# What I learned shipping $418K of CTV inventory for PI firms in 2024.

*Field Notes No. 05 · Channel deep-dive · 2026-04-22 · 10 min read*

CTV looked indefensible until it didn't. Eight months on the buying side of Vibe, MNTN, and a sliver of NBCU direct — what the CPL chart hides, how to model branded-search lift, and the one operational rule that matters more than creative.

I avoided Connected TV for plaintiff-side firms for three years running. The CPMs were ugly, the attribution was worse, and every legal-marketing vendor pitch I sat through ended with the phrase "incrementality study" — meaning the agency couldn't show their spend produced cases, but had built a framework where you'd pay them to keep trying.

In April of 2024 I broke that rule for one firm — a 14-attorney PI shop in metro Atlanta — at a $20K monthly budget on **Vibe.co**, a CTV buying platform pitched specifically at SMB legal. By September I'd rolled it to four more firms across Vibe and MNTN, with a small direct-to-publisher carve-out on NBCU. Total 2024 CTV spend across the cohort: **$418,000**.

It mostly worked. Not in the way I expected. Here's what I learned.

- **$418K** — 2024 CTV spend across 5 firms
- **$591** — Headline CTV CPL — misleading
- **+22%** — Branded-search lift in CTV markets

## § I · The CPL chart is a lie — Why the $591 CTV CPL hides the actual unit economics.

The blended CTV CPL across our cohort in 2024 was **$591**. Next to Google Search at $288 or Meta at $372, that number looks like a category error. It's not. It's measurement error.

Two things the headline CPL ignores:

- **Phone-call inquiries vs. form fills.** CTV drives inbound calls disproportionately — most prospects see an ad on Hulu, screenshot the phone number, and dial later. On the cohort's CRM, CTV-attributed calls signed at **22.4%**; form fills signed at 8.1%. If you blend the two and report a single CPL, you're hiding a structural difference in lead quality.
- **Branded-search lift.** In the four markets we ran CTV continuously for >90 days, branded-search query volume (i.e., people Googling the firm by name) lifted **22% on average** vs the pre-CTV baseline. That lift carried lower CPL than non-branded search. We were crediting that lift to "branded Google Ads improving" until we ran a clean off-week and watched it collapse.

Cost per signed case on the CTV cohort, properly attributed (including call inquiries and branded-search lift): **$2,940**. Better than our cohort average for standard auto MVA. The $591 CPL is genuinely a category error in the CRM, not a real performance number.

> Most firms running CTV are reporting the wrong number to their managing partner and reaching the wrong conclusion. The headline CPL is a lie, the signed-case CPA is the truth.

## § II · The platforms, ranked — What we bought and what worked.

| Platform | 2024 spend | Signed-case CPA | Notes |
| --- | --- | --- | --- |
| Vibe.co | $262K | $2,710 | Best fit for SMB firms. Self-serve, geo-tight, decent reporting. |
| MNTN | $108K | $3,140 | Higher CPMs, stronger creative tools. Worth it for firms that already had broadcast-style spots. |
| NBCU direct | $48K | $4,420 | Premium inventory, premium price. Tried for two months on one firm; pulled back. |

### Vibe.co was the workhorse

Geo-targeting down to ZIP, household-income filters, decent reporting, no minimum spend that mattered. For a firm spending $15–$30K/month on CTV, Vibe is structurally the best fit. The reporting still feels two generations behind Google Ads, but the unit economics work.

### MNTN was creative-driven

Two of the firms we ran on MNTN had pre-existing broadcast spots from local TV buys. MNTN's creative-optimization layer (their "QuickFrame" tool) re-cut those spots into 15s and 30s variants without us needing to ship to an editor. The CPM premium was real (~28% above Vibe in the same geos) but for firms with strong existing brand assets, the per-signed-case math worked.

### NBCU direct was a vanity play

Pure premium inventory, Today Show and local NBC affiliate mornings, $4,420 signed-case CPA. The managing partner liked telling his clients his ad ran on Today; the CFO didn't like the math. We pulled it back to programmatic-priority pricing and the per-signed-case CPA settled to about $2,900 — fine, but not better than Vibe.

