone list. Cell matches carry the SMS program, address matches carry the phones, and the same geography weights the radio buy. Everything is deduplicated against itself and against allied programs through the America Votes overlay, so a voter counted in one channel is sequenced in the others rather than paid for twice.
Voter file access, and this one is the critical path. We do not hold the Georgia file for this program. Access comes through GIPL, through one of your partners in the state, or we purchase it on your behalf and bill it at cost. America Votes is the obvious first call, and an introduction to their data team is the fastest way to settle it. If the file has to be bought, we will price it before we scope the universe so the number sits in your budget rather than arriving as a surprise. Nothing else in section 04 can be built until this is resolved, which is why it belongs in week one.
The 2025 State Voices match-back. We would like whatever match-back exists on that universe, even if it is only the contact list without vote history. Knowing who GIPL already touched lets us suppress, sequence, and measure incremental reach rather than paying twice for the same voter. If that data is not recoverable, we will build the universe from the file and the interest overlays and note the limitation in reporting rather than paper over it.
05 Creative
We send framing, lighting, and camera app settings, walk the leader through the script by phone, and they record on their own device. We handle the edit, captions, graphics, and disclaimer. Keeps per-asset cost low and keeps the content credible, because the low-production look is the persuasive part.
One day in metro Atlanta, one in a non-metro hub, with our crew and our equipment. For the four or five leaders whose delivery carries the program, and for the B-roll library that makes every other cut better. We are a four-hour drive from Atlanta and we own our gear.
At the recommended tier, 12 to 18 faith leader cuts, each in :15 and :30, plus 9:16 vertical for Meta and Reels, plus static and motion ballot-completion units. Scripts and direction come from us. Selfie-style does not mean unscripted.
The strongest idea in this document
Your plan has the voter guide moving through mail and digital. We would build it as a mobile-first page on the new Action Fund site with a ballot lookup, a plain-language explanation of what the PSC controls, and the two races laid out. Then we run the September education units against it. That changes what the media buy produces: every click becomes a known supporter with an email address.
Marqus made the point that the goal is for voters to know who GIPL Action Fund is cycle after cycle, through the 2027 municipals and into 2028. A voter guide landing page is how paid media builds that list instead of renting attention for three weeks.
In Wisconsin we ran ballot information for Leaders Igniting Transformation on TouchTunes jukeboxes and gas station pump screens, with scan-to-learn-more built in. Gas pump screens are the right place in the world to talk about the cost of energy, and the format is cheap, uncontested, and nothing like the political mail and TV that voters are tuning out by late October. We would scope digital out-of-home as an add-on for the non-metro hubs rather than a substitution for core video.
06 Media plan: the digital line
These are the $70,000, $130,000, and $210,000 digital lines in your plan. SMS, radio, and paid phones are priced separately and covered in section 07.
A voter who sees the connected TV spot receives two additional impressions on their desktop or mobile device within the following week. The impression establishes the message and the follow-up impressions burn it in. In surveys we see materially higher message recall than programs without device follow-up. On a race where the whole job is teaching voters that an office exists, that repetition across screens is the mechanism.
Day 0
CTV impression
The spot runs on the television, in full screen, unskippable.
Within 7 days
Device impression
Same voter, same message, on desktop or mobile.
Within 7 days
Device impression
The second follow-up is where recall gets built.
Pricing is , cost per completed view, at $0.074. Fifty percent of the video must be in view for 15 seconds on in-stream and in-read placements, 10 seconds on desktop in-page, 5 seconds on mobile in-page. You are paying for views, not for served impressions.
Voter universe
325,000
voters reached
CTV frequency
9x
floor, per voter
Meta frequency
10x
floor, per voter
Billable views
1,256,757
completed, at $0.074 CPV
| $70,000 | $130,000 (recommended) | $210,000 | |
|---|---|---|---|
| Voter universe | 150,000 est. | 325,000 est. | 500,000 est. |
| VDX TV Magnify Surround | $50,000 | $93,000 | $150,000 |
| CTV frequency | 10x | 9x | 9x |
| VDX impressions | 1.3M–1.6M | 2.4M–3.1M | 3.9M–4.8M |
| Meta | $17,400 | $32,000 | $52,200 |
| Meta frequency | 12x | 10x | 10x |
| Meta impressions est. | 870,000 | 1.6M | 2.6M |
| Meta matched est. | 75,000 | 160,000 | 250,000 |
| Direct buys from publishers | $2,600 | $5,000 | $7,800 |
| Billable views | 675,676 | 1,256,757 | 2,027,028 |
| Total | $70,000 | $130,000 | $210,000 |
YouTube and pre-roll. Cheap incremental frequency against the same matched audience, useful for the September education phase.
