The owner's Monday morning starts at 6:30 a.m. with twelve numbers on one page. That is the dashboard. Not a Looker tab the owner opens once a quarter. Not a CFO spreadsheet that ships Friday at 9 p.m. Not a vendor portal with 47 widgets that takes 20 minutes to read. Twelve numbers. One page. Refreshed daily by the time the truck rolls. Read in 8 minutes. One priority set for the day. The L4 owner consumes the L3 service-manager dashboards, the L3 dispatch dashboards, the L3 ride-along scorecards, the marketing recap, and the RC&D rollup — and pulls the top twelve into a single artifact that drives the day. Booking %. Missed-call %. After-hours capture. Average ticket. MPR. Financing close %. RPT (service). RPT (replacement). Recall %. CSR show rate. GLSA ROAS. RPL (revenue per lead). Built from ServiceTitan or Sera or Housecall Pro plus CallRail plus Avoca plus GLSA plus Hatch plus the accounting system, joined nightly, formatted ruthlessly. This lesson is the dashboard spec — what each number measures, where it comes from, what the 2026 target band is, and why these twelve and not the other forty.

Why Twelve Numbers, and Why One Page

The owner of a 4-12 truck shop has a finite attention budget at 6:30 a.m. Eight minutes of dashboard reading is the operational ceiling sustainable across 50 weeks of the year. Push the format to 15 minutes and the owner reads it three days a week instead of five. Push to 20 and the owner reads it Mondays only. The 8-minute discipline is what keeps the dashboard alive. The number of metrics the owner can scan, interpret, and convert to a single daily priority inside that 8 minutes is 12 — the ceiling for sustained pattern recognition that Atul Gawande-style checklist research and operations literature both converge on. Fewer than 10 metrics under-covers the operating surface; more than 14 produces attention fatigue and the dashboard becomes a reference document the owner consults rather than an instrument the owner reads.

One page is the second discipline. Two pages doubles the cognitive overhead — the owner has to remember what page one said while reading page two. The phone screen is the format constraint: the owner reads the dashboard on a phone while drinking coffee or sitting in the truck cab, not on a desktop in the office. One page that fits a phone screen — twelve rows, three columns (current value, trailing-4-week average, delta) — is the form factor that survives 50 weeks of daily reading.

The dashboard does not produce decisions. It produces priorities. The owner reads twelve numbers, identifies the one or two off-trend, and sets the day's priority around the off-trend. The decisions happen at the standup, in the coach call, in the comp-plan meeting — not at the dashboard. The dashboard surfaces what needs attention; the owner allocates the attention. Conflating the two — building a dashboard that prescribes decisions — produces over-engineered AI dashboards that the owner stops trusting after the third bad recommendation. The twelve numbers tell the owner where to look; the owner decides what to do.

The Four Funnel Numbers — Booking %, Missed-Call %, After-Hours Capture, CSR Show Rate

The first four numbers cover the inbound call funnel — every dollar in the shop starts with a phone call landing somewhere. Booking % is the headline funnel metric: numerator is booked appointments, denominator is total inbound calls. 2026 target band is 80-85% on a human CSR floor, 80-87% effective floor when AI receptionist plus human CSR are running together. Baseline at most under-trained shops is 65%. Every booking point at a typical $5M shop is worth roughly $9K-$14K of annual revenue. The dashboard reports the combined effective floor — not the CSR floor alone and not the AI floor alone — because the owner cares about total funnel throughput, not vendor attribution.

Missed-call % is the inverse leak: numerator is calls that hit voicemail, abandoned during hold, or never answered; denominator is total inbound. The Avoca-built-around-it baseline is 22%. 2026 target is under 5%. Source: CallRail Conversation Intelligence plus Avoca's missed-call dashboard plus the FSM appointment record. The dashboard surfaces the gap between calls in and bookings registered — if the shop is at 78% booking but 18% missed-call, the marketing manager's funnel narrative is incomplete and the morning's priority is the CSR floor coverage or the AI receptionist routing rules.

