Prepared by Steve Hill for Intermountain Home Services — proposed scope across 8+ brands, reporting to the CMO, marketing through customer service.
"Ship something useful in month two, earn trust, then propose larger architecture in quarter two. Every dashboard answers a specific decision — if I can't name the person making it, it doesn't get built."
Five operating rules that govern what gets built, what gets cut, and how trust compounds over the first 90 days.
A trustworthy dashboard beats an elegant model no one believes.
Month two deliverables are real and live; month three is where we get ambitious.
Reporting without a feedback loop is wasted effort. Every dashboard ties back to a decision.
Build with brand GMs and Jacqueline, never at them.
If I can't name the person making the decision, it doesn't get built.
Three phases, 90 days, each with weekly detail and a deliverables contract at the end.
Understand the org, map the data, and let stakeholders tell me where the pain is before proposing a single dashboard.
Shadow the marketing org across 8+ brands. 30-min meetings with each brand GM. Sit in on one CS call review and one marketing standup. Understand where decisions happen and where friction lives.
Map every system per brand: ad accounts (GA, Google Ads, Meta), CRM/lead platform, call tracking, CS ticketing, billing, web analytics. Document schema and access. Mechanical work — but where gaps surface.
60-min interviews with Jacqueline, Blythe, 3–4 brand GMs, and a CS lead. Key question: "What 2–3 decisions are you trying to make that data doesn't support today?" Listen for patterns across brands.
Review existing dashboards, recurring reports, ad-hoc requests. What's consumed weekly? Ignored? Duplicated across brands? Document the "report graveyard" — tools built for last quarter's question.
This is where trust gets earned — real dashboards live in Power BI, standardized taxonomy in flight, and a weekly KPI cadence Jacqueline can rely on.
Power BI dashboard: first-touch attribution by channel and brand. Spend, leads, attributed revenue by source. Shared weekly. Establishes credibility and proves data can flow.
Standardize campaign tagging and lead flow. Propose one schema that works for all 8+ brands without forcing anyone into a box. This is political — do it carefully.
Recurring Friday dashboard to Jacqueline and brand GMs. Lead volume, conversion, CAC, spend, revenue. Nothing fancy — just what we said we'd measure.
Close the loop: which lead sources convert to completed service, which don't. Overlay win/loss and completion rates on the attribution dashboard. Where marketing meets reality.
Iterate on the attribution dashboard and KPI deck. Fix data issues. Strengthen trust before layering on sophistication.
With the foundation live and trusted, the job shifts from surfacing numbers to recommending moves — multi-touch attribution, CAC/ROAS benchmarking, and an incrementality test plan queued for Q3.
Add a second layer: first-touch, last-touch, linear side-by-side. Lets Jacqueline and brand GMs see influence, not just first-click noise.
Comparative analysis by brand. Which brands are over-investing relative to yield? This is the question Jacqueline likely cares about most.
Propose testing framework for paid-search branded terms (notoriously inflated). Outline methodology for measuring true incremental lift and ROI. A Q3 execution plan.
Identify one recurring manual report (brand KPI pull or ad-spend reconciliation) and automate it. Document and hand off to a brand ops person or Jacqueline's team.
Synthesize 90 days of learning into a Q3 plan: which brands need deeper measurement, incrementality readiness, MMM opportunity, and schema investments.
Success conditions and the risks I'd actively manage so neither becomes a surprise at Day 45.
Each GM may argue "our business is different." Mitigation: show comparative insight first (cross-brand benchmarks), then propose schema as benefit, not mandate.
UTM tagging broken, CRM schema inconsistent, CS missing lead source. Mitigation: ship MVP with caveats documented. Fix quality over time — don't let perfect be the enemy of useful.
If CS can't tie completed service back to the original lead source, the feedback loop breaks. Mitigation: phase 1 directional (flagged as estimates); phase 2 adds instrumentation.
Three specifics about Intermountain — not generic analytics wisdom — that make this plan the right fit for this seat, at this moment.
One standardized attribution model and shared taxonomy unlock cross-brand benchmarking that brand GMs can't generate alone. This is Intermountain's competitive advantage as a holding company.
With a paid-search and performance-marketing background, you already think in CAC, ROAS, and incrementality. I don't need to sell why — only execute fast.
Most home-services companies measure marketing in isolation. Closing the loop to CS completion and profitability is where the real insight lives — from "we spent X on leads" to "we profitably acquired Y service-revenue customers."
This plan is a hypothesis, not a contract. The first 30 days will tell us what to keep, cut, or re-sequence. I'm looking forward to talking through it.