Skip to main content
Main analysis: Tondo Smart 2025: Defense Is Lifting Margins, but 2026 Is Still a Proof Year
ByMarch 26, 2026~8 min read

Tondo Smart: How Much of the Backlog Really Turns Into Revenue and Cash by End-2026

Tondo's backlog looks larger than the near-term revenue it actually supports. This follow-up separates the headline figure from the 2026 recognition bucket and the much longer-dated performance obligations, then tests what that means for end-2026 cash visibility.

What Sits Behind the Backlog Number

The main article framed 2025 as Tondo's proof year. This follow-up isolates the part of the 2026 story that is easiest to overread: the backlog. The headline number looks strong, but it combines three very different things: revenue expected within 2026, multi-year service layers that stretch far beyond it, and future revenue recognition tied to orders that have already been supplied.

That is the core point. At 31 December 2025 backlog stood at NIS 23.2 million, but only NIS 6.0 million of that sat in the 2026 recognition bucket. Near the publication date, on 10 March 2026, backlog had risen to NIS 25.4 million, yet only NIS 7.6 million was still allocated to 2026. The remaining NIS 17.8 million was pushed into 2027 and beyond.

That matters because the headline figure invites a simple reading: there is backlog, so 2026 must be largely spoken for. The numbers themselves argue for a more careful interpretation. Even after the February and March updates, less than one-third of backlog falls into 2026, and the company explicitly says the expected-recognition schedule includes revenue that will be recognized from orders already supplied. So even the 2026 bucket is not pure future execution.

The headline backlog remains long-dated even after the March 2026 update
Metric31 December 202510 March 2026What it means
Total backlogNIS 23.2 millionNIS 25.4 millionThe headline grows, but the shape stays long-dated
Expected recognition in 2026NIS 6.0 millionNIS 7.6 millionThis is the truly near-term layer
Expected recognition from 2027 onwardNIS 17.2 millionNIS 17.8 millionMost of backlog still sits outside 2026
2026 share of backlog25.9%30.0%Even after the update, still less than one-third

Remaining Performance Obligations Show an Even Heavier Long Tail

There is another number that has to be read alongside backlog: remaining performance obligations. Here the company excludes transactions whose original expected duration is up to one year and leaves only the longer-dated layer. At 31 December 2025 that amount stood at NIS 14.7 million. Of that, NIS 3.8 million is expected in 2026, NIS 1.7 million in 2027, and NIS 9.2 million only in 2028 and later.

That sharply changes the read. The headline backlog says there is business in hand. The remaining-performance-obligation table says a meaningful part of that business stretches much further out than a first glance suggests. In fact, 62.3% of the disclosed remaining performance obligations sit only in 2028 and beyond. That is not a near-term execution bucket. It is a long contractual tail.

Remaining performance obligations are concentrated beyond 2027

The gap between NIS 23.2 million of backlog and NIS 14.7 million of remaining performance obligations is not an accounting contradiction. It tells you backlog is a blended metric. On one side, the company says the remaining-performance-obligation table excludes contracts whose original duration is up to one year. On the other, the backlog footnote says the recognition schedule also includes revenue from orders already supplied. These are two different cuts through the same activity, not two versions of the same measure.

LayerAmountWhat it captures
Backlog at 31.12.2025NIS 23.2 millionAll orders expected to generate revenue, including a layer already supplied
Remaining performance obligationsNIS 14.7 millionOnly transactions whose original duration exceeds one year
Gap between the two measuresNIS 8.5 millionActivity that does not enter the long-duration disclosure

So anyone taking the NIS 23.2 million headline and building a fast end-2026 revenue case from it is missing the structure. Backlog says there is scale. Remaining performance obligations say a large part of that scale is materially longer-dated.

The Two Main Civil Projects Explain Why Conversion Is Slower Than the Headline

Once the analysis moves from the aggregate tables into the two main civil projects, the same pattern repeats. There is real activity and real backlog, but the economic value is not sitting neatly inside the next twelve months.

Netivei Israel

For Netivei Israel, the initial expected project scope was about NIS 50 million, with room for additional expansion orders. By the end of 2025 the company had already recognized cumulative revenue of roughly NIS 42 million from the project, while carrying a contract liability of NIS 8.2 million, of which only NIS 1.1 million was current. The combination of recognized revenue and that contract-liability layer already comes close to the initial project envelope.

