Propane Delivery Route Optimization SaaS for Independent Fuel Dealers
11.9 million American households heat with propane, and most of their deliveries are still scheduled using degree-day formulas from the 1970s. A single wasted stop costs $75-150. Tank monitors can cut stops by a third, but fewer than 15% of residential tanks have one. Nobody has combined real-time tank data with route optimization in a single platform.
The Problem
Propane is the forgotten last-mile logistics problem. According to the National Propane Gas Association (NPGA), propane is used in 50 million American homes, with 11.9 million households relying on it as their primary heating fuel. A 2018 PERC study found the US propane industry generates $46 billion in direct economic impact annually and employs over 57,000 workers. The vast majority of residential propane reaches customers the same way it has for decades: a bobtail truck drives to your house, hooks up a hose, and fills your tank.
The scheduling of those deliveries is where the problems start. Most propane dealers use one of two methods. "Will-call" means the customer calls when they're running low. The dealer scrambles to dispatch a truck, often to a single stop outside the normal route, burning $75-150 in delivery cost for a 150-gallon fill that generates maybe $50 in gross margin. "Automatic delivery" means the dealer estimates when a tank will need filling based on degree-day calculations: a formula that correlates heating demand with outdoor temperature, calibrated to each customer's historical usage. This method was a genuine innovation in the 1980s. It's now 40 years old and still the industry standard.
The problem with degree-day forecasting is that it guesses. It doesn't know the customer installed a heat pump last summer. It doesn't know the in-laws visited for two weeks and cranked the thermostat. It doesn't account for the increasingly erratic weather patterns that make historical degree-day averages unreliable. When the algorithm guesses wrong in one direction, the truck arrives at a tank that's still 60% full, delivering 200 gallons instead of 500, which means the delivery cost stays the same but the revenue per stop drops by more than half. When it guesses wrong in the other direction, the customer runs out, calls in a panic, and the dealer sends an emergency delivery that destroys the day's route efficiency.
According to LP Gas Magazine's analysis, the difference between arriving at a tank at 40% full versus 20% full is 33% fewer delivery stops for the same total gallons delivered. On a 3,200-gallon bobtail making stops at 1,000-gallon tanks, that's the difference between 6.4 stops per load and 4.3 stops. For a dealer running 5 trucks through heating season, that's thousands of eliminated stops per year, at $75-150 per stop.
The Gap in the Market
Software exists for propane dealers. But the market is fragmented into point solutions that don't talk to each other, and legacy platforms built before smartphones existed.
| Company | What They Do | What's Missing |
|---|---|---|
| ADD Systems | The dominant back-office ERP for fuel dealers. Founded 1973. Handles billing, accounts receivable, tank inventory, and basic degree-day scheduling. Installed base of hundreds of dealers. | On-premise or hosted legacy architecture. No native tank monitoring integration. No real-time route optimization. Pricing starts around $50K+ for implementation. The company's bread and butter is accounting, not logistics intelligence. |
| Cargas Energy | Cloud-based fuel delivery management. CRM, dispatch, billing, and driver mobile app. Integrates with Tank Utility for monitoring data. Based in Lancaster, PA. | Better UX than ADD, but still primarily a back-office system. Tank monitor integration exists but doesn't drive automated route building. Route optimization is basic. No wholesale procurement intelligence. |
| Tank Utility (Generac) | Wireless propane tank monitors. Cellular-connected gauges that report tank levels to a cloud dashboard. Raised seed funding with strategic participation from Generac, later acquired. Hardware: ~$200/monitor. | Pure monitoring play. Tells you the tank level but doesn't build the delivery route. Doesn't integrate deeply with dealer dispatch systems. The dealer still has to manually translate tank readings into delivery decisions. |
| Blue Cow Software | Fuel delivery management with dispatch and driver tracking. Divested by REPAY in 2023. Serves heating oil and propane dealers. | Changed ownership, uncertain product roadmap. Limited tank monitor integration. No degree-day replacement or predictive routing. |
| Trimble (TMW) | Enterprise fuel logistics. Introduced a heating fuel delivery module for their transportation management system in 2018. | Enterprise pricing and complexity. Designed for large fleets with dedicated IT staff. A 5-truck independent propane dealer isn't going to implement Trimble. |
The pattern: back-office systems don't optimize routes. Route systems don't integrate tank monitors. Tank monitors don't build delivery schedules. And the only company that tried to bridge all three (Tank Utility) got acquired by Generac, a generator company that cares about monitoring as a feature for generator fuel management, not as a platform for propane delivery optimization. The independent dealer with 5-15 trucks, 3,000-8,000 customers, and no IT department is cobbling together spreadsheets, phone calls, and 40-year-old degree-day tables.
