9 Michigan Senate Primary Market Mistakes That Cost You Value

A troubleshooter for diagnosing why a Michigan Senate primary market campaign loses value — triage “lost value,” align objectives, correct geography and audience targeting, match inventory to intent, and reduce frequency waste with sequencing.

Triage the Problem

You can’t fix a primary-market buy until you name what “value” was supposed to be. Then you isolate where value leaks, using the same checklist every time.

Treat this like incident response. Gather proof first, diagnose second, change last.

Define “lost value”

Lost value is any spend that fails its intended job, even if the platform shows “delivery.” You’re looking for symptoms you can verify, not vibes.

• Low reach against the intended universe
• Weak persuasion signals despite steady spend
• Wasted frequency on the same people
• Bad geography or district spillover
• Wrong audience or faulty targeting logic

The moment you can label the symptom, you can choose the right fix instead of guessing.

Collect the evidence

Pull the artifacts first so later changes don’t erase the trail.

1. Export buy orders, invoices, and insertion orders.
2. Save placement logs, ad IDs, and targeting settings.
3. Capture screenshots of live placements and previews.
4. Download pacing charts and delivery summaries.
5. Pull post logs, approvals, and compliance notes.

If you don’t have receipts, you don’t have a diagnosis.

Map the funnel

Campaign line items fail when they’re judged against the wrong goal. Split the funnel so awareness, persuasion, and turnout each get their own scoreboard.

Imagine a persuasion video buy getting blamed for low volunteer signups. It might be doing fine; it’s just being graded on turnout.

Set a baseline

Pick a few checks you run on every buy, every time. Non-negotiables beat complicated dashboards.

• Delivered vs ordered impressions and dates
• Geography match to district and precinct plan
• Audience match to targeting brief
• Creative and landing alignment by message

When the baseline fails, stop optimizing and start correcting the plumbing.

Mistake 1: Wrong Objective

In a Michigan Senate primary, the wrong KPI turns real spend into background noise. You buy volume, not movement, because the objective, inventory, and measurement disagree.

Spot the mismatch

Wrong-objective campaigns usually look “busy” while persuasion stays flat.

• High impressions, no downstream actions
• Clicks treated as persuasion proof
• Cheapest reach bought during persuasion
• Video views optimized with no recall read
• Leads counted without validation

If your dashboard is green but your field team feels nothing, the objective is lying.

Reframe the goal

Pick one outcome per flight, or your optimization will average into mediocrity. When you can’t measure the true outcome directly, define a proxy you’ll actually trust.

Example: early persuasion can use completed video views plus brand search lift checks, while GOTV can use validated signups or pledge actions. The key is writing down what “good” looks like before you launch.

Clarity forces tradeoffs, and tradeoffs create signal.

Correct the setup

Fix the plumbing before you judge performance.

1. Update the platform objective to match this flight’s single outcome.
2. Map conversion events to that outcome, and remove competing “success” events.
3. Set reporting to the same north-star metric and its agreed proxy.
4. Split budget by primary phase: persuasion, persuasion-to-ID, then GOTV.

Once the machine aims at one target, creative and targeting finally get a fair test.

Mistake 2: Bad Geography

Bad geography is silent waste. You think you’re buying Michigan voters, but you’re buying whoever your settings allow.

District lines, media markets, and county priorities rarely match neatly. If you don’t reconcile them, value leaks across every platform.

Find geo drift

Geo drift happens when your targeting and your reporting don’t share the same map. You catch it by comparing ZIP, DMA, and county views side by side.

1. Pull ZIP, DMA, and county delivery reports for each platform.
2. Flag spend outside intended counties, plus border bleed near state lines.
3. Check for underweight delivery in priority counties versus your plan.
4. Compare mobile and CTV geo distributions; they often diverge.
5. Repeat after every major budget shift or creative swap.

If three reports disagree, your targeting isn’t “wrong.” It’s undefined.

Fix geo controls

Most geo mistakes come from defaults. Fix them with explicit inclusions, hard exclusions, and documented settings.

• Build county inclusion lists for the whole flight.
• Add exclusions for out-of-state border zones.
• Use tight radii only when you can justify them.
• Set location to “presence,” not “interest.”
• Record each platform’s geo logic in your QA notes.

Treat geo like compliance: written, repeatable, and reviewed.

Validate with logs

Fixes aren’t real until delivery evidence matches your map. Validation protects you from quiet regressions and vendor hand-waving.

Use placement reports to confirm where impressions actually ran. Request publisher affidavits when inventory is syndicated. Spot-check delivery screenshots that show local context and timestamps.

If you can’t prove where it ran, you can’t defend what you bought.

Mistake 3: Audience Misfire

Primary voters aren’t the general electorate. If your targeting is built for November, you waste money in August.

Audience misfires usually come from sloppy definitions, stale models, and missing suppressions. Fix those first, then judge performance.

Audit audience sources

You can’t fix targeting until you know what’s driving it. Different sources behave differently, and recency quietly matters.

