The Nine-Month Rule: When to Start Looking for Space
The ToHa towers from above, September 2023
There is a pattern in how Israeli technology companies approach office space, and it is very common. The instinct is to stay light, stay flexible, and decide late.
For headcount, hiring plans and most operating commitments, that instinct is correct. Applied to real estate it produces a predictable and expensive outcome, because the office market does not reward late decisions. It punishes them.
The inventory is committed before it exists
Take ToHa2, the Amot tower due to be completed at the end of 2026. Google signed for about 60,000 m² there in June 2024, for occupancy from January 2027, and by August 2026 Amot reported signed contracts or advanced negotiations covering 75% of the marketable space.
Read those numbers carefully, because it describes something counterintuitive. The buildings are not finished. The floors do not exist yet in any usable sense. And they are already spoken for.
So a company that decides to begin looking when it needs space is not entering a market and surveying the options. It is arriving after the fact, to examine what nobody else wanted.
Residual inventory does not price as a discount. It prices as the only option, which is a different thing entirely and considerably more expensive.
Large floor plates are not a liquid market
Enterprise requirements in Tel Aviv's core corridors are often measured in whole floors, and in the newest towers a single floor is large: ToHa2's floors were expected to be 2,500 to 3,000 square metres each, according to Globes.
The number of buildings that can deliver contiguous space at that scale in Yitzhak Sadeh, Yigal Alon and the surrounding cluster is small. Small enough that anyone working the market can list them without notes. And they are full: occupancy on Yigal Alon and in the Hassan Arafa/Yitzhak Sadeh area stood at 99% in the first half of 2026, according to a Newmark Natam survey reported by Globes.
This matters because liquidity assumptions imported from other decisions do not apply. When one of those floors is taken, it is gone for the length of the lease, which in large prime deals often runs six to ten years (Wiz's 2026 lease at Landmark B is for six years, Google's ToHa2 lease for ten). There is no secondary market that meaningfully replaces it, and no amount of budget conjures another one.
Flexibility requires alternatives. Where alternatives are countable on one hand, deciding late is not flexibility. It is deciding with fewer options.
Working the timeline backwards
The useful way to think about this is to start from the date the space must be operational and work back.
A fit-out programme runs months, not weeks, and that is after design is settled and permits are in hand. Lease negotiation on a significant floor is rarely quick. Before that comes the search itself, and before that the internal work of deciding what is actually needed, which is usually the step that takes longest because it requires agreement rather than execution.
Assembled honestly, that chain lands somewhere around nine months for a company of any size, and longer if the requirement is unusual or the internal decision is contested.
Nine months before you need it is therefore not early. It is roughly on time. Anything later is compression, and compression is paid for in rent, in specification, or in both.
What starting early actually costs
Very little, which is the part companies consistently misjudge.
Starting early does not mean signing early. It means knowing the inventory, understanding the pricing, having relationships with the relevant landlords, and being in a position to move when the right floor appears.
A company that has done that work can act inside days when an option surfaces. A company that has not will spend its first weeks learning what the market contains, by which point the floor it wanted has been taken by somebody who already knew.
The cost of preparation is some management attention. The cost of arriving unprepared is a worse building at a higher rent for the better part of a decade.
The practical rule
If your headcount plan shows you outgrowing your current space within roughly eighteen months, the market work should start now.
Not the search. Not the negotiation. The work of knowing what exists, what is coming, and who controls it.
By the time the need is obvious, the good floors have been gone for a year.
