Parkomat: Why the Service Layer Still Does Not Stabilize Earnings
Parkomat's service and maintenance revenue rose to NIS 5.98 million in 2025, but cost of sales still reached NIS 7.05 million, leaving the layer loss-making despite a clear improvement. Once 12 to 24 months of warranty, project-bundled service periods, and lingering post-handover legal threads are added, it becomes clear why recurring revenue still does not act as an earnings stabilizer.
What This Follow-up Is Isolating
The main article argued that 2025 looked better on cash flow than on normalized profitability. This follow-up isolates the layer that should have softened that tension: service, maintenance, and warranty. On first read, this is supposed to be recurring revenue. In practice, even after a sharp improvement in 2025, it still does not behave like a stabilizer.
The numbers make the point quickly. Service and maintenance revenue rose to NIS 5.983 million, up 42.7% from 2024. Cost of service still reached NIS 7.050 million, so the layer ended the year with a gross loss of NIS 1.067 million. That was better than the NIS 2.119 million gross loss in 2024, but it was still the third straight year of gross losses in this line. Just as important, service was still only 8.4% of group revenue. A line this small, and still loss-making, cannot stabilize the company. At best it reduces dependence on projects a little. For now, it does not even do that.
This is not a side issue. The 2025 service gross loss wiped out 11.8% of the group's total gross profit and was more than seven times the company's NIS 145 thousand pretax loss. That is why it is a mistake to treat service as an automatic defensive layer. In Parkomat's 2025 economics, it is still a drag on earnings rather than a buffer.
The Problem Starts in the Contract Structure
The contract terms explain why recurring revenue looks cleaner on paper than in the underlying economics. The company's standard sale agreement includes 12 to 24 months of warranty, and the customer can extend that warranty for an additional payment. At the same time, once a parking system is handed over, the company signs a service and maintenance agreement with the building representatives or the building management company. In other words, the service layer does not begin as a pure aftermarket contract. It begins as part of the promise to deliver a functioning system, and only later tries to turn into paid maintenance revenue.
The Ahuzot Hahof Bograshov project shows that structure almost without abstraction. The contract was worth about NIS 28 million, included a 24-month warranty period, and during that period service and maintenance were included in the total consideration. Only after the warranty period does the customer have the exclusive right to buy additional maintenance from the company for extra consideration. Put simply, Parkomat first has to service the installation inside the project price, and only later tries to monetize it as recurring revenue.
| Layer | What the contract says | Economic implication |
|---|---|---|
| Standard warranty | 12 to 24 months of warranty, with a paid extension option | Post-handover friction begins immediately, before a mature service contract exists |
| Post-handover service | Usually signed with resident representatives or a management company | The service layer shifts from the developer world into ongoing field maintenance |
| Bograshov | 24 months of service and maintenance are included in the project consideration, with paid maintenance only afterward | Some future upside is delayed, while the operating burden starts on day one |
That is the core issue. The line between project economics and service economics is not clean. The company itself describes a warranty period in which it fixes, at its own expense, defects caused by a product defect or defective work, and only after that period enters into a separate service agreement for service calls, troubleshooting, and ongoing maintenance. That means higher service revenue is not the same thing as a mature profitable aftermarket engine. For service to stabilize earnings, a much larger part of the installed base has to move from contractual obligation into paid maintenance. By the end of 2025, that shift was still not visible.
The 2025 Numbers Show Where It Breaks
Once 2025 service cost is decomposed, the issue is clearly not one odd overhead line. Labor and related costs alone were NIS 4.271 million, equal to 71.4% of service revenue. Add spare parts and subcontractors, and the total reaches NIS 5.806 million, or 97.0% of revenue. Before vehicles, travel, office costs, depreciation, and other expenses, almost the entire revenue line is already gone.
This is not a G&A problem. It is field economics. The lines that consume most of the revenue are technicians, parts, subcontractors, and travel. Those are exactly the areas an aftermarket platform has to price correctly if it is supposed to become a recurring profit stream. That is why the 2025 improvement still needs to be read carefully. The gross loss narrowed, but it did not disappear because the economic structure has not yet flipped.