The Solution
A single platform that replaces the degree-day clipboard with data-driven delivery intelligence:
1. Tank monitoring integration layer (hardware-agnostic): Rather than selling proprietary hardware, the platform integrates with existing tank monitors from Tank Utility, Otodata, SkyBitz, and other cellular gauge manufacturers via their APIs. For unmonitored tanks, the system runs an improved degree-day model that incorporates National Weather Service forecast data (not just historical averages), customer-reported lifestyle changes, and delivery history pattern recognition. The goal is to know every tank's level within 10% accuracy, whether monitored or modeled.
2. Predictive delivery scheduling engine: Instead of building routes the morning of delivery, the system projects tank levels 7-14 days forward and clusters deliveries by geography, minimizing total miles driven per gallon delivered. The optimizer targets a fill threshold of 20-25% remaining (the sweet spot where tanks are low enough to accept a large, profitable delivery but not so low that runout is imminent). For a 1,000-gallon tank, that means delivering 550-600 gallons per stop instead of the industry-typical 300-400. Fewer stops, more gallons per stop, lower cost per gallon delivered.
3. Dynamic route builder with driver mobile app: Each morning, drivers get an optimized route on their phone with turn-by-turn navigation, estimated gallons per stop, and any special instructions (gate codes, dog warnings, access notes accumulated from previous drivers). Real-time adjustments when emergency will-call requests come in, automatically inserting the call into the nearest driver's route rather than dispatching a separate trip.
4. Wholesale procurement timing alerts: Propane prices fluctuate daily based on Conway and Mont Belvieu hub pricing. The platform tracks rack prices from multiple suppliers in the dealer's region, inventory levels in their bulk storage, and projected delivery volume over the next 7-30 days. When the model identifies a buying opportunity (rack price dip coinciding with adequate storage capacity and upcoming high-demand weather), it alerts the purchasing manager. Even a $0.05/gallon improvement on a 30,000-gallon transport load is $1,500.
5. Customer self-service portal: Customers check their own tank level (if monitored), see their estimated next delivery date, and request early fills without calling the office. Eliminates the #1 source of inbound phone calls for most propane dealers. Includes autopay enrollment, delivery confirmations via text, and seasonal budget billing setup.
The Math: What Wasted Stops Actually Cost
Take a mid-size independent propane dealer in rural New Hampshire. 5,000 residential customers, 8 bobtail trucks, delivering approximately 2.5 million gallons per heating season (October through March). The average residential propane price in New England was $3.77/gallon as of March 2026 (EIA weekly survey), with wholesale at approximately $1.16/gallon. Gross margin: $2.61/gallon.
Current state (degree-day scheduling):
Average delivery: 350 gallons to a 1,000-gallon tank (arriving at ~45% full). 2.5M gallons ÷ 350 gallons/stop = 7,143 delivery stops per season. At an estimated $100/stop fully loaded (driver time at $35/hour, truck operating cost, fuel, insurance, amortized vehicle cost), total delivery operations cost: $714,300/season.
Optimized state (monitor-informed scheduling):
Average delivery: 575 gallons to a 1,000-gallon tank (arriving at ~22% full). 2.5M gallons ÷ 575 gallons/stop = 4,348 delivery stops per season. Same $100/stop: total delivery cost: $434,800/season. Route optimization reduces per-stop cost by an additional 10-15% through better sequencing (less windshield time between stops). Conservative 10% reduction: $43,480.
Annual savings: $322,980. That's $279,500 from eliminated stops plus $43,480 from route efficiency. On 2.5 million gallons, that's $0.129/gallon in delivery cost reduction. When your gross margin is $2.61/gallon, a $0.13 efficiency gain is a 5% margin improvement.
Add procurement timing optimization. If the dealer buys 2.5M gallons per season across roughly 83 transport loads (30,000 gallons each), and better timing saves $0.03/gallon on average, that's another $75,000/season. Total annual value: approximately $398,000.
