Keep fuel, vehicle meters and service work together.

In this guide

Evaluate a fleet maintenance system with one vehicle

AI-generated editorial illustration; not a customer photograph or product screenshot.

A useful trial should answer a specific question: can the next person make the correct service decision from the vehicle record, without phoning the person who last worked on it? Use one ordinary vehicle and one difficult record, such as a vehicle with a replacement odometer or incomplete history. A clean demonstration record alone will hide migration problems.

For a mixed fleet, decide whether each asset is maintained in kilometres, miles or engine hours. Keep that unit consistent from the source reading through the service plan. A generator sharing the workshop may belong in the maintenance system even though it has no road mileage. Do not convert hours into distance to make a dashboard look uniform.

A worked service example

AI-generated editorial illustration; not a customer photograph or product screenshot.

This is an illustrative planning example, not a recommended service interval. Suppose the approved service plan says a task is next due at 85,000 km. The current reading is 84,720 km, leaving 280 km to the trigger. A later reading of 85,030 km puts it 30 km beyond the trigger. In the trial, check the asset reading, the due work and the resulting job history. A reminder appearing is only the first half of the test; the technician must be able to record what was done and a manager must be able to find it later.

If the incoming source says 52,835 miles but the asset is configured in kilometres, review the mapping before syncing. Incorrect units can advance maintenance by thousands of kilometres. If the source device has been replaced, establish how the new baseline relates to lifetime usage rather than accepting a lower number silently.

Compare maintenance scope before comparing prices

Fleetio's maintenance offering explicitly separates in-house and outsourced workflows. That is a useful evaluation distinction. PreventiveHQ currently focuses on maintenance records, fuel entry and selected Samsara usage readings; it does not provide a maintenance-shop payment network.

What fuel data can and cannot prove

For an illustrative full-to-full interval, 420 km travelled divided by 35 litres gives 12 km/litre, or 8.33 litres/100 km. A partial fill does not provide the same evidence. Fuel spend also excludes tyres, repairs, depreciation, financing and insurance. Compare the same measurement basis before drawing conclusions about operating cost.

Record those distinctions in your pilot worksheet. Measure whether service records are complete, readings are current and overdue work has an owner. A short pilot cannot establish a durable reduction in fleet downtime.

Use the fleet service worksheet, review the Samsara setup guide, and compare workspace limits using your real vehicle and technician counts.

What fleet maintenance software should help you decide

The first buying question is not how many charts a product contains. It is which decisions are currently delayed or made with unreliable information. A small delivery operation might struggle to know which vehicle can be allocated tomorrow. An equipment owner might have accurate service dates but poor records of engine hours. A workshop may know exactly what is due while still losing time because the required parts were never received.

Write down three recurring decisions and the records each requires. For vehicle availability, that may include unresolved defects, scheduled work and a responsible person's release decision. For service planning, it includes the approved interval, the current reading and the last relevant work. For purchasing, it includes the needed part, current quantity, outstanding order and expected delivery. A useful trial shows whether these records can be found and trusted.

Then separate must-have capabilities from conveniences. A live dispatch map can be essential for a transport control room while offering little help to a mechanic reconstructing a repair. A complete accounting integration can be mandatory for one business and unnecessary for another that only wants reliable service records. Making that distinction before a demo prevents an attractive feature list from replacing your operating requirements.

Choose the operating model before choosing the product

When your own team performs the work

An in-house workshop needs a dependable handoff from the person reporting the problem to the person planning and performing the repair. Evaluate job assignment, equipment identification, parts use, time recording and closeout together. If the technician must copy the same information into a separate spreadsheet at the end of the day, include that remaining administrative effort in your assessment.

Have a technician use the product on the device and network they actually have. A desktop demonstration does not establish that the same task is comfortable on a phone. Test long asset names, a realistic procedure and an incomplete job. If offline use is required, verify the exact supported app, synchronization behavior and conflict handling; do not infer offline availability from the phrase “mobile friendly.”

When service is outsourced

An outsourced maintenance process has a different control point: the authorization of external work. Your team may need to approve estimates, retain supplier documents and reconcile work that a third party performed. Decide who can authorize additional spend and how the final service evidence returns to the vehicle record.

PreventiveHQ can hold maintenance work and historical records, but its fleet connector does not provide a network of external workshops or automatically pay them. If you need those services, compare a specialist offering against that requirement. Avoid evaluating two products as equivalent merely because both display a vehicle history.

When vehicles and stationary equipment share a team

A small business may maintain vans, forklifts, compressors and generators with the same technicians. A maintenance workspace can be useful when these assets share procedures, parts and work-order practices. Their operating measures should remain distinct. Road distance, engine runtime and production counts describe different kinds of use.

