Propane Delivery Software

Propane delivery software for knowing which tanks to fill before they run dry.

The fill list lives in one person's head, the routes get drawn on a whiteboard, and the invoice never quite matches the meter reading. We turn that into one system built around your delivery operation, so forecasting, routing, tickets, inspections, and billing all read from the same customer record.

See our work

Bring one stuck workflow. Leave with a build, buy, repair, or wait recommendation.

Trusted by

Perceptional logoMusgrave GroupUrShipper logoBrux Dental SolutionsBella Skin Institute LogoEnergia RewardsDraftly logoTuneClub LogoSekou LMS logoLogo of food order management app gulaSnelwegDealsGrubly logoPSi logoInstantor Rewards logologo of Mobile app for events, membership clubs, and communitiesAldiFest retail campaign logoVidmattic logoEMS Connect logoWorx Squad logologo of Online Web App For Making Intrologo of Referral and Viral Marketing PlatformConcurrences logoGE logoTicketstop logo

The brief

Start with what is not working.

Good software decisions begin with the constraint, not a list of features or a preferred technology.

01

Nobody knows which tanks need a fill this week until a customer rings with the gauge already low.

02

Bobtail routes get planned by habit, so trucks cross each other and burn miles they did not need to drive.

03

The delivery ticket, the meter reading, and the invoice disagree, and the office finds out when the customer does.

Plain answer

Propane delivery software built around your company forecasts when each tank needs a fill, plans bobtail routes, and ties every delivery ticket and meter reading to the invoice. RaftLabs builds that system for propane companies whose generic tools cannot see tank levels, inspections, or contract pricing in one record.

What to remember

  • Use this when customers run out, or nearly run out, because nobody saw the fill coming.
  • The first release should take one delivery from forecast to paid invoice, not collect disconnected features.
  • The tank record, the route, the ticket, and the invoice should all read from the same customer record.
  • Quote your own team's screens as their own scope: dispatch, exception handling, and billing review are a second product, not a single admin-panel line.
  • Hire freelancers for one small, well-defined task; use one accountable party when the work needs coordination across screens, rules, data, and system connections.

Fit

Custom software should earn its place in a delivery operation.

A feature request is not enough. The case is real when the same problem shows up in run-outs, wasted miles, disputed invoices, or inspection records nobody can find.

A fit

Fills are forecast from memory and habit, so automatic customers still run out and will-call calls arrive in a panic.

Your pricing, contract, and inspection rules are settled enough to write down, and they are how your company keeps customers.

The company can name one person who owns the workflow and what 'done' looks like for the first release.

Not a fit

A maintained fuel delivery product fits once someone sets it up properly and dispatch actually uses it.

The team wants software to settle arguments about pricing or delivery policy that are still changing weekly.

The brief is a feature list with no single delivery path, user, or business result attached.

Not sure which side you are on? Use the first call to make that decision. A useful answer may be to buy, connect, repair, or wait.

Start with the delivery, not the software.

A propane delivery looks simple from the road. From the office it is three jobs wearing one name. The guessing job: which tanks need a fill this week, based on how cold it has been, how fast each household burns, and who is on automatic delivery versus who rings when the gauge is already low. The driving job: getting bobtails to those tanks in an order that does not waste the day. The paper job: tickets, meter readings, inspections, and invoices that all have to agree.

Most propane companies run those three jobs in three places. The forecast lives in one person's head. The route gets drawn up each morning from habit. The ticket gets written in the cab and retyped in the office, where a transposed meter reading becomes a billing argument. Nothing is wrong with any single step. The cost sits in the joins, where the same facts are handled again and quietly drift apart.

A second product hides inside the first. Customers see prices, deliveries, and a bill they trust. Your team lives in the other half: fill lists, route changes, unsigned inspection forms, contract prices that differ by customer, and a record of who promised what. Quotes that squeeze all of that into an "admin panel" line item are where budgets break, so we quote your team's screens as their own job from the start.

Custom propane delivery software makes sense when one owned system can hold the whole delivery better than the patchwork does. The first release should take one delivery from forecast to paid invoice. It should not try to replace every tool you own at once.

We start where the delivery leaks money.

The screen is only one part of the job. The real scope includes your forecasting rules, your pricing and contract terms, the connections to your other systems, who can see and change what, and who runs it after release.