After-hours capture rate is the third funnel number and the one most directly attributable to the AI stack. Baseline at shops without an Avoca, Jobber AI Receptionist, Housecall Pro AI Agents, or ServiceTitan Voice deployment is 0-15% — calls land in voicemail at 6:42 a.m. or 9 p.m. and the homeowner is on a competitor's truck by 7:20 a.m. Monday. 2026 target with a deployed AI receptionist is 80%+ — the AI answers, books direct to the FSM, sends the on-call tech text only for true emergencies. The dashboard reports the after-hours capture rate as a standalone number because it isolates the AI receptionist's contribution from the human CSR floor's.

CSR show rate closes the funnel: numerator is appointments that showed; denominator is appointments booked. Baseline is 84% — the homeowner who booked at 2 p.m. Tuesday and forgot, or never received the morning confirmation, or got cold feet on the dispatch fee. 2026 target is 92%+. Source: ServiceTitan appointment outcomes joined to CallRail's original booking call. The dashboard surfaces the booking-to-show conversion as the operational reality between "we booked it" and "we ran it" — the headline that closes the four-funnel-numbers block.

The Four Margin Numbers — Average Ticket, MPR, Financing Close %, Recall %

The funnel produces calls that run; the margin numbers report what those calls earn. Average ticket is the headline margin metric. 2026 median for residential service is $450-$600 per completed job; high-mix shops with mature good-better-best presentation discipline run $800-$1,500; replacement tickets run $8K-$45K residential and $25K-$250K commercial. The dashboard reports average ticket against the previous trailing-4-week — the absolute number matters less than the trend, because shop mix shifts seasonally and average ticket is a coarse aggregate of service, repair, and replacement mix.

MPR — Membership Penetration Rate is the second margin number and one of the most under-coached. Numerator is closed jobs that result in a membership upsell; denominator is total eligible closed jobs. Baseline at under-coached shops is 22%. 2026 target is 35-50%; top-quartile Nexstar-style shops run 60%+. MPR matters because membership revenue is the recurring base — the $19-$29/month per home that compounds into a 3,000-member roster worth $700K-$1M of recurring revenue with 80%+ gross margin. Source: FSM membership records joined to closed-job records. The dashboard surfaces MPR weekly trend because membership pitch discipline degrades fastest when the field is busy and the techs default to "diagnose, repair, leave" without the 30-second membership pitch the L2 Ch4 lesson scripted.

Financing close % is the third margin number and the single biggest dollar lever on the replacement ticket. Numerator is jobs $5K+ where the homeowner chose to finance through Wisetack, GreenSky, or Synchrony; denominator is jobs $5K+ total. Baseline is 14%. Wisetack's published 2026 benchmark for the trades is 28-40%. The lift from 14% to 30% on a shop running 200 replacement quotes per year at $14K average ticket is $30K-$50K of margin that previously walked away because the kitchen-table conversation didn't pivot to payment. The dashboard reports financing close % weekly because the lift is fragile — a Comfort Advisor who stops doing the soft-pull-at-the-door workflow loses 4-6 close points within two weeks.

Recall % is the fourth margin number and the cost-of-quality metric. True recall is the same customer, the same complaint, the system still broken — distinct from callback (tech missed something on the original visit) and warranty (manufacturer part failure). 2026 industry median true-recall rate is 4-7%; top-quartile shops run under 2%. Each recall costs roughly $185-$340 in tech labor plus parts plus dispatched windshield time, all unbillable. A shop at 5.4% recall on 1,500 annual jobs is bleeding $15K-$27K of pure margin to operational gaps that the L3 Ch5 RC&D workflow exists to close. The dashboard reports true-recall % separated from callback and warranty so the number is operationally legible — the owner knows whether the spike is internal (tech training gap), supply (bad part batch), or process (incomplete diagnostic).