This is our original calculation, and we believe it represents a reasonable estimate for a dealer of this size. The inputs (delivery gallons, stop costs, fill levels) are drawn from industry sources cited throughout this article. The actual savings will vary substantially depending on the dealer's geography, customer density, current operational efficiency, and the proportion of tanks that can be monitored.
Revenue Model
Picture the dispatcher's morning today at a 5,000-customer dealer: she arrives at 5:30 AM, pulls up a spreadsheet of degree-day projections, cross-references yesterday's will-call list, and starts manually building routes on a printed map. By 6:15, the first driver is already asking which stops to hit. Now picture the alternative: the system built last night's routes automatically based on projected tank levels, weather forecasts, and geographic clustering. The dispatcher reviews and approves. Drivers open their phones and go. That shift from reactive to predictive is what the platform monetizes.
| Revenue Stream | Amount | Notes |
|---|---|---|
| Platform SaaS (per truck/month) | $299/truck/month | Core platform: scheduling engine, route optimization, driver app, customer portal. Annual contract with monthly billing. |
| Tank monitor data integration | $2/monitored tank/month | API fees for ingesting and processing tank level data from third-party monitors. Hardware-agnostic. |
| Procurement intelligence module | $499/month flat | Wholesale price tracking, supplier comparison, buy-signal alerts. Phase 2 product. |
| Customer portal white-label | $199/month | Branded customer self-service portal with tank levels, delivery scheduling, autopay. Reduces inbound calls by 30-50%. |
| Implementation + data migration | $2,500-5,000 | One-time. Migrate from legacy ADD/Cargas systems, import customer database, configure routes. |
Unit economics on an 8-truck dealer with 1,200 monitored tanks: Monthly recurring: 8 × $299 + 1,200 × $2 + $499 + $199 = $5,090/month = $61,080/year. Implementation: $5,000. For a dealer saving $398,000/year, the platform cost represents 15.3% of the savings. LTV at 5-year retention: $305,400 in SaaS revenue. Customer acquisition cost via industry trade shows (Eastern Energy Expo, NPGA Southeastern Convention), LP Gas Magazine advertising, and NPGA channel partnerships: estimated $8,000. LTV:CAC ratio: 38x.
Market Size
TAM: The IBISWorld fuel dealers industry report (NAICS 454312) identifies approximately 7,200 fuel dealer establishments in the US (including heating oil and propane). LP Gas Magazine's 2024 Top Propane Retailers ranking shows the top 3 alone (AmeriGas, Ferrellgas, Suburban Propane) operate 2,883 outlets with 8,800+ bobtails. Industry estimates place the total number of propane-focused retail operations at 3,000-5,000 companies. At an average fleet size of 6 trucks and $299/truck/month SaaS, the core platform TAM is 3,500 companies × 6 trucks × $299/month × 12 months = $75.3M/year. Adding tank monitoring integration, procurement modules, and customer portals brings the total addressable to approximately $120M/year.
SAM: Independent dealers with 3-20 trucks (too small for enterprise solutions like Trimble, too large to manage on paper). Approximately 2,200 companies. SAM: 2,200 × 6 trucks × $299/month × 12 = $47.3M/year for core platform alone.
SOM (year 3): 120 dealers averaging 7 trucks each at $5,000/month blended ARPU = $7.2M ARR. 5.5% penetration of SAM.
Why Now
Consolidation is forcing efficiency. The propane industry is consolidating rapidly. LP Gas Magazine tracked significant M&A activity in 2024-25, continuing a decades-long trend of larger players acquiring independents. The dealers who survive as independents need the same operational efficiency that the nationals achieve through scale. Software is how a 5-truck dealer matches the per-gallon delivery cost of AmeriGas's 2,640 bobtails.
Tank monitor costs have cratered. Five years ago, a cellular tank monitor cost $300+ and required a $15/month data plan. Today, devices like Tank Utility's standard gauge retail for under $200 with connectivity included. At 2% annual market penetration growth, the installed base of monitored residential propane tanks is growing steadily. But the monitoring data is only valuable if it feeds a delivery optimization system. The monitors are the sensor layer. The opportunity is the intelligence layer on top.
Driver labor is scarce and expensive. CDL-B drivers with hazmat endorsements (required for propane delivery) are increasingly difficult to hire. The national CDL driver shortage has been well-documented, and propane delivery adds seasonal demand compression: 80% of residential deliveries happen in 6 months. When you can't hire another driver, the only option is making each driver more productive. Eliminating 33% of stops means your existing 8 drivers do the work that previously required 12.