The warranty provision also did not signal a system that was calming down. The provision rose to NIS 672 thousand at the end of 2025 from NIS 541 thousand at the end of 2024, an increase of 24.2%. This is not a large balance-sheet item, but the direction matters. If the layer that is supposed to stabilize the business were beginning to work in Parkomat's favor, there should be at least an early sign that the past burden is fading. End-2025 still did not provide that signal.
The Obligation Does Not End at Handover
The post-handover layer is also visible in the contingent-liabilities note. In August 2023, a NIS 700 thousand claim was filed in connection with a project where a parking system had been installed. In February 2025, a roughly NIS 6.3 million claim was filed over a robotic parking system installed about nine years earlier, and that case was later deleted after a court-approved compromise to appoint an agreed expert. In August 2025, another claim was filed for roughly NIS 6.191 million in total, with the company jointly and severally exposed for about NIS 3.431 million. In the cases where exposure remained, the company's legal advisers believed dismissal was more likely than acceptance, so no provision was recorded.
| Service and warranty thread | Accounting status at year-end 2025 | Why it matters |
|---|---|---|
| August 2023 claim | No provision | Shows that friction around delivered systems can remain open after handover |
| February 2025 claim on a system installed about 9 years earlier | Claim deleted after an agreed-expert compromise | The operating tail can remain alive for years |
| August 2025 claim | No provision, with Parkomat sued for about NIS 3.431 million jointly and severally | Even when the company disputes liability, the service layer still demands time, handling, and defense |
It is important to stay fair here. In some of these proceedings the company argues that the damage was caused by water penetration or building defects rather than by the parking mechanism itself, and its legal advisers did not see a need to recognize a provision. So this is not evidence of a systemic product failure. It is evidence of a narrower but still important point: the warranty, service, and maintenance layer is not a brief event that ends at delivery. It remains part of the economics of the project long after the initial revenue has already been recognized.
What Has to Change Before Service Really Stabilizes the Business
The path to turning service into a real stabilizer is fairly clear, and Parkomat is not there yet. First, a larger part of the installed base has to move from warranty periods and bundled service into paid maintenance agreements. As long as most of the work is still done inside warranty or inside the project consideration, service is more obligation than aftermarket.
Second, the cost structure has to move. If field labor, parts, and subcontractors absorb nearly the full revenue line, the company is not yet enjoying operating leverage from a larger installed base. For service to generate profit, Parkomat has to show either better pricing, lower failure intensity, or both.
Third, the warranty tail has to shorten. That does not mean service calls or legal claims disappear altogether, which would be unrealistic in this kind of business. It means the warranty provision starts to settle, legal threads around delivered systems become less prominent, and service stops consuming a double-digit share of the group's gross profit.
Until that happens, the better way to read service at Parkomat is not as a second engine that already stabilizes the business, but as part of the cost of turning a signed project into a working installation. It is a necessary layer. It is still not a cushioning layer.
Conclusion
The easy mistake in reading Parkomat is to assume that every increase in service revenue automatically moves the company toward a steadier profit model. The 2025 numbers show the opposite. Service grew, but remained loss-making. The contract structure delays part of the monetization until after warranty. And the friction around delivered systems still shows up both in the warranty provision and in legal threads.
Current thesis in one line: Parkomat's service layer still does not stabilize earnings because it begins as a warranty obligation and an expensive field-service model, and only later tries to become paid recurring revenue.
What changed versus the initial read: It is now clearer that the issue is not only that service is loss-making, but that the loss sits on top of a contract model in which part of the service is bundled into project consideration, so revenue growth does not automatically convert into margin.
Strongest counter-thesis: One could argue that this read is too conservative because the installed base is growing, and as more systems leave warranty and move into paid maintenance contracts, the same service network could start to show much better operating leverage and profitability.
What could change the market read over the short to medium term: Not another rise in service revenue on its own, but evidence that service is moving toward breakeven, that the warranty provision stops rising, and that the installed base is actually converting into paid maintenance agreements.
Why this matters: In a robotic parking business, recurring revenue deserves that label only when it absorbs the friction of handover and starts producing margin after installation, not when it merely extends project cost into later years.