A fill list before anyone runs out

For each automatic delivery tank, the system tracks degree-days, meaning how cold the days have been and for how long, alongside tank size, usage history, and the last fill. It estimates the current level and builds the week's fill list from the tanks that genuinely need it, not from a fixed rotation that fills half-full tanks and misses empty ones.

Routes that do not cross each other

Due tanks are grouped by area and sequenced into bobtail routes that respect truck capacity, so drivers work one patch at a time instead of passing each other on the same road. Urgent fills slot into the nearest active route, and dispatch sees the change at the same moment the driver does.

Tickets and meter readings captured once

At the stop, the driver records the meter reading and gallons pumped against the right customer and the right tank. That single entry prices the delivery ticket, so what was pumped, what was ticketed, and what gets billed start from the same number instead of being copied three times.

Tank records that follow the tank

Every company tank carries its record with it: size, set date, location, rental terms, fill history, and service history. Tank sets, swaps, and pickups update the record once, and nobody has to reconcile a yard list against a billing list against what the driver remembers.

Inspections with a signature on the record

Safety inspections and leak checks are completed at the tank, with the result and the customer's signature captured at the time. The signed record stays on the customer file, ready to produce whenever the question is asked, instead of living on paper in a cab or a filing cabinet. Companies running wider field crews face the same record problem at larger scale in field service management software.

Which one is yours?

Bring it to a 30-minute call. Leave with a build, buy, repair, or wait recommendation in writing. Free, whether or not we ever work together.

The season

Built for how propane demand actually arrives.

Propane demand follows the weather, not the calendar. A mild month forgives a weak system. A cold snap exposes every gap in it at the same time, when you can least afford the scramble.

Automatic and will-call, side by side

Automatic customers are forecast and routed before they get low. Will-call customers are queued by estimated level when they ring, so the genuinely urgent fill is visible in the list instead of being whoever shouted loudest on the phone.

Pricing that matches the contract

Each customer carries their own price: per-gallon rate, contract terms, tank rental, and any agreed charges. The ticket prices itself from that record, so a long-standing customer's contract rate is what lands on the invoice, not the standard rate someone picked from a list.

Billing that matches what was pumped

Invoices are generated from delivery tickets, and tickets come from meter readings taken at the tank. When the numbers disagree, the mismatch is flagged for your office to review before the bill goes out, which is when a billing question is still cheap to answer.

Proof

The fifth attempt became a product customers could keep using.

UrShipper had already worked with four development vendors. Its founder describes the communication and decisions during the recovery.

The case study documents a 14-week recovery, five carrier connections, a Shopify fulfilment workflow, and client-reported results. It is one comparable project, not a promise that every inherited codebase can follow the same plan.

Read the UrShipper case study

Which part of the delivery is costing you the most right now?

Bring it to a 30-minute call. Leave with a build, buy, repair, or wait recommendation in writing. Free, whether or not we ever work together.

How it works

Close one expensive risk before opening the next.

Each stage answers a buyer question and leaves behind a decision you can inspect. The next stage only opens when the expensive uncertainty in the current one is small enough to accept.

  1. 01
    Understand

    Trace one real delivery week

    Where do the run-outs, miles, and billing disputes actually come from?

    Follow one week of deliveries from the fill list to the final payments. Include the run-out nobody forecast, the route drawn from habit, the ticket retyped in the office, and the inspection form that never made it back from the truck.

    Decision produced

    A map of one week from forecast to paid invoice, with the guessing, retyping, and failure points named, and the cost of each worth changing.

    Risk closed

    Paying to automate the visible symptom while the expensive join between forecast, route, and billing stays broken.
  2. 02
    Shape

    Choose the first useful release

    What needs to exist together for one delivery to run end to end?

    Choose the path from the way your deliveries run, not from a feature catalogue. Name the forecast inputs, your pricing and contract rules, the inspection steps, who approves what, and the result the first release must complete.

    Decision produced

    The first-release boundary: one complete path, the users it serves, the rules it encodes, and a clear definition of done.

    Risk closed

    A small-looking feature list that quietly depends on forecasting rules, contract pricing, and three people's habits.
  3. 03
    Prove

    Review working software

    Will it forecast a real week correctly?

    Review the product against your own history, not a tour of finished screens. Run the forecast over a cold week you already lived through and compare its fill list with what actually ran low. Record changed assumptions while there is still time to act on them.

    Decision produced

    Working checkpoints tested with your real tanks, your delivery history, your prices, and the people who will use the system every day.

