Performance Baseline & Unit Economics

May 15-26, 2026 (12-Day Test Window)

Purpose: Establish a data-backed reference for your records going forward. All numbers come from the Meta Marketing API pull on May 26, 2026.

Campaigns: APTL Cold Broad (AZ Palm Tree Lights, lead-optimized) + TLP Cold Broad W1 (Trimlight Phoenix, traffic-optimized)

---

1. Test Window Summary

Key context: Trimlight has no Meta Pixel installed on trimlightphoenix.com, so even if visitors submit the JotForm, Meta cannot attribute conversions. All Trimlight metrics are traffic-only.

---

2. APTL Campaign Analysis

Daily Performance Trend

The Fatigue Pattern

The data shows a clear front-loading of leads:

• Days 1-3 (May 15-17): 7 of 10 leads. CPL range $11.73-$18.48. Avg CPL: $15.14
• Days 4-7 (May 18-21): 2 leads. CPL range $34.95-$38.99. Avg CPL: $36.97
• Days 8-11 (May 22-25): 1 lead (May 23 only). CPL: $35.12. Three zero-lead days.
• Day 12 (May 26): Partial day after pause, 0 leads.

The learning phase delivered strong results in the first 72 hours, then performance degraded steadily. By day 8, the campaign was spending $30-43/day with zero conversions. This is textbook creative fatigue in a narrow geo audience (Scottsdale/PV ZIPs, ~11K reach total).

CTR also declined: 4.62% on day 1 down to 3.13% by day 11, with a notable dip to 2.60% on May 22 -- the lowest point.

Per-Ad Creative Performance

Observations:

• DayNightSplit (single) consumed 93% of total spend ($371 of $398). Meta's algorithm heavily favored this creative. It generated 9 of 10 leads, but at a blended $41.23 CPL -- inflated by the zero-lead tail days.
• New Leads Ad had the lowest CPL at $9.08, but only received $9 in spend and 263 reach. Too little data to draw conclusions, but the signal is interesting: 1 lead from 263 people = 0.38% conversion rate, vs. the DayNightSplit's 0.089% (9 leads / 10,067 reach).
• DayNightSplit (carousel/multi?) received almost no spend ($17) and generated zero leads. Meta deprioritized it early.

The spend concentration problem: When Meta's algorithm picks a "winner" early, it starves the other creatives of data. The DayNightSplit dominated spend before we had enough data on the alternatives. This is a known issue with Advantage+ creative optimization at low budgets.

CPL Benchmarks

Realistic CPL range for fresh creative: $15-25 in the first 72 hours, degrading to $35+ as frequency builds.

Audience Targeting Notes

• Audience size: ~11,000 unique reach over 12 days
• Geo: Scottsdale/Paradise Valley/Arcadia ZIP codes (85253, 85255, 85258, 85254, 85260, 85018, 85028)
• Optimization: Lead generation (on-platform instant form)
• Ad set: Single cold broad audience

At $35/day average spend hitting only 11K unique people in 12 days, frequency was building fast. The Scottsdale luxury ZIP audience is inherently small. Fresh creative or audience expansion is required to maintain performance.

What Worked

1. Day/Night split creative concept -- the transformation angle clearly resonates. 9 of 10 leads came from it.
2. Lead gen objective with instant forms -- frictionless conversion at the top of funnel.
3. First 72 hours -- CPL of $15.14 proves the offer/audience fit exists when creative is fresh.
4. CTR of 3.62% -- well above Meta benchmarks for home services (typically 1.5-2.5%). The creative stopped scrolls.

What Didn't Work

1. Single creative carrying all spend -- no rotation meant fatigue hit the entire campaign simultaneously.
2. No creative refresh plan -- 12 days of the same image to a small audience. Should have swapped at day 5-7.
3. Budget too thin to test alternatives -- at $35/day, Meta concentrates spend on one ad. Need $50-75/day minimum to test 3+ creatives properly.
4. No retargeting layer -- people who clicked but didn't convert had no follow-up ad sequence.