The enabling tech stack finally exists at indie-dealer prices. Open-source vehicle routing solvers (Google OR-Tools, VROOM, OSRM) can now compute optimized multi-stop routes in milliseconds on commodity cloud hardware. Five years ago, building a real-time delivery optimizer required licensing enterprise logistics software or hiring operations research PhDs. Today, a small engineering team can build on top of freely available TSP and VRP solvers, pair them with weather APIs from the National Weather Service (free) and tank telemetry from cellular IoT platforms, and run the whole thing on a $500/month cloud budget. The infrastructure cost of building this product dropped by an order of magnitude between 2020 and 2025.
Warm winters are breaking degree-day models. The 2023-24 winter was the warmest on record across much of the US (EIA winter fuels outlook). Degree-day models calibrated on 10-year averages consistently over-predicted demand, sending trucks to tanks that didn't need filling. Climate variability makes historical statistical models less reliable every year, creating demand for real-time measurement.
Startup Costs
| Category | Cost | Notes |
|---|---|---|
| Software engineering (scheduling engine + route optimizer, 10 months) | $320K | 2 backend engineers + 1 frontend + 1 mobile developer. Core algorithms: tank level prediction, delivery clustering, route TSP solver. Integration with Tank Utility, Otodata, SkyBitz APIs. |
| Mapping and routing infrastructure | $40K | Google Maps Platform or Mapbox for geocoding, routing, and driver navigation. Annual licensing. |
| Pilot program (10 dealers, subsidized onboarding) | $50K | Free implementation + first 3 months for early adopters in New England and Upper Midwest. Travel for on-site support. |
| Industry trade shows and marketing (year 1) | $35K | Eastern Energy Expo, NPGA Southeastern Convention, LP Gas Magazine advertising, state propane association meetings. |
| Data acquisition and weather APIs | $15K | NWS forecast data (free), historical degree-day databases, wholesale propane price feeds. |
| Cloud infrastructure (year 1) | $24K | AWS/GCP hosting for route optimization compute, real-time data ingestion from tank monitors, driver app backend. |
| Legal and compliance | $16K | Software licensing agreements, dealer contracts, data privacy compliance. NIST guidelines for IoT data handling. |
| Total | $500K |
Limitations
The 33% delivery stop reduction figure comes from LP Gas Magazine's tank monitoring analysis, which models the math for arriving at tanks at 20% full versus 40% full. This represents a theoretical maximum for tanks with monitors installed. In practice, most dealers will have a mix of monitored and unmonitored tanks for years during the transition. A dealer with 20% of tanks monitored might see a 7-10% stop reduction in year one, not 33%. The full savings materialize only at high monitor penetration, which requires the customer or dealer to invest $150-200 per tank in hardware that we don't control or sell.
The $100/stop delivery cost estimate is an industry-cited figure that includes fully loaded driver labor, vehicle depreciation, fuel, and insurance. Actual costs vary enormously by geography. A dealer in densely settled suburban Connecticut with 3-mile stop spacing has very different per-stop economics than a dealer in rural Montana with 15-mile gaps between customers. Route optimization savings are proportionally larger in spread-out territories, but the baseline delivery cost is also higher, making the ROI calculation dependent on local conditions.
Our TAM calculation assumes 3,500 propane-focused retail companies. This is an estimate; the actual count is not centrally tracked. The IBISWorld fuel dealer figure of 7,200 includes heating oil dealers (concentrated in the Northeast), which face a structurally declining market as oil-to-gas conversions accelerate. If we're overstating the propane dealer count by 30%, the TAM drops to $84M, which is still a large market for a vertical SaaS startup.
The addressable market faces a structural headwind from building electrification. The Inflation Reduction Act provides up to $8,000 in heat pump rebates, and several states are pushing electrification building codes. Propane's 11.9 million household base will likely shrink over the 10-15 year horizon a startup needs to reach full maturity. The near-term opportunity is real (heat pump adoption is slow in rural areas without natural gas infrastructure, which is exactly where propane dominates), but a company building for this market should plan for a customer base that plateaus or contracts by the early 2030s rather than grows indefinitely. This makes efficient market capture more urgent, not less.
Wholesale procurement optimization assumes the dealer has some flexibility in timing purchases, which is not always the case. During peak winter demand, dealers may be supply-constrained and buying at whatever price is available. The $0.03/gallon average savings figure is our estimate based on observed rack price volatility; it is not derived from a controlled study.