Digital out-of-home. Gas station pump screens in the non-metro hubs, scoped as an add-on rather than a substitution for core video.
The other three channels. SMS, radio, and paid phones run off this same universe and are priced in section 07. Running them together is how the frequency numbers above actually get hit.
07 SMS, radio, and paid phones
Digital teaches the office at scale. These three close the distance. A text gets answered by a human, a gospel radio spot reaches a voter who is not on Meta, and a live caller can ask whether somebody knows the PSC is on their ballot and hear the answer. All three run off the universe in section 04 and carry the creative from section 05.
CTV/OTT and Meta at frequency. Teaches the office, drives the voter guide.
10% commission on placement
Two waves, education then ballot completion. Real people answer replies.
$0.08 per message, all in
Gospel and urban AA, terrestrial in the hubs and streaming in metro Atlanta.
10% commission on placement
Live callers on ballot completion, for the least online part of the universe.
Vendor cost, pass through
Pricing is usage based at eight cents per message sent, and that number is all in. It includes live responses from actual people on our team and the labor behind them. When a voter texts back asking what the PSC is, or where they vote, or to be taken off the list, a person answers. That is the part most vendors bill separately or quietly automate.
Your plan budgets $15,000 to $60,000 for SMS. At eight cents a message that is the volume below. We are sending millions of texts into five states this cycle including Georgia, with script and built in.
| Spend | Messages sent | What it covers |
|---|---|---|
| $15,000 | 187,500 | Low end of your SMS line. One education wave plus a narrow ballot-completion wave in the priority counties. |
| $30,000 | 375,000 | Both waves across the metro Atlanta counties and the non-metro hubs, with room for a final 72-hour push. |
| $52,000 | 650,000 | Two touches against a 325,000-voter universe, matching the recommended digital tier. |
| $60,000 | 750,000 | Top of your SMS line. Adds a third touch and a replies-driven follow-up wave. |
Volume math at the stated rate, not a promise about cell match rates. The deliverable universe is the cell-matched share of the file, and we report it before the first send rather than after.
The September wave drives to the voter guide landing page, which is what turns a text into a known supporter with an email address. The October wave carries the ballot-completion ask, early vote urgency from the 13th, and Election Day urgency in the final 72 hours. Sends are suppressed against the rest of the program so the same voter is sequenced rather than hit twice in a day.
Gospel and urban AA formats reach your faith universe at a fraction of terrestrial radio cost, and a large share of our radio placement already runs there. We buy and produce radio in-house, so the same faith leader reads become audio spots with no additional talent cost and no second production line item. Placement carries the same 10 percent commission as the rest of the media, and production is included with the buy.
Terrestrial for the non-metro hubs, where gospel and urban AA stations still carry real reach and rates stay reasonable through October. Digital audio and streaming for metro Atlanta, where the targeting is addressable against the same universe and the pricing does not get pulled around by the governor's race. A pastor your voters already listen to on Sunday reading the same script they recorded on camera is the cheapest frequency in the plan.
Live callers on ballot completion, aimed at the oldest and least online part of the universe. Phones are the only channel in this program where a voter can ask a question and get an answer in the same breath, which matters on a race where the first honest reaction is usually "what is the Public Service Commission."
Dials run off the same universe. The script confirms that two PSC seats are on the ballot, points to the voter guide, and asks the voter to commit to a voting plan. Vendor cost is pass through at cost and the per-dial rate gets confirmed with the vendor at scope. Script, training, list management, and reporting sit inside the retainer.
Phones also produce the cleanest read in the program on whether the education flight is working, because a live caller can ask a voter directly whether they know the PSC is on their ballot. That is a daily signal on the exact thing the brand lift survey measures once at the end.
Why one vendor across all four
Four vendors running four programs off four lists will double-serve some voters, miss others, and hand you four reports that do not reconcile. One universe, one creative system, and one team means the connected TV impression, the text, the radio spot, and the phone call are counted against the same voter and ordered on purpose. It also means the congregation-level reporting in section 10 can tell a pastor how many people in their community were reached, across every channel, by them.
08 Phase 0: the C4 brand and website
Nothing in the media plan can run until both exist, because every ad needs a destination and every disclaimer needs to name an entity that looks like it exists. This is the first thing we would build, and it runs in parallel with universe construction rather than in sequence with it.
| Deliverable | What it includes | Timeline |
|---|---|---|
| Brand system | Primary logo, secondary and stacked lockups, favicon, color palette, type system, social templates, one-page brand guide | 72 hours from creative direction approval |
| Website | Homepage, about, PSC voter guide landing page with ballot lookup, email and SMS signup, donate page if applicable, mobile-first throughout | 4 to 7 days after brand approval |
| Compliance treatments | Ad disclaimers, paid-for-by treatments across every format and aspect ratio, site footer disclosures | With brand delivery |
| Social build-out | Profile and channel art sized for every platform, launch post templates | With brand delivery |
We would want to see whatever direction Jay already has, plus the current GIPL identity, because the Action Fund should read as clearly related to GIPL and clearly its own entity. Voters who know the C3 should recognize the C4 without the two being confused for compliance purposes.