That means the older flagship civil project is not hiding a major unreported 2026 conversion surprise. It remains important, but much of its economic weight has already passed through revenue or sits in a contract layer that stretches beyond the coming year.

Jerusalem

Jerusalem looks similar, only smaller in scale. By the end of 2025 the company had recognized cumulative revenue of NIS 4.0 million from the project and also carried a contract liability of about NIS 1.8 million, with only about NIS 0.2 million classified as current. Then, in February 2026, the company updated cumulative project orders to about NIS 5.7 million, explicitly noting that this amount includes ongoing management services for 10 years.

That is a useful clue. NIS 5.7 million sounds like a fresh step-up, but it broadly reconciles with the NIS 4.0 million already recognized by year-end 2025 plus the roughly NIS 1.8 million carried as contract liability. The new Jerusalem pilot matters strategically because it expands the product and service set on top of infrastructure already installed. At this stage, though, it does not fundamentally change the 2026 recognition path. It mainly reinforces the project's multi-year tail.

Most of the disclosed project contract-liability layer sits beyond the next year

That chart is the heart of the issue. In both civil projects, most of the disclosed contract layer at the end of 2025 is non-current. So when the company talks about backlog, what sits underneath increasingly looks like service and management revenue recognized over time, not just fast product delivery packed into 2026.

Even the Post-Balance-Sheet Additions Extend the Story

After the balance-sheet date, two announcements can easily invite a more aggressive 2026 reading: the Shikun & Binui Netivei Hatzafon project and the Jerusalem pilot. Both are positive. Both also reinforce the same structural message.

In Netivei Hatzafon, the company reported in February 2026 an order of about NIS 2.5 million. Implementation is expected during the first half of 2026, but after that the company is expected to provide ongoing technology-management services until mid-2030. So even when Tondo wins a new order, the economic structure still mixes near-term deployment with a multi-year service layer.

Jerusalem makes the same point even more clearly. The February 2026 update said the cumulative project amount of about NIS 5.7 million includes not just edge units but also ongoing management services for 10 years. That is good news for customer stickiness and for the company's ability to widen its footprint on infrastructure already in place. It is not automatically good news for a sharp acceleration in recognized revenue by end-2026.

What matters here is that nearly every positive civil update adds contractual depth, not just speed. That can be strategically attractive. It can also improve the quality of the customer relationship. But it means backlog is not a shortcut to a one-year revenue forecast.

So How Much Really Turns Into Revenue and Cash by End-2026

On revenue, the disclosure allows a fairly firm answer. Based on the information available near publication, the hardest number is NIS 7.6 million of expected 2026 recognition out of NIS 25.4 million of backlog. Even that figure is not a pure future-delivery layer, because the company says it includes revenue expected from orders already supplied. The conservative reading, then, is that Tondo should convert a meaningful part of backlog by end-2026, but not most of the headline backlog.

On cash, the answer is narrower, and it is important not to blur that limitation. The disclosure does not provide a project-level collection calendar, detailed billing terms, or a project cash bridge from backlog to cash. So the NIS 25.4 million backlog cannot be translated reliably into an end-2026 cash forecast. What can be said is that the disclosed contract structure, the long service tail, and the non-current classification across most of the project balances do not support a reading in which most of the backlog rapidly becomes both revenue and cash.

Bottom line: Tondo's backlog is real, but it is stronger as evidence of multi-year visibility than as evidence of fast 2026 conversion. It shows that the company is embedding itself in real projects. It does not show that most of that value will sit in recognized revenue by end-2026, and it certainly does not justify assuming the same pace for cash.

The right test for 2026 is not whether the company keeps printing backlog. It is whether most of the NIS 7.6 million marked for 2026 is recognized on time, and whether new wins start to bring a larger near-term supply layer rather than only another long service tail. That is where the line runs between backlog that looks good on paper and backlog that proves real conversion.

Disclosure: Deep TASE analyses are general informational, research, and commentary content only. They do not constitute investment advice, investment marketing, a recommendation, or an offer to buy, sell, or hold any security, and are not tailored to any reader's personal circumstances.

The author, site owner, or related parties may hold, buy, sell, or otherwise trade securities or financial instruments related to the companies discussed, before or after publication, without prior notice and without any obligation to update the analysis. Publication of an analysis should not be read as a statement that any position does or does not exist.

The analysis may contain errors, omissions, or information that changes after publication. Readers should review official filings and primary sources before making decisions.

Found an issue in this analysis?Editorial corrections and sharp feedback help keep the coverage honest.
Report a correction