Strongest Counterargument
ADD Systems has been selling software to fuel dealers since 1973. They have deep relationships with hundreds of dealers, decades of domain expertise, and an installed base that most companies would struggle to displace. If route optimization and tank monitor integration become table-stakes features, ADD can build (or acquire) them and ship them to their existing customer base with zero new sales effort. Their 50-year head start in understanding fuel dealer workflows, billing complexity, regulatory requirements, and seasonal cash flow patterns is a moat that no startup can replicate in a few years of coding. The switching cost for a dealer already on ADD is enormous: migrating billing history, customer records, tax configurations, and driver workflows is a multi-month project that most operators won't undertake for a better routing algorithm.
The response: ADD's architecture is its anchor. Their platform was built for on-premise deployment in an era of Windows servers and local databases. While they've added hosted options, the core product reflects its 1973 origins in data model and user experience. More critically, ADD's business model is back-office software: billing, receivables, tank inventory tracking. Delivery optimization is a logistics problem that requires real-time data ingestion, geospatial computation, and mobile-first interfaces — capabilities that are architecturally distant from an ERP system. ADD could acquire a routing startup and run it as a separate product — Intuit did this successfully with Mailchimp, and plenty of legacy software companies have modernized through bolt-on acquisitions. That path is real, and a startup in this space should assume ADD is watching. The bet is on execution speed: can a focused startup build and prove the delivery intelligence layer faster than ADD can build, buy, or partner its way there? History suggests the focused startup moves faster in the first 3-5 years, but the window closes once the market is proven. The practical strategy is interoperability, not displacement: build as the intelligence layer that makes ADD's billing data smarter, and make acquisition by ADD (or a national propane distributor) a feature, not a bug.
What You Can Do
If you're an independent propane dealer: Start measuring your delivery efficiency today. Pull your last 1,000 delivery tickets and calculate your average gallons per stop. If it's below 400 on 1,000-gallon tanks, you're arriving too early. Next, calculate your cost per stop: total delivery department costs (driver labor, vehicle expenses, fuel) divided by total stops. If it's above $80, route optimization should be your top operational priority. You can capture 10-15% of the savings with nothing more than a spreadsheet: sort your automatic delivery customers by expected fill date and geography, then build routes that cluster nearby customers on the same day. It's tedious but free.
If you're building this: Start with 5 pilot dealers in New England, where heating season is longest, propane prices are highest ($3.77/gallon average retail), and margins have the most room for optimization. Build the scheduling engine first — the tank level prediction and route optimization core. Integration with existing back-office systems (ADD, Cargas) is essential for adoption. No dealer will rip out their billing system. Position as the intelligence layer that sits between the tank monitor and the dispatch system. The beachhead is not "replace ADD" but "make ADD smarter."
If you're an investor evaluating this space: Vertical SaaS for propane delivery is pre-category. No venture-backed company is building a delivery intelligence platform for independent fuel dealers. Tank Utility proved the monitoring hardware market but was absorbed into Generac's ecosystem. ADD Systems and Cargas serve this market profitably but focus on back-office functions rather than delivery logistics intelligence. The projected unit economics look strong on paper ($5K/month ARPU on a customer saving $30K+/month), but both the ARPU and savings figures assume full module adoption and high monitor penetration, which are unproven at scale. The 5-year retention assumption underlying the LTV calculation is speculative for a market that has never used subscription delivery software. Counterbalancing that risk: the strategic exit market includes AmeriGas/UGI, Ferrellgas, and Suburban Propane, all of which need delivery optimization and have a history of acquiring operational technology. The key diligence question is whether an independent dealer's willingness to pay for logistics software survives a mild winter when the pain of inefficiency is less acute.
The Bottom Line
Propane delivery is a $46-billion industry running on logistics technology from the Carter administration. Every winter, thousands of bobtail trucks drive millions of miles visiting tanks that don't need filling yet, while other customers run out and make panicked phone calls. The sensor technology to fix this exists. The routing algorithms exist. The mobile platforms exist. What doesn't exist is a single product that combines real-time tank monitoring data, predictive delivery scheduling, dynamic route optimization, and wholesale procurement intelligence into one platform designed specifically for the independent fuel dealer who runs 5-15 trucks and can't afford a Trimble implementation. The propane industry spent the last decade automating its back office. The next decade is about automating its trucks.