We live in a rapid response world with our candidate clients. We have turned broadcast-quality creative in 24 hours off B-roll when a race demanded it. More time makes better work, and the capacity is there when a funder commits late and money has to go up fast.
09 Measurement and reporting
Allie raised a specific frustration: past digital vendors moving past the metrics without making sure anyone understood them. Marqus raised a related one: needing to prove to funders, and more importantly to congregations, that the money moved actual people. Those are two different reporting problems and we build for both.
Live, refreshed every hour. Every channel reports into the same place: impressions and views by medium for digital, messages sent and delivered and replied for SMS, spots aired and affidavits for radio, and dials, contacts, and completed conversations for phones. Performance by individual creative, , average seconds watched per viewer, frequency delivered against target. Every metric has a hover explainer, so nobody has to email us to find out what a view-through rate is. We join your weekly calls. Being in the room when messaging gets decided is the part of this engagement we actually want.
GIPL Action Fund · PSC program
Updated hourly
Daily completed views, education flight through ballot completion
Illustrative layout only. The numbers above are placeholders, not projections.
Live dashboard, weekly call, mid-flight optimization memo with reallocation recommendations.
End-of-flight report: universe reached, frequency delivered, cost per voter reached, creative performance, brand lift results.
Congregation-level delivery summaries: how many voters in their community saw their leader, how many times, on what devices.
The third one is the report most vendors will not build for you, and it is the one that compounds. A pastor who can see that 4,000 people in their community watched their video is a pastor who records another one in 2027.
We recommend fielding a brand lift survey at $130,000 or above. It measures whether voters in the treated universe can identify what the PSC does and that it is on their ballot, against an untreated control. That is the actual objective of this program, and impressions do not measure it. It is the cleanest number you can hand a funder, and the cleanest number for deciding what to do differently in 2028.
10 Timeline
| Date | What happens |
|---|---|
| Week of Aug 3 | Engagement starts. Creative direction call, brand discovery with Jay, and voter file access resolved through GIPL, a partner, or purchase. VoteShift syncs with the America Votes data team and universe construction begins. |
| Week of Aug 10 | Brand system delivered. Faith leader recruitment and scheduling with Marqus. Scripts drafted for education phase, video and text and radio off one message set. SMS compliance and number provisioning start. |
| Week of Aug 17 | Website live including PSC voter guide landing page. Self-shoot coaching kits out to first wave of faith leaders. Universe locked with AV, then split into cell, address, and geography layers for SMS, phones, and radio. |
| Late Aug | Two in-person shoot days, metro Atlanta and one non-metro hub. Education creative into edit. Radio spots cut from the same faith leader reads. Phone script written and callers trained. |
| Sept 8 | Education flight launches across digital and radio. Gospel and urban AA placement begins. SMS education wave sends. Everything drives to the voter guide, and email and SMS capture is live. |
| Sept | Direct mail deploys per your plan. GOTV creative production runs in parallel across video, audio, and text. Weekly calls begin. First A/B reads back from the SMS wave. |
| Oct 5 | Voter registration deadline. Registration messaging sunsets, GOTV creative loads in. |
| Oct 13 | Early voting opens. Ballot-completion flight launches. Faith leader GOTV video, make-a-plan and social pressure messaging. SMS wave two, radio GOTV rotation, and paid phones all layer in on the same universe. |
| Oct 13 to 30 | Peak spend across all four channels. Daily optimization, creative rotation against performance, phone contact rates read daily against the education assumption, mid-flight memo in week one. |
| Oct 30 | Early voting ends. Final 72-hour Election Day urgency creative across digital, SMS, and radio. Phones run through close of polls. |
| Nov 3 | General election. |
| Nov 10 | End-of-flight reporting, funder report, congregation-level summaries. |
| Dec 1 | Runoff, if needed. |
11 Pricing and fees
The retainer scales with how much of the program we run. Placement carries a 10 percent commission on every channel, SMS is usage based, and everything else is either a one-time project fee or a pass-through you can audit.
| Per month | Scope | What we are running |
|---|---|---|
| $5,000 | Digital and creative | The digital program, the faith leader creative series, static and motion units, the live dashboard, weekly calls, and all reporting. |
| $6,500 | Add SMS or radio | Everything above, plus one of the two: the SMS program end to end, or radio production and placement. |
| $8,000 | The full program | Digital, SMS, radio, and paid phones, managed off one universe with one creative system and one report. |