## § III · Creative — The one rule that mattered more than format or length.

Across the cohort we tested 22 creatives in 2024 — 15-second and 30-second spots, testimonials and offer-driven, polished studio and iPhone-shot. The one variable that predicted performance most cleanly was not length, not aesthetic, not offer.

It was **whether the creative named a specific event**.

"Hurt in a car accident? Call us." — bad. Generic, low recall, no urgency.

"Were you in an accident on I-285 last weekend? You have ninety days to file in Georgia." — meaningfully better. Specific location, specific timeframe, action-oriented. We ran a variant of that script in every market where we had statute-of-limitations urgency to leverage, and the per-signed-case CPA on those creatives ran 30–40% below generic spots.

## § IV · House arrest — The operational rule that matters most.

The single biggest performance variable on CTV is not creative and not platform — it's **operator restraint**. Once a campaign has delivered for four weeks and the unit economics are converging, the temptation is to tweak: shift dayparts, narrow geos, raise bids, change CTAs. Don't. Every single time we touched a converging CTV campaign in 2024, the model regressed and took 9–14 days to recover.

The operating rule we landed on: **once a CTV campaign enters its fifth week with stable CPA, the only acceptable changes are creative rotation and budget level**. Geographic targeting, dayparts, devices, frequency caps — all frozen.

## § V · The lift model — How to measure CTV branded-search effect.

Here's the cleanest framework we landed on for measuring CTV's downstream lift:

1. **Baseline:** Capture 30 days of branded-search impression share and click volume *before* any CTV campaign launches.
2. **Run period:** Run CTV for 90 days at a consistent budget level.
3. **Off week:** At day 90, pause CTV completely for one full week. Capture branded-search volume for that week.
4. **Measure:** The delta between baseline volume and off-week volume is the residual brand effect. The delta between off-week and a representative run-period week is the active CTV-driven branded lift.

In our four-firm cohort, the active CTV-driven branded lift averaged **22% over baseline branded-search query volume**. That lift is what makes the per-signed-case CPA math work in our favor. Most agencies don't run the off-week test because they're afraid the lift will be smaller than they've been claiming. The truth, in our data, is that the lift is real but smaller than the breathless decks suggest.

## § VI · The verdict — When CTV makes sense for a PI firm.

- You're spending at least **$15K/month** across existing acquisition channels. Below that, CTV will cannibalize other channels you haven't optimized yet.
- You have a consistent geographic concentration — single state or a few metros. CTV doesn't scale well across patchy national footprints for small firms.
- Your intake answers inbound phone calls within 90 seconds, 6am–10pm. CTV drives calls more than forms; if your phone coverage is bad, the spend will land somewhere else.
- You've already got branded-search bid coverage. CTV's lift shows up there first; if competitors are bidding on your firm name and you're not, you'll lose the lift you paid to create.

Cohort detail reflects active CTV campaigns run by [Mass Tort Marketing Agency](https://www.masstortmarketingagency.com) across five PI firms in 2024. Per-firm spend, signed-case CPA, and lift measurements are aggregate; specific firm names redacted under client confidentiality. [Read the 2024 annual postmortem](/field-notes/seven-figures-pi-ads-2024) for the full multi-channel context.

## Related Field Notes

- [The seven figures I spent on PI ads in 2024 — what worked, what didn't.](https://mvaleads.org/field-notes/seven-figures-pi-ads-2024)
- [Why I'm done with programmatic display for plaintiff-side, after three years and $148K of trying.](https://mvaleads.org/field-notes/done-with-programmatic-display)
- [Why my Meta CPM doubled the week iOS 17 dropped, and how we rebuilt the funnel.](https://mvaleads.org/field-notes/meta-cpm-doubled-ios-17)

## Related guides on this site

- [Pricing benchmarks](https://mvaleads.org/pricing)
- [Cost of MVA leads](https://mvaleads.org/cost-of-mva-leads)
- [PPC vs buying leads](https://mvaleads.org/resources/ppc-vs-buying-mva-leads)
- [2026 MVA Lead Cost Report](https://mvaleads.org/reports/mva-lead-cost-report-2026)

---
Canonical: https://mvaleads.org/field-notes/ctv-vibe-pi-2024
Author: Tarun Kapoor, founder, MVA Leads (https://mvaleads.org/about)