Power BI for Home Services: 6 Templates Contractors Actually Use (And When to Skip It)

Power BI Pro is $14 per user per month, per Microsoft's pricing page. That part is cheap. The templates that actually work in a home service shop are not.

This post walks through six Power BI dashboards that contractors in the $1M-$10M range build and keep using, plus the questions they still cannot answer. The honest framing: Power BI is a great tool if you already have a Microsoft 365 stack and a half-time analyst. If you do not, the sticker price is the smallest part of the bill.

Power BI's pricing tiers are worth understanding before the build conversation. Pro at $14/user/mo lets users publish and share reports. Premium per user (PPU) at $24/user/mo adds paginated reports and larger dataset sizes. Power BI Premium capacity (under the Microsoft Fabric banner) starts at $5,068/month for F64 per Metrica's 2026 breakdown. For most shops under $10M in revenue, Pro is the only tier that makes sense.

For the broader BI tool comparison, see 8 home service BI tools compared. For the question-shape problem that dashboards cannot solve, see 5 questions every contractor asks that no dashboard will ever answer.

1. Revenue by Department Dashboard

Build cost: 20-40 hours at $150/hr = $3,000-$6,000 the first time, per industry rate data in the real cost of building an AI agent for your home service business.

What it shows. Revenue split by HVAC, plumbing, electrical, and install. Time-sliced by day, week, month, year. Compared against prior period. Usually broken down further by tech.

What it does well. This is Power BI's home turf. Pull ServiceTitan or Jobber data via API or an ETL connector, model the star schema, build the visuals, and the result is clean. DAX measures for period-over-period comparisons work exactly the way DAX was designed.

Where it breaks. ServiceTitan and Jobber do not expose the full relational schema. API limits force incremental data pulls that miss edits to historical records. The first time a department is renamed or a new service line is added, the dashboard breaks silently. Budget 4-8 hours of maintenance every quarter.

Who it is right for. Shops with 15+ techs where a daily department-level view drives dispatch decisions.

Text Sully: "Show me revenue by department this month vs last month, and flag anything down more than 10%." Sully pulls it directly from the CRM connection with no ETL build. The question changes next week to "break out HVAC installs by day" and Sully still answers.

2. Technician Performance Scorecard

Build cost: 40-60 hours = $6,000-$9,000.

What it shows. Revenue per tech, average ticket, close rate, callback rate, membership sales. Stack-ranked. Color coded against benchmarks.

What it does well. Power BI's matrix visuals handle tech stack rankings well. Integration with Rilla Voice or CallRail can add conversation-level data if you buy a connector or have a developer pull their API.

Where it breaks. Callback rate is the hardest metric. It requires matching a follow-up job to an original job across time, which most CRM schemas do not model natively. Building that join in DAX takes real work. And once built, it breaks when the CRM adds a new "job type" the logic does not account for.

Tommy Mello's A1 Garage, at $200M+ in revenue, uses Rilla Voice scorecards alongside KPI dashboards. The dashboard shows the what. Rilla shows the why. Two different tools.

Who it is right for. Shops doing monthly tech reviews with KPI bonus structures.

Text Sully: "Which 3 techs have the biggest callback rate increase in the last 60 days, and what do the Rilla transcripts say about their problem jobs?" Sully cross-references both systems in one query.

3. Lead Source + Ad Spend Dashboard

Build cost: 80-120 hours = $12,000-$18,000. Ongoing maintenance: high.

What it shows. Google Ads, Meta, LSA, Angi, and Yelp spend joined to CRM lead sources and closed revenue. Cost per lead, cost per booked job, revenue per ad dollar by channel.

What it does well. With Microsoft Fabric or a Power BI dataflow pulling from a warehouse, this is doable. Google Ads and Meta have native Power BI connectors. Properly built, it is the single highest-ROI dashboard a contractor can have.