Keep separate plans where the maintenance policies differ, even when two assets belong to the same broad category. The value of a common workspace is that the team can find the work and its evidence; it should not force different equipment into one inappropriate interval or checklist.

Establish a vehicle register that survives migration

Migration is often the first serious test of the software. Real registers contain duplicate names, retired vehicles, incomplete serial numbers and service notes mixed into identification fields. Take an unchanged copy of the source before cleanup. Record the decisions that resolve duplicates, because the same decision will affect any imported service history.

Choose a stable asset code for each physical vehicle. Keep registration numbers, VINs and descriptive names as separate references where supported. Registration or fleet allocation can change, while the maintenance history still needs to refer to the same physical equipment. If a spreadsheet row represents a trailer and tractor together, decide whether they should own separate service histories before import.

Use a small batch containing the difficult cases. Verify a vehicle with leading zeros in its code, one with an existing usage reading and one retained only for historical work. Confirm that the import associates each record with the intended site. A successful upload is not sufficient: compare the resulting records with the source and have someone other than the importer perform the check.

Move identification, schedules and completed work through their respective formats. The asset register creates identity; the schedule import establishes future maintenance; the history import preserves completed work. Keeping them separate makes it easier to reconcile counts and avoid accidentally turning an old service entry into a new active job.

Design a defect-to-repair handoff

AI-generated editorial illustration; not a customer photograph or product screenshot.

A useful defect report captures an observation, not an unsupported diagnosis. For example, a driver can describe when a new noise occurs and identify the vehicle without deciding which component must be replaced. Preserve that original observation so the technician can compare it with the eventual finding.

During triage, decide what work is authorized, who owns it and which approved procedure applies. If a vehicle's operational use needs to change, follow the organization's established decision process. A task status in software is not, by itself, an engineering assessment or a regulatory release. The people responsible for those decisions need an understandable record.

At closeout, distinguish what was found, what was done and how the result was verified. “Repaired” gives the next technician less useful information than a concise account of the finding, action and verification. Include the relevant time and parts through the available work-order workflow. If the symptom returns, the earlier record should help the team investigate rather than restart from memory.

For a trial, deliberately include one job that cannot immediately be completed. Perhaps the part is unavailable or access is delayed. Review how the team records that situation and who follows it up. Software that only performs well when every job closes immediately may not solve the handoffs causing your actual backlog.

Connect telematics without hiding data-quality problems

An integration can remove repeated entry, but it also creates a dependency on source permissions, mapping and freshness. Document which provider statistic feeds each maintenance asset. A provider's vehicle name may look familiar while its stable identifier points to a different unit. Check the identifier and a recent known reading before enabling the mapping.

The PreventiveHQ Samsara connector imports selected odometer or engine-runtime readings into compatible asset meters. Customers supply their own authorized token. Eligible connected workspaces are polled on a fifteen-minute schedule, and a manager can run a manual sync. A scheduled poll is not a promise of a new reading every fifteen minutes: the provider may have no newer sample, permissions may have changed or connectivity may fail.

Samsara's planned-maintenance documentation describes the source statistics used for usage-driven maintenance integrations. The important implementation questions are which statistic is selected, what unit it uses and how the receiving system handles repeated or out-of-order observations. Those questions matter more than displaying a provider logo on the page.

Test the connection failure path as well as the initial success. Know where the last-sync information and safe error message appear. Assign someone to review stale mappings and unexpected readings. If credentials change, verify the mappings again rather than treating the previous authorization as permanent evidence that the connection remains correct.

Keep parts availability separate from purchasing activity

AI-generated editorial illustration; not a customer photograph or product screenshot.

Ordering a part does not mean it is physically available. The maintenance team needs to distinguish current stock, the quantity ordered and the quantity actually received. A supplier confirmation can support planning while still leaving the job waiting for delivery.

For an illustrative test, order ten filters and receive six. The stock record should reflect six received filters, with four still outstanding. Repeat the same receipt submission to check that it does not double the delivered quantity. Then cancel the remainder and verify that the six received filters remain in stock. This is a more useful purchasing demonstration than simply creating a printable order.

Agree on units before testing. A box, a litre and an individual item are not interchangeable. If the supplier sells packs while technicians use individual pieces, define how your team converts and records the quantity. Test a fractional quantity when the stock unit genuinely permits one rather than rounding it away for convenience.

PreventiveHQ purchasing supports suppliers, draft and ordered purchase records, partial receiving and stock updates. Xero export creates a draft purchase order for review; it does not settle the supplier's invoice. Keep tax, freight and accounting reconciliation in the appropriate process and verify the exact boundary in the purchasing guide.