The Two Truck Numbers — RPT (Service) and RPT (Replacement)

The funnel and margin numbers measure throughput and per-job economics; RPT measures the truck's daily output. RPT service is revenue per service truck per day. 2026 shop baselines per Built on Tenth, Level CFO, and MarginPlug 2026 data: $1,600-$2,400/day for residential service trucks at most shops; top-quartile target is $2,800-$3,500/day. The dashboard reports RPT service as the trailing-7-day rolling average to smooth weekend and holiday volatility. Source: ServiceTitan or Sera or HCP job records aggregated by tech and day. The number captures dispatch yield (the dispatcher's hidden math of who-gets-which-call) plus tech productivity (close rate × average ticket × calls per day) plus equipment-mix exposure (the tech with the freon mix gets the better tickets in July).

RPT replacement is revenue per replacement truck per day — separated from service because the economic model is fundamentally different. Replacement trucks run 1-3 jobs per day at $8K-$45K residential per job; service trucks run 5-8 calls at $450-$600 per call. The 2026 top-quartile target for replacement trucks is $2,500-$4,000/day residential and $5K+ for commercial. Source: same FSM as RPT service, filtered to replacement-job categories. The dashboard reports both numbers because the Comfort Advisor team's productivity sits inside RPT replacement; the service-tech team's sits inside RPT service; conflating them masks the underlying performance.

The two RPT numbers report against the trailing-4-week as the trend column. RPT moves slowly — week-over-week noise dominates the signal unless the trailing window smooths it. The owner reads "RPT service $2,180 / 4-wk avg $2,140 / +$40" and infers the dispatch and ride-along workflows are holding; "RPT service $1,980 / 4-wk avg $2,180 / -$200" triggers the L3 Ch3 dispatch review and the L3 Ch4 ride-along scorecard pull. The trend column is the leverage; the absolute number is the context.

The Two Marketing Numbers — GLSA ROAS and RPL by Source

The last two numbers report what the marketing manager owns. GLSA ROAS is Google Local Service Ads return on ad spend. 2026 baseline with mature manual optimization sits at 3-4x; AI bidding through Ryze AI or similar 2026 GLSA bid platforms lifts the baseline by 30-50% — so the target band on a shop running AI bidding is 4-6x. Below 3x means the GLSA account needs marketing-manager intervention (negative-lead disputes, geo-tuning, service-category rebalance); above 5x means the spend ceiling can probably move higher and the marketing manager should propose a budget lift in the Friday recap. Source: GLSA dashboard joined to ServiceTitan or HCP closed-revenue. The integration is the discipline that produces the closed-revenue ROAS rather than the cost-per-lead ROAS that GLSA reports natively.

RPL — Revenue per Lead — by source is the last number and the highest-context one. It is a single dashboard row with the top 3-5 lead sources listed: GLSA RPL, PPC RPL, organic RPL, Hatch reactivation RPL, direct-mail RPL. Each source produces a RPL figure: closed revenue divided by lead count over the trailing-30-day window. The owner reads which channel is producing the highest RPL and which is leaking. A $5M shop running GLSA at $720 RPL, organic at $1,840 RPL, Hatch at $2,200 RPL, and direct mail at $290 RPL has an obvious next-week reallocation toward Hatch and organic. The dashboard surfaces the comparison; the marketing manager owns the reallocation in the Friday recap. Source: CallRail attribution joined to ServiceTitan closed-revenue plus marketing-spend system.

The two marketing numbers close the dashboard's twelve-row format. Funnel (4), margin (4), trucks (2), marketing (2). Each block measures a distinct operational surface; no number duplicates another; the format is calibrated for 8-minute morning reading on a phone. The metric definitions are stable across shop sizes from 4-truck Jobber operations to 80-truck ServiceTitan platforms — only the data-source integration depth scales.

The Data Pipeline and the Refresh Cadence

The twelve numbers are joined from six systems. FSM platform (ServiceTitan, Sera, Housecall Pro, FieldEdge, BuildOps) supplies booking %, average ticket, MPR, financing close %, recall %, RPT service, RPT replacement, and CSR show rate. AI receptionist (Avoca, Jobber AI Receptionist, Housecall Pro AI Agents, ServiceTitan Voice) supplies after-hours capture rate and contributes to booking %. CallRail Conversation Intelligence supplies missed-call %, contributes to booking %, and supplies the lead-source attribution that feeds RPL. GLSA dashboard supplies GLSA ROAS. Marketing-spend system (GLSA spend, NiceJob, Podium AI Employee, Birdeye AI Employee, Hatch) supplies the spend denominators for RPL and the cost-per-booked-call diagnostic. Accounting system (QuickBooks, Xero, FSM-integrated) supplies the gross-margin context that converts revenue numbers to margin contribution.

