Disclosure: the author holds shares in Innoviz (INVZ) and in Mobileye (MBLY), which is mentioned in this piece as a customer of Innoviz, as of the date of publication. Those holdings may change at any time without notice. Nothing here is a recommendation to take any action in any security. See the full disclosure and disclaimer at the foot of the page.
Innoviz published its second-quarter report on 5 August 2026. We are writing about it today, 6 August.
What was reported
| Quarter | Year ago | Change | |
|---|---|---|---|
| Revenue | 18.1 | 9.7 | +87% |
In millions of dollars
That is the highest revenue level in the company's history. And per the report, it comes from a combination of non-recurring engineering services - NRE - and actual LiDAR unit sales.
And that composition matters. NRE is what a carmaker pays for bespoke development, and it arrives before production begins. Unit sales are the recurring revenue that follows. The larger the unit share becomes, the more mature the model.
The guidance that did not move
Full-year 2026 guidance was left unchanged: $67 to $73 million.
And that is a point worth pausing on. After a first half containing a record quarter, the company chose not to raise guidance. That may reflect caution, and it may reflect that the second half depends on delivery dates outside its control.
And the news that changes the picture
The first defense order
On 30 July - after quarter end - Innoviz received its first defense orders.
- Initial size: $3.5 million
- From a major defense customer
- Several hundred LiDAR units
- With the potential for additional orders
$3.5 million is small against guidance of $67 to $73 million. But the significance is not in the size - it is in the opening of a channel.
Until now Innoviz has been an automotive company. A model in which every dollar of revenue depends on carmaker timetables, which are known for slipping. A defense order operates at an entirely different pace - shorter, and usually with less dependence on the start of mass production.
And the company notes it is in discussions with dozens of prospective customers in the field.
The automotive programs
The company reports ten global OEMs, and programs on track to start of production (SOP) with three names:
- Volkswagen
- Mobileye
- Daimler Truck
And Mobileye is the interesting detail here. It is not a carmaker but a supplier of vision and autonomous driving systems - meaning Innoviz sells to it as a component inside a larger system.
That is a different distribution channel from selling directly to a manufacturer: it may be broader, because Mobileye itself sells to dozens of carmakers - but it also places Innoviz further from the end customer.
What deserves watching
Three points
First, SOP is a date that slips. "On track to start of production" is accurate phrasing but not binding. In the automotive industry timetables move - and every delay pushes revenue with it.
Second, guidance was not raised despite a record quarter. That is a choice, and it says something about confidence in the second half.
And third, the defense order is a beginning, not a trend. $3.5 million from one customer, with "potential" for more. That potential will be measured in coming quarters, not today.
My Angle
A personal opinion of Ilan Abramov - not advice, not a recommendation
I hold Innoviz, and so I try to be more exacting with it rather than less.
What is good in this report is real: revenue that nearly doubled to a record $18.1 million, and a mix that is starting to include actual unit sales rather than only development services. A LiDAR company selling hardware in growing volume is a company leaving the proof-of-concept stage.
And what I flag as a question: the guidance that did not move. After a record quarter, a company leaving full-year guidance at $67 to $73 million is saying - deliberately or not - that the second half will not be strong enough to justify an update.
And what interests me more than the numbers is the defense order. $3.5 million is nothing at this company's scale. But the defense market runs at a different pace from the automotive market - orders arrive faster, and do not depend on a mass-production start that slips from year to year. If that channel genuinely opens, it changes the company's risk profile more than any single automotive program.
What I will check next quarter: whether a second defense order arrived. One is an event. Two is a channel.
