I landed in Tel Aviv on February 4, 2026. I signed a lease on March 18. In between I spent about ten hours scrolling Yad2, messaged a dozen Facebook listings, and viewed exactly one apartment: the one I now live in.
The gap between those two facts is the point of this article. For six weeks I made no progress at all. Then I changed one thing, and three days later I had a place. What changed was not effort, or budget, or luck. It was timing.
Why the first six weeks produced nothing
I booked an Airbnb in Shapira for a month. A private room in an apartment shared with four other people, $400 a week.
At the end of that month I renewed for a second. Not because I liked it, but because I had not started looking. I was doing aliyah admin, seeing family I had not seen in years, taking trips north to Haifa, and trying to work out where in the country I actually wanted to live. Every one of those was a reasonable use of a day. Together they ate six weeks.
This is the part I would want someone landing tomorrow to hear. Nobody drifts because they are lazy. You drift because the apartment search is the only task with no deadline attached, and everything else has one. The Airbnb quietly becomes the plan.
By the second month I wanted out for two reasons: I wanted my own space, and I was paying more per week for a room than I would eventually pay for an apartment.
What went wrong on Facebook Marketplace
I started there because it is in English and the barrier to sending a message is nearly zero.
I messaged about twelve listings. Two people replied. Both conversations died before I saw anything. No explanation, no "it's taken", just silence.
Ten of twelve never answered at all. I do not know why, and that is the problem with Marketplace: you get no signal. A listing that was rented three weeks ago looks identical to one posted this morning.
Why the Yad2 listings were already gone
Yad2 is where the actual inventory is. I gave it about ninety minutes a day for seven days, so roughly ten hours in total.
In those ten hours I contacted zero landlords.
Not because I was hesitant. Because Yad2 shows you two things that make the decision for you: how many views a listing has, and when it was last updated. The listings I was finding had between 300 and 1,500 views.
A listing with 1,500 views is not an opportunity. It is a record of something that already happened. Somewhere in those 1,500 people were the ten who called on day one, and one of them has already signed. By the time a listing surfaces in a manual browse, you are not early, you are archaeology.
So I would open a listing, look at the view count, and close it. Ten hours of that. The search was not failing because the apartments were bad or my budget was wrong. It was failing because I was arriving late to every single one.
What actually changed
I am an engineer, and I have scraped platforms before in previous work. I used an AI tool to help me write it. It took a couple of hours.
The bot did one thing: check Yad2 every fifteen minutes for new rental listings in Tel Aviv and the surrounding cities, Ramat Gan and Givatayim included, filtered to my budget and size, and post anything new to a Slack channel.
It did not find better apartments. It found the same apartments, earlier.
At that point I still did not know which neighborhood I wanted. My requirements were vague on purpose: close to a big park, and no more than 4,000 shekels a month. Widening the search to the greater Tel Aviv area rather than Tel Aviv proper turned out to matter more than any other filter.
What the alert looked like
Three days after I switched it on, this arrived:
The timestamp reads 5:31 AM because I took this screenshot in June while traveling in New York, and Slack renders old messages in whatever timezone you are currently in. In Israel it was 11:31 in the morning on March 18. The building number is removed.
Ramat Gan, a short walk from Park HaYarkon, 49 square meters, two rooms, exactly at my 4,000 shekels budget. Three other listings came through in the same batch.
I called the owner straight away. He confirmed later that I was the first person to reach him. I saw the apartment that afternoon and told him I would take it before I left.
One viewing. That was the whole search.
To be precise about the timing claim: my scraper could not read Yad2's publish date reliably, which is why the alert says "lo nimtza" in that field. The fifteen minutes is my polling interval, not something the listing told me. What I can say is that the listing was at most fifteen minutes old when it reached Slack, and that the owner told me nobody had called before me.
Did I need Hebrew?
Less than I expected, and this surprised me.
The listings are in Hebrew. I read them with Chrome's built-in translation, which was accurate enough for the things that matter: price, rooms, size, floor, whether there is an elevator.
The owner was Israeli and spoke reasonable English on the phone. Every message after that first call was WhatsApp, in English.
I want to be careful here, because "you don't need Hebrew" is the kind of claim that gets people into trouble. Reading listings without Hebrew is a solved problem. Whether the person on the other end speaks English is luck, and it will not always go your way. But the language barrier was not what cost me six weeks. Timing was.
What I would tell someone landing tomorrow
Start the search in week one, even badly. You will not know where you want to live. Start anyway. Vague criteria plus early contact beats perfect criteria plus late contact, every time.
Treat the view count as the only number that matters. Under a hundred views, call now. Over five hundred, it is probably already gone. You will save yourself hours by filtering on this before you read the description.
Widen the map. I found my place in Ramat Gan because I stopped restricting the search to Tel Aviv. The apartment I could not find inside the city existed twenty minutes away at the price I had set.
Understand what the short-term rental is costing you. Mine cost more per week than the lease I eventually signed. Six weeks of that is real money, and none of it was a decision I consciously made.
And know that the fundamental problem is not one you can outwork. I did not find my apartment by trying harder. I found it by seeing a listing fifteen minutes after it was published instead of three days after. Ten hours of manual browsing produced nothing. A bot running while I did other things produced a lease in three days.
That is the entire difference, and it is why I ended up building Sia.