---

3. Trimlight Campaign Analysis

Traffic-Only Optimization

The Trimlight campaign is optimized for Landing Page Views (LPVs) because there is no Meta Pixel on trimlightphoenix.com. This means:

• Meta optimizes for people likely to click and load the page, NOT people likely to submit the contact form
• Even if 50 visitors submitted the JotForm, we'd have zero conversion data in Meta
• The algorithm is serving ads to "clickers" not "buyers" -- fundamentally different audience
• No retargeting is possible without the pixel
• No lookalike audience can be built from converters

This is the single biggest gap in the Trimlight campaign. Every dollar spent without a pixel is a dollar of wasted learning data.

Daily Performance

Unlike APTL, the Trimlight campaign shows no dramatic fatigue pattern. LPVs remained relatively stable (78-159/day). This makes sense: traffic optimization doesn't hit fatigue walls as hard as lead gen, because the conversion threshold is lower (click vs. form submit).

Cost/LPV range: $0.21 (best, May 18) to $0.35 (worst, May 22). Variance is modest.

CTR trend: Started at 3.03%, drifted down to 1.74% by May 25. This suggests some creative wear but not catastrophic fatigue.

Per-Ad Creative Performance

Observations:

• TC1b (5% Math) dominated spend at 71% ($238 of $336). Reached 38K people. Solid workhorse at $0.29/LPV. Uses a corporate Trimlight Christmas house photo -- on-brand.
• TC1a (Never the Ladder) had the highest CTR at 4.70% -- significantly above the others. But it used a stock landscape lighting photo, not a Trimlight product image. Paused on May 25 because it's off-brand. The high CTR likely came from a visually appealing but misleading image.
• TC1d (Price Clarity) received minimal spend ($19) despite being on-brand. "$25/linear foot" pricing in cold traffic is problematic -- homeowners don't know their roofline footage, so the number is meaningless without context.
• TC1e (Scottsdale Backyard) launched May 25 with only $6.98 spent so far, but the early signal is the best of the group: $0.20/LPV. Uses a real Trimlight install photo (patriotic colors, Scottsdale home). Worth watching closely.

The "$25/ft" Pricing Problem

TC1b and TC1d both reference "$25/linear foot" in copy. For a cold audience:

• Most homeowners have no idea how many linear feet their roofline is
• "$25/ft" doesn't translate to a meaningful total cost in their head
• Without pixel data, we can't tell if the pricing attracts or repels qualified leads
• Recommendation: remove pricing from cold traffic ads; reserve for retargeting where intent is higher

Cost per LPV Benchmarks

The Pixel Gap Problem

Without a pixel:

1. No conversion optimization -- Meta serves ads to people who click, not people who buy. These are different audiences.
2. No attribution -- if a Trimlight website visitor submits a JotForm, we can't tie it back to the ad that drove them.
3. No retargeting -- can't serve follow-up ads to people who visited but didn't convert.
4. No lookalike audiences -- can't build audiences modeled on actual customers.
5. Wasted learning data -- 12 days and 67K reach worth of conversion signals are lost forever.

Estimated impact: Industry benchmarks suggest traffic-optimized campaigns run 3-5x higher effective CPL than conversion-optimized campaigns on the same creative. Installing the pixel and switching to conversion optimization could cut the true cost per lead by 60-80%.

---

4. Creative Learnings

Video vs. Static Performance

No direct A/B test was run during this window. All APTL ads were static images. All Trimlight ads were static images with copy variations.

However, the creative plan (meta-creative-plan-v2.md) designed an explicit Test 1: static C1r vs. video C11 at $25/day each. This test was not executed during the 12-day window.

Runway Gen-4 video assets exist (13 videos, 5 stills, 3 color variations) but were not deployed as ads. These were generated for APTL, not Trimlight. Future phases should test video creative, especially for Instagram Reels placements where motion content typically outperforms static by 1.5-2x on CTR.

On-Brand vs. Off-Brand Imagery

TC1a (Never the Ladder) used a stock landscape lighting photo rather than an actual Trimlight installation. It achieved the highest CTR of any Trimlight ad at 4.70% but was paused because:

• The image showed a competitor-style product, not Trimlight's channel-mounted LED system
• High CTR on misleading imagery attracts the wrong audience expectation
• Brand consistency matters for franchise compliance

Takeaway: Visually appealing lifestyle photography drives engagement, but it must show the actual product. The TC1e launch (real Trimlight install, Scottsdale setting) is the right direction -- early $0.20/LPV supports this.