Where it breaks. Every ad platform has different attribution windows, data-refresh lags, and schema quirks. LSA and Angi do not offer clean native connectors. Google Ads changes report field names twice a year. LocaliQ's 2025 benchmarks peg non-branded HVAC Google Ads leads at $149 and non-branded plumbing leads at $167, so the stakes on getting attribution right are real. A broken dashboard here costs real money.

Who it is right for. Shops spending $10K+/month across 3+ ad channels with a marketing lead who trusts the numbers enough to act on them.

Text Sully: "Which ad channel had the best profit-per-lead this month, factoring in labor and parts cost on closed jobs?" Sully joins ad spend, leads, jobs, and accounting data in one query. The Power BI equivalent takes a month to build and breaks on the next Google Ads API change.

4. Membership Program Dashboard

Build cost: 30-50 hours = $4,500-$7,500.

What it shows. Active members, new sign-ups, cancellations, member revenue vs non-member, renewal rate, member lifetime value.

What it does well. Membership data is typically clean inside the CRM since it is billing-adjacent. Power BI handles cohort analysis well with PowerQuery and DAX time intelligence.

Where it breaks. Renewal logic varies by shop. Some CRMs auto-renew, some require manual renewal. Some offer tiered memberships with different renewal timelines. DAX can model this but it is custom logic per shop, which means every implementation is bespoke.

The 2026 Jobber Home Service Economic Report notes that retention is now weighted heavier than acquisition in contractor strategy. Membership programs are the cleanest retention signal, and the dashboard matters.

Who it is right for. HVAC and plumbing shops with active maintenance agreement programs.

Text Sully: "Which members are within 60 days of renewal, have not had a visit in 9 months, and spent over $500 on their last non-member service?" Sully filters, ranks, and drafts the outreach text in one conversation.

5. Cash Flow + AR Aging Dashboard

Build cost: 40-60 hours = $6,000-$9,000.

What it shows. Invoices outstanding by age bucket (0-30, 31-60, 61-90, 90+). Cash in vs cash out. DSO (days sales outstanding). Compared against prior month.

What it does well. QuickBooks Online has a reasonable Power BI connector via Microsoft or a third party. Joining AR data with CRM customer data lets you see which aging invoices belong to high-value customers, which is useful for collection decisions.

Where it breaks. QuickBooks' Power BI connector pulls a subset of fields. Anything custom (class tracking, memo fields, attachments) usually needs a custom API build. If you run QuickBooks Desktop, the connector is essentially unusable and you need a third-party sync tool.

Who it is right for. Shops doing $3M+ where AR balances regularly exceed $100K and DSO moves matter to operations.

Text Sully: "Who's 60+ days overdue, owes more than $2K, and had a job closed in the last year that they paid on time?" Sully pulls the list with customer name, amount, last contact date, and payment history.

6. Marketing ROI by Zip Code

Build cost: 60-100 hours = $9,000-$15,000.

What it shows. Map visual of zip codes colored by customer count, average ticket, ad spend reach, and net profit. Used for decisions on where to increase LSA coverage, where to drop billboards, where to run door hangers.

What it does well. Power BI's map visuals with Azure Maps backing are genuinely good. Zip-level rollups happen fast in DAX once the data model is in place.

Where it breaks. Ad platforms only report at regional or DMA level natively, so zip-level attribution requires building CRM-side lead-source inference. That inference is error-prone. A zip code dashboard based on inferred data is 70-80% accurate, which is enough to guide direction but not enough to defend a tech in a team meeting.

Who it is right for. Multi-city operators evaluating market expansion or ad-spend reallocation.

Text Sully: "Which zip codes have the highest average ticket but lowest ad spend this quarter? Rank by opportunity." Sully answers. The Power BI version requires the map visual plus the full attribution model plus someone to interpret it.

When to skip Power BI entirely

Power BI is not the right tool for a lot of contractor questions. Skip it when:

You have fewer than 10 techs and no analyst. The total cost of ownership (licenses plus build plus maintenance) runs $15K-$40K/year once you count analyst time. A 10-tech shop doing $2M in revenue gets marginal value from that spend.