The pipeline runs nightly. The owner's 6:30 a.m. dashboard reads the data as of midnight the prior day. Real-time dashboards are a vendor pitch, not an operational requirement — the owner does not need to see at 9:14 a.m. that the 9:02 a.m. call booked; they need to see at 6:30 a.m. Tuesday what Monday's numbers were against the trailing week. The nightly join is the cadence that supports the 8-minute read at the morning's first cup of coffee without producing the constant ping of real-time alerts the owner learns to ignore.

The format is text or PDF, not a dashboard tool. Looker, Sigma, Metabase, ServiceTitan's native reporting, Power BI — all produce richer artifacts than the twelve-row text the dashboard requires. The owner does not open a dashboard tool every morning at 6:30 a.m. The owner reads a Slack message, a text, an email, or a PDF that pushes to their phone. Format discipline is what makes the cadence stick. The L3 service-manager builds the data pipeline and produces the artifact; the L4 owner consumes the artifact. The handoff is the same as every other L3-to-L4 boundary in the trades program.

The Eight-Minute Daily Routine

The dashboard exists to power the owner's daily 8-minute routine. The routine has four steps. Step one (2 minutes): open the dashboard, scan all twelve numbers, identify the off-trend ones (delta column outside the noise band). Step two (2 minutes): read the four alerts that fire when any of the twelve cross a threshold — missed-call % above 8%, recall % above 4%, dispatch override rate above 25%, complaint flag from CallRail sentiment crossing the warning band. Step three (2 minutes): read the AI-summarized customer story — one positive or negative customer narrative pulled from CallRail or the FSM that lands the metrics in a human moment. Step four (2 minutes): set the day's one priority — the off-trend metric that warrants the owner's intervention or the alert that warrants the standup conversation.

The 8-minute routine compounds. Over 50 weeks of daily reading, the owner has invested roughly 33 hours in dashboard consumption — and produced 250 daily priorities that drove operational decisions, training huddles, comp-plan tweaks, vendor conversations, and standup escalations. The owners who run the routine for a full year operate the shop on data-driven priorities; the owners who skip it operate on the loudest fire of the morning. The difference compounds quarterly into the operating cadence that distinguishes top-quartile from median shops.

The routine has two failure modes. The first is dashboard skip — the owner stops reading because the format degraded (too many metrics, attachments added, real-time alerts replaced the morning summary). Fix: reassert the twelve-row text format and audit any change that broke the cadence. The second is priority paralysis — the owner reads the dashboard, identifies four off-trend metrics, and tries to address all four in one day. Fix: discipline the one-priority rule; the other three off-trend metrics get queued for the Friday weekly review or the standup huddle.

The Friday 25-Minute Review and the Quarterly 90-Minute Review

The 8-minute daily routine sits inside a weekly and quarterly cadence. Friday 25-minute review: the owner reads the trailing-week view of all twelve numbers plus the marketing manager's Friday recap (covered in L3 Ch6 Lesson 3) plus the service manager's L3 Ch2 dashboard (covered in L3 Ch2 Lesson 4) plus the L3 Ch3 dispatch report plus the L3 Ch4 ride-along scorecard summary. Twenty-five minutes is the operational ceiling for sustained Friday engagement. The owner reads the consolidated view, identifies the week's wins, the week's gaps, the next week's two-or-three priorities, and walks into Monday's standup with the priorities queued. The Friday review produces decisions; the daily routine produces priorities; the distinction matters.