Copy Themes

What resonated: Transformation and curiosity angles. The day/night split is the proven concept. Math/logic ("only use holiday lights 18 days/year -- why not have them all 365?") performs as a solid evergreen angle.

What's unproven: Price in cold traffic. The $25/ft copy may be filtering out people who don't understand the metric, or it may be attracting cost-conscious shoppers who won't convert at Trimlight's premium price point. Need pixel data to know.

The "$25/ft" Issue in Cold Traffic

This pricing appears in TC1b (the top spender) and TC1d (Price Clarity). Problems:

1. Cold audience doesn't know their linear footage
2. The number sounds cheap per unit but the total ($2,500-$10,000+) is a surprise at quote time
3. May attract price-comparison shoppers rather than quality-first buyers
4. Without conversion tracking, we can't measure if this copy converts or just clicks

Recommendation: Strip pricing from cold traffic ads. Use benefit-focused copy ("Done in a day. Looks incredible every night."). Reserve pricing for retargeting ads where visitors have already researched and have context.

---

5. Audience Insights

Reach vs. Engagement

APTL reached a small audience (10.8K) with moderate frequency (1.59x over 12 days). The narrow Scottsdale/PV geo targeting limits the audience pool. At this rate, frequency would exceed 3.0 within 3-4 weeks -- the typical threshold where creative fatigue becomes severe.

Trimlight reached a much broader audience (67K) at near-1.0 frequency, meaning most people only saw the ad once. This is healthy for a traffic campaign but means the ads haven't had enough repeated exposure to build familiarity. The broader West Valley + North Phoenix geo gives more room to scale.

Geographic Signals

No zip-level breakdown is available from the API pull. However:

• APTL targets Scottsdale/PV/Arcadia ZIPs (85253, 85255, 85258, 85254, 85260, 85018, 85028) -- high-income, palm-heavy neighborhoods
• Trimlight targets West Valley + North Phoenix + Central Phoenix (53 approved franchise ZIPs) -- a much larger geography
• The 6:1 reach ratio (67K TLP vs. 11K APTL) reflects this geographic difference

For future reference: When the pixel is installed, geographic conversion data will reveal which ZIPs convert best. The buyer personas predict Paradise Valley (85253) and North Scottsdale (85255, 85258) as the highest-value APTL ZIPs, and these should be monitored for disproportionate lead volume.

Day-of-Week Patterns

APTL lead pattern: 7 of 10 leads came Thursday-Saturday in the first week. Weekend leads (Saturday = 3, Sunday = 1) suggest homeowners browse and convert while relaxing. However, the fatigue confound makes it impossible to isolate day-of-week vs. creative freshness.

TLP LPV pattern: Sundays and Saturdays consistently delivered the highest landing page views (159 and 157 on the two Sundays, 118 and 124 on the two Saturdays). Midweek (Tue-Thu) saw lower engagement. This aligns with homeowner browsing behavior -- weekend consideration of home improvement projects.

---

6. Benchmarks & Targets

What "Good" Looks Like

Based on the 12-day data, industry comps for home services, and this specific market:

APTL (Lead Gen)

Context for CPL targets: Jack's unit economics show ~$1,595 blended revenue/tree at 66% gross margin = ~$1,053 gross profit per tree. At a $30 CPL with a 15% close rate, the cost to acquire a customer is $200 -- a 5.3x return on ad spend. Even at $40 CPL, the math works ($267 acquisition cost, 3.9x ROAS). CPL becomes problematic above $75 at a 15% close rate.

Trimlight (Traffic -> Leads after Pixel)

Budget Recommendations

---

7. Recommendations for Next Phase

Priority 1: Install the Meta Pixel on trimlightphoenix.com

This is the single highest-impact action. Until the pixel is live:

• Every Trimlight ad dollar optimizes for clicks, not customers
• No conversion data flows back to Meta's algorithm
• No retargeting audiences build
• No lookalike audiences can be created
• The campaign is effectively a billboard, not a direct response machine

Action: Install the Meta Pixel base code + configure a "Lead" standard event on the JotForm thank-you page. Test with Meta's Pixel Helper Chrome extension. This is a 30-minute task for someone with website access.