Check each audience input:

• Voter file match: match rate, district coverage, last refresh
• Modeled segments: training data, update cadence, confidence thresholds
• Interest proxies: exact interests used, how broad they are
• Lookalikes: seed quality, size, and similarity settings
• Contextual: placements, keywords, and topic exclusions

If you can’t explain your source, you can’t defend your spend.

Catch common errors

Most audience mistakes are boring and repeatable. They also get expensive fast.

• Mixing party universes across districts
• Missing absentee and early voters
• Skipping suppression of opponents’ universes
• Using overbroad “politics” interests
• Reusing a general-election persuasion pool

If the universe is wrong, the creative never gets a fair test.

Repair and retest

Don’t patch the old audience. Rebuild it, then prove it with clean comparisons.

1. Recreate segments from the current voter file cut.
2. Add suppressions for opponents, wrong districts, and known non-targets.
3. Split test by audience type, not by tiny tweaks.
4. Compare delivery, frequency, and cost by segment.
5. Keep winners, then iterate one variable at a time.

Treat audiences like hypotheses, not possessions.

Ensure privacy compliance

Audience repairs can cross policy lines without you noticing. Primary data is political data, and platforms treat it that way.

Confirm your permitted data use, required disclaimers, and each platform’s political advertising rules before scaling. If compliance is fuzzy, stop and get it clean.

The fastest way to lose value is to “fix” targeting and trigger a denial or takedown.

Mistake 4: Inventory Mismatch

Choosing the wrong inventory breaks the chain between message and voter. You think you bought persuasion, but you often bought cheap impressions in the wrong rooms.

Classify inventory

Inventory labels hide real differences in attention, context, and targeting. In a Michigan Senate primary, each channel wins for a different reason.

CTV/OTT: big-screen attention, strong completion, weaker local news context.
Broadcast TV: mass reach fast, high trust environments, limited targeting.
Cable: zone and program control, solid frequency building, fragmented audiences.
Digital video: flexible targeting, variable quality, needs verification.
Display: efficient retargeting, low persuasion alone, heavy placement risk.
Audio/podcasts: high engagement, strong host adjacency, harder message recall.
Social: rapid testing, strong fundraising hooks, brand safety varies by placement.

Buy the format for the job, or your “reach” becomes expensive noise.

Identify wasteful placements

Bad inventory usually looks like “scale,” then performs like fog. Spot it before it eats your persuasion budget.

• Long-tail CTV apps with unclear content
• Made-for-ad sites posing as “news”
• Autoplay video with sound off
• Tiny units nobody can read
• No political adjacency or brand controls

If you can’t explain where it ran, you can’t defend why it worked.

Reallocate intelligently

You don’t need more spend to fix mismatch. You need tighter rules and cleaner supply.

1. Pause placements failing viewability, completion, or fraud thresholds.
2. Shift budget to verified sellers and authenticated inventory.
3. Add channel-specific whitelists and block known junk domains and apps.
4. Tune frequency caps by channel to prevent waste and burnout.

Clean supply turns the same message into a different outcome.

A good starting point is implementing ads.txt (Authorized Digital Sellers) checks across your programmatic buys.

Mistake 5: Frequency Waste

Too little frequency leaves voters unconvinced, even if your targeting is perfect. Too much frequency turns persuasion into irritation and drains budget fast.

You fix it by treating frequency like a system. Caps, sequencing, and flighting do the heavy lifting.

Diagnose frequency issues

Bad frequency rarely looks like “too high” or “too low” everywhere. It shows up as uneven pressure, hidden pockets, and performance that stops moving.

• Reach climbs, but frequency barely moves
• Frequency spikes by daypart or device
• Certain ZIPs get hammered repeatedly
• Outcomes plateau despite steady spend
• Retargeting outpaces prospecting impressions

If you see pockets of oversaturation, you’re not optimizing—you’re leaking.

Apply frequency controls

You need guardrails that match how each channel delivers impressions. Otherwise, the algorithm will happily spend into the easiest, not the best.

1. Set frequency caps by channel and objective.
2. Switch to reach-optimized buying when persuasion stalls.
3. Separate prospecting and retargeting with different caps.
4. Stagger flights to match the primary calendar moments.
5. Review caps weekly and adjust for inventory shifts.

Control frequency, or frequency controls you.

Use message sequencing

Repetition works when each impression adds a new piece of meaning. Sequencing turns frequency from “same ad again” into a guided narrative.

Imagine a four-step order: introduce the candidate, contrast with the opponent, prove with credible specifics, then remind close to voting. Build rotations so the next impression unlocks the next message, not the same one on a loop.

If your frequency is rising, your information density must rise too.

Fix the Biggest Leak First—and Prove It

1. Run triage first: define “lost value,” collect evidence, map the funnel, and set a baseline you’ll keep constant while testing.
2. Start with the highest-leverage fixes in order: objective alignment → geo controls → audience sources/compliance → inventory reallocation → frequency caps and sequencing.
3. Change one variable at a time, annotate the change, and validate with logs and funnel signals—not just top-line spend.
4. Once the leak is plugged, document the corrected setup as your new default so the same mistake doesn’t creep back in during the next flight.