Quarterly 90-minute review: the owner reads the rolling 13-week view of all twelve numbers plus the quarterly L3 attribution audits (the dashboard's underlying ROI defenses) plus the tool-ROI roll-up for the AI stack (Avoca contribution, Rilla contribution, Dispatch Pro contribution, Hatch contribution, ResponsiBid contribution — each translated to dollar terms). Ninety minutes is the operational ceiling for sustained strategic engagement. The owner ends the quarterly review with three decisions: which AI tools to renew, which to kill, which new pilots to start in the next 90 days. The decisions translate to the next quarter's budget approval and the next quarter's roadmap milestones.

For platform-owned shops and franchisees reporting up to Wrench Group, Authority Brands, Apex Service Partners, Sila Services, Path Light Pro, Redwood Services, or ARS/Rescue Rooter HQ, the quarterly review feeds the platform QBR slide deck. The twelve dashboard numbers map directly to the platform QBR's standard slide structure: funnel slide (the four funnel numbers), margin slide (the four margin numbers), truck slide (the two RPT numbers), marketing slide (the two marketing numbers), plus the EBITDA waterfall with AI-ROI line items. The owner walks into the QBR with the deck pre-built, edits the executive summary in 30 minutes, and the conversation moves to strategy rather than data assembly. The dashboard is the single artifact that scales from daily 8-minute reading to quarterly PE-board defense without restructuring.

Key Takeaways

• Twelve numbers, one page, 8 minutes daily. Funnel (4): booking %, missed-call %, after-hours capture, CSR show rate. Margin (4): average ticket, MPR, financing close %, recall %. Trucks (2): RPT service, RPT replacement. Marketing (2): GLSA ROAS, RPL by source. The format is the message.
• 2026 target bands: booking % 80-85% (baseline 65%), missed-call % under 5% (baseline 22%), after-hours capture 80%+ (baseline 0-15%), CSR show rate 92%+ (baseline 84%), average ticket $450-$600 service median, MPR 35-50% (top-quartile 60%+, baseline 22%), financing close % 28-40% on $5K+ jobs (baseline 14%), recall % under 2% (industry median 4-7%), RPT service $2,800-$3,500/day top-quartile from $1,600-$2,400 baseline, RPT replacement $2,500-$4,000/day, GLSA ROAS 4-6x with AI bidding (3-4x mature manual baseline), RPL by top 3-5 channels weekly.
• Six data sources, nightly join. FSM (ServiceTitan / Sera / HCP / FieldEdge / BuildOps), AI receptionist (Avoca / Jobber AI Receptionist / HCP AI Agents / ServiceTitan Voice), CallRail Conversation Intelligence, GLSA dashboard, marketing-spend system (NiceJob / Podium AI Employee / Birdeye AI Employee / Hatch), accounting system (QuickBooks / Xero / FSM-integrated). L3 service manager builds the pipeline; L4 owner consumes the artifact.
• Format discipline beats tool sophistication. Twelve rows, three columns (current, trailing-4-week, delta), phone-screen size, text or PDF push — not a Looker tab the owner has to log into. Format is what keeps the cadence alive across 50 weeks of daily reading.
• The daily 8-minute routine has four steps: scan twelve numbers, read four alerts (missed-call %, recall %, dispatch override rate, sentiment complaint flag), read one AI-summarized customer story, set the day's one priority. Routine compounds quarterly into the operating cadence that distinguishes top-quartile from median shops.
• The dashboard produces priorities, not decisions. Decisions happen at the standup, the coach call, the comp-plan meeting. Conflating the two produces over-engineered AI dashboards the owner stops trusting.
• Friday 25-minute review consolidates twelve numbers plus L3 dashboards plus marketing recap; produces next week's two-or-three priorities. Quarterly 90-minute review consolidates 13-week view plus tool-ROI roll-up; produces three decisions (renew, kill, pilot).
• For platform-owned shops and franchisees, the dashboard maps directly to the QBR slide deck: funnel slide, margin slide, truck slide, marketing slide, plus EBITDA waterfall with AI-ROI line items. Owner walks into the QBR with deck pre-built; edits executive summary in 30 minutes; conversation moves to strategy.
• Two failure modes: dashboard skip (owner stops reading because format degraded — fix: reassert the twelve-row text format) and priority paralysis (owner tries to address all off-trend metrics same day — fix: discipline the one-priority rule).