Your primary data lives in tools without good connectors. LSA, Angi, and some voice AI platforms have weak or no Power BI integration. Every missing connector is a custom build.

Your real questions are "why" not "what." Dashboards answer what. Power BI shows you that revenue dropped 12%. It does not tell you why. If you spend most of your analytics time asking why, a dashboard is the wrong tool. See 5 questions every contractor asks that no dashboard will ever answer.

ServiceTitan's own 2026 data backs this up. Their State of AI in the Trades report shows 44% of contractors cite integration complexity as a top AI adoption barrier and 38% cite understanding how to use AI tools as a barrier. Those are the same barriers that stall Power BI builds.

One ServiceTitan user on Capterra summed up the dashboard fatigue directly: "My least favorite thing about ServiceTitan is the reports." ServiceTitan's reporting is essentially a packaged Power BI experience. The complaint applies to both.

Where AI chat fills the gap

The six templates above are the dashboards worth building. The questions that fall outside those six templates are where a conversational AI layer earns its keep.

ServiceTitan AI vs standalone AI for contractors covers the build-vs-buy question when the CRM vendor ships their own AI. The short version: embedded AI is limited to the vendor's data. A standalone AI can reach across the CRM, accounting, email, calls, and ads at once. That cross-source reach is exactly what dashboards cannot do well, and it is what a single chat query can do in 30 seconds.

The real cost math

Building and maintaining the six dashboards above at a 15-tech HVAC shop:

Line item	Cost
3 Power BI Pro seats	$504/yr
Power BI Premium per user (2 seats)	$576/yr
ETL tool (Fivetran or Airbyte)	$6,000-$18,000/yr
Analyst time (part-time, 10 hrs/wk)	$60,000-$90,000/yr
Initial build (6 dashboards)	$40,000-$65,000 one-time
Ongoing maintenance	$15,000-$25,000/yr

Total first-year cost: $120,000-$200,000 for a shop doing $3M in revenue. That is a 4-6% revenue hit on dashboards before you have answered the first "why" question.

A chat AI like Sully reads the same data sources with no dashboard build, no ETL tool, and no dedicated analyst. Different shape of answer, different cost structure.

The two-hire test

Before committing to Power BI, run this test. Can you name the person who will build the dashboards? Can you name the person who will maintain them when a connector breaks?

If both answers are "we will figure it out," Power BI is the wrong tool. The sticker price on licenses is the smallest line item. The build and maintenance hours are what actually kill most home service Power BI projects.

Shops that skip this test end up 9 months in with three half-built dashboards, no one to own them, and a $40K sunk cost. This is the most common failure pattern in contractor BI adoption.

Contractor voices on the Power BI experience

Power BI feedback from actual users is mixed in predictable ways. The Reporting Hub's analysis describes the typical new-user experience as hitting a "brick wall" the moment the dashboard needs to do more than a basic chart.

SQLBI's 7 reasons DAX is not easy is required reading for any contractor weighing whether they can build Power BI dashboards in-house. DAX is the formula language behind every non-trivial Power BI measure. It is powerful, and it is unforgiving. A single missing comma changes an entire calculation.

Reddit discussions on Power BI licensing document businesses repeatedly underestimating the complexity of the Pro vs PPU vs Fabric decision before purchase. Contractor shops that buy the wrong tier end up over-licensed by 30-50% on a typical deployment.

The verdict

Power BI is genuinely good if you have a Microsoft-native stack, a half-time analyst, and questions that fit the dashboard shape. Six templates will carry most shops: revenue by department, tech performance, ad ROI, membership, AR aging, and zip-level ROI.

Build those. Then stop.

Everything else (the ad-hoc questions, the cross-source questions, the "why" questions) belongs in a conversational AI layer on top of the same data. Not because Power BI is bad. Because the question shape is different.

The dashboard shows you that HVAC revenue dropped 12% this month. The chat AI tells you that three senior techs were out for a certification week, two of your top-selling CSRs took PTO, and a Google Ads campaign paused when the billing card expired. Different layer. Different job.