Priority 2: Fresh APTL Creative (Fatigue Timeline)

The data proves creative fatigue onset at day 5-7 in this audience. The APTL campaign needs:

• Minimum 3 active creatives at any time to distribute frequency
• Weekly creative rotation -- kill the bottom performer every Monday, add 1-2 new concepts
• Use the Runway video assets -- 13 videos already generated, none deployed as ads yet. Test video vs. static (the planned Test 1 from the creative plan)
• Deploy the Phase 2 concepts from the creative plan: C12 (Pool Reflection), C13 ("One Tap Away"), C14 (Price Carousel)

Fatigue budget math: At $50/day to 11K reachable people, frequency hits 3.0 in ~10 days. Creative must rotate faster than frequency builds.

Priority 3: Build Retargeting Audiences

Even without a pixel on Trimlight, APTL can start building retargeting pools now:

• Page engagers (liked, commented, shared, saved) -- audience building passively
• Video viewers (if video ads are launched) -- 25%, 50%, 75% completion audiences
• Instagram profile visitors -- requires active IG profile
• Website visitors (requires pixel on azpalmtreelights.com -- check if this is installed)

Retargeting typically delivers 2-4x better CPL than cold traffic. The creative plan budgets $20/day for retargeting -- this should activate as soon as audiences reach 300+ people.

Priority 4: Trimlight Creative Refresh

• Pause or rewrite TC1b copy to remove the "$25/ft" pricing from cold traffic
• Monitor TC1e (Scottsdale Backyard) -- early $0.20/LPV signal is promising. Give it 72+ hours and 1,000+ impressions before drawing conclusions.
• Get franchisor creative approval locked in so new on-brand concepts (from Trimlight media library photos) can be tested
• Grant system user "Manage Page" permission on the Trimlight Phoenix Facebook page so ads can be created via the CLI/API instead of manually in Ads Manager

Priority 5: Seasonal Ramp Planning

Jack's business does 90% of revenue in 90 days (mid-September through mid-December). Current May spend is off-season. The plan should be:

The May test window's primary value is learning -- not lead volume. Use June-August to systematically test creative concepts, audiences, and copy angles so the September ramp deploys only proven material at scale.

Priority 6: Attribution & Reporting

Current state: no CallRail, no GA4 goals, no Looker Studio dashboard. The approved attribution stack (Meta Pixel + GA4 + CallRail + n8n -> Jobber -> Looker Studio) should be built before September ramp so that:

• Every lead source is tracked (Meta, Google, organic, yard sign, referral)
• Cost per acquisition is measurable across channels
• Jack and the team managing ads can make data-driven budget allocation decisions

---

Appendix: Creative Asset Inventory

Currently Used in Ads

Available but Not Deployed

From the Runway asset manifest (generated May 14, 2026):

• 13 videos (10 hero/lifestyle shots + 3 color variations) -- 5-10s each, Gen-4 Turbo
• 5 still images (day/night split, luxury home, color palette grid, close-up glow, entrance)
• Key unused assets for immediate testing: Shot 1 (The Reveal -- color-shifting dolly push), Shot 4 (Dinner Party -- lifestyle), Shot 6 (Pool Reflection -- aspirational), Shot 7 (App Control -- tech angle)

Creative Plan Queue (Not Yet Produced)

From meta-creative-plan-v2.md, 18 APTL concepts and 8 Trimlight concepts are designed. The immediate priority queue:

1. C12 -- Pool Reflection (lifestyle/aspiration)
2. C13 -- "One Tap Away" (tech/convenience)
3. C14 -- Price Transparency carousel (direct response)
4. R1-R5 -- Retargeting concepts (requires retargeting audience to activate)
5. TC1-TC8 -- Trimlight concepts (blocked on franchisor approval)

---

Generated May 26, 2026. Data from Meta Marketing API live pull.