Measure cost and downtime on a consistent basis

AI-generated editorial illustration; not a customer photograph or product screenshot.

A lower fuel total could mean better efficiency, fewer kilometres travelled, cheaper fuel or missing receipts. Compare the underlying volume, distance and price before interpreting the total. Keep currencies separate unless you have an explicit exchange-rate policy and record the rate and date used for conversion.

For downtime, define when the clock starts and stops. Is the measure time awaiting any maintenance attention, time unavailable for operation, or hands-on repair time? A vehicle can wait two days for a part and then require one hour of repair. Combining those periods without explaining them makes it harder to identify the actual constraint.

Use the same vehicle population and observation window in comparisons. If a problem vehicle leaves the fleet midway through a pilot, a lower fleet average may partly reflect the change in population. Keep that limitation visible. Similarly, better recordkeeping can initially increase recorded defects because issues that were previously informal now appear in the system.

The first measures should establish record quality: current readings, identifiable assets, complete closeouts and overdue work with owners. Once those are stable, longer-term measures can help evaluate maintenance performance. Do not use an illustrative calculation or a short setup test as proof of customer savings.

Run a structured 30-day evaluation

AI-generated editorial illustration; not a customer photograph or product screenshot.

Days 1–5: define the pilot and reconcile identity

Choose a representative subset of the fleet, a technician, a planner and the person who will judge the outcome. Write the three decisions the pilot should improve. Capture the current process and its recurring failures. Clean the selected asset records, establish the meter units and check the initial import against the source.

Avoid migrating the entire fleet before the team has completed a real workflow. A small pilot can expose a mistaken field mapping or unclear role without multiplying the cleanup across hundreds of records. Keep the original source recoverable throughout the evaluation.

Days 6–15: execute ordinary and difficult work

Create approved maintenance plans for the pilot assets and review them before activation. Record ordinary service work, an inspection finding and a job with a delay. If you connect Samsara, compare at least one imported reading with the source vehicle record and test how the team notices a stale or failed sync.

Review records daily with the people entering them. Look for instructions that require explanation, repeated copying and fields that are routinely left incomplete. Fix the process while the sample is small. A pilot is an opportunity to find friction, not an exercise in producing a perfect-looking dashboard.

Days 16–25: test review and reconciliation

Have a manager reconstruct a completed job without asking the original technician. Reconcile one fuel interval and one parts receipt against the supporting evidence. Review a recurring service that became due during the pilot and establish whether it reached the right person with enough information to act.

Test the handover between users. If only the person who configured the workspace can interpret the records, the process is still dependent on that person's memory. Record any training, missing feature or integration issue needed before a wider rollout.

Days 26–30: make a purchase decision from evidence

Compare the pilot against the acceptance criteria written at the start. Include the ongoing administrative work, required subscription tier and any external tools that remain necessary. Decide which gaps are acceptable, which have a workable manual process and which prevent adoption.

A useful outcome can be “the team can now find and reconcile every pilot service record” even when there has not been enough time to assess breakdown rates. Buy on demonstrated fit and a credible rollout plan. Do not require the pilot to manufacture a dramatic financial return to justify an otherwise practical operational improvement.

Questions to settle before subscribing

Can this replace a live fleet dispatch system?

PreventiveHQ's current fleet workflow is centered on maintenance records, fuel entry and supported meter imports. It does not replace live dispatch mapping or route optimization. If those functions are required, retain the relevant system or evaluate a specialist product with a demonstrated connection to the maintenance process.

Do drivers need to become maintenance planners?

They should not need to decide the technical scope merely to report an observation. Decide which intake method and account permissions fit your operation. The maintenance team still owns triage, work authorization and the procedure used to address the report.

Does connecting Samsara import the whole provider account?

No. The current connector handles selected vehicle usage statistics mapped to maintenance assets. It does not imply support for every Samsara API, diagnostic workflow or inspection record. Review the connection guide and test the exact statistic you plan to use.

Can old service records be retained?

Completed work can be imported through the dedicated validated history CSV. Prepare asset identity first and preserve original source identifiers and dates. Attachments, old user identities and detailed historical stock or labor transactions require a separate migration plan; they are not implied by a completed-work import.

How should we compare the commercial cost?

Use your actual users, assets and sites against the current pricing and limits. Include setup time, training and any external systems you still need. Compare a common required workflow across vendors before comparing the headline subscription amount. A lower price for a product missing a mandatory integration is not the same offer.

Start with one trustworthy service record

The next step is to choose a pilot vehicle and an approved maintenance task. Download the fleet worksheet, prepare its identity and usage data, and start a trial workspace. Use the acceptance checks in this guide to decide whether the workflow fits your team. The aim is a service record that another person can confidently use for the next decision.