    Risk closed

    Discovering the gaps in the middle of a cold snap, when every truck and every customer needs the system at once.
  4. 04
    Transfer

    Release, measure, and hand over

    Did the new system improve the original problem, and can another capable team operate it?

    Move one route or one delivery area onto the system first. Compare run-outs, miles per delivery, and billing disputes with the old way before adding the next workflow.

    Decision produced

    A controlled release, comparison with the starting point, operating notes, an account ownership map, and a recommendation for what should happen next.

    Risk closed

    Measuring activity instead of results, or depending on the original vendor to keep the system running.

What it costs

Every project starts at $10K.

The first phase turns the costliest patched workflow into a scoped build with written acceptance criteria.

Starting investment

Starts at $10K

Workflow complexity, systems to connect, data migration, and user roles set the scope. After a 30-minute call, you get a written build, buy, repair, or wait recommendation plus first-phase direction. Free, whether or not you build with us.

Work with us

Bring one stuck workflow. Leave with a decision.

In a 30-minute call, we will help you decide whether to buy, connect, repair, build, or wait. If custom software is not the sensible next move, that is still a useful answer.

  • One named workflow owner and one complete path for the first phase.
  • Scope, price, what 'done' looks like, and exclusions agreed before paid work starts.
  • Code, cloud, data, and provider access remain under your control.
  • Post-launch warranty cover for defects in the agreed scope.

Frequently asked questions

Straight answers to what buyers ask before they scope. Still stuck? .

The system forecasts from each tank's own history. For automatic delivery customers it tracks degree-days, which in plain words means how cold the days have been and for how long, because colder weather burns propane faster. It combines that with the tank size, past usage, and the last fill to estimate the current level and flag the tanks due this week, while there is still time to route them calmly.

Routes are built from the fill list, not from habit. The system groups due tanks by area, sequences the stops so a bobtail works one patch instead of crisscrossing it, and shows gallons against truck capacity before the driver leaves. When a will-call order or an urgent fill lands mid-day, it goes into the route that is already nearest, and the change is visible to dispatch and the driver at the same time.

One record carries the whole delivery. The driver records the meter reading and gallons pumped against the customer and tank at the stop, that reading prices the ticket at the customer's own rate or contract price, and the invoice is generated from the ticket rather than retyped in the office. If the pumped amount and the billed amount ever differ, the exception is flagged for review instead of reaching the customer first.

Every tank your company owns or rents out carries its own record: size, location, who it was set for and when, rental terms, and its full fill and service history. A tank set, a swap, or a pickup updates that record once, so the office, dispatch, and the driver are all looking at the same tank instead of three versions of it in three places.

Each inspection and leak check is recorded against the tank and the address at the time it happens, with the result, the person who did it, and the customer's signature captured on the spot. The signed record stays attached to the customer file, so when anyone asks later whether a check was done and what it found, you can produce the record itself rather than reconstruct it from memory or a paper form in a truck cab.

Will-call customers order their own fills, so a cold snap brings a flood of calls from people whose gauges are already low, all at once. The system cannot change that habit, but it can rank the queue by estimated tank level and days of heat left, show dispatch who is genuinely urgent, and offer your regular will-call customers a move to automatic delivery with the forecast to back it up.

The first paid phase may cover a bounded audit, a proof of concept, or a small first release that takes one delivery from forecast to paid invoice. The final price depends on how deep the workflow goes, how many kinds of users it serves, what it needs to connect to, whether your customer and tank history needs moving over, and whether drivers need phones or tablets in the cab. The agreed phase is priced in writing before it starts.

Timing depends on scope. A first release that covers one complete path, such as forecast to route to ticket to invoice, is a smaller engagement than one that also adds a customer portal, tank monitoring connections, and links to your accounting system. The schedule is set once that first path and what 'done' looks like are clear.

You judge what you can see, not the code. Every checkpoint is working software you click through yourself, tested against the workflow agreed at the start: forecast a week you already lived through and check the fill list against what actually ran out, then follow one delivery from route to invoice. Progress is reported in plain language: what changed, what is next, and what it costs. Our job is to make the work checkable; your job is to check it against your business.

You own the code built for you and control where it lives, the cloud accounts, the data, analytics, and third-party subscriptions. Outside services and open-source pieces keep their own licence terms. The operating notes and handover material stay current so another capable team can take over.