Fixed. I just stopped using Graph and switched to a different API. Still, the issue on Opendoor’s website has persisted for three days. I don’t understand why they stopped showing homes on the site.

Idea: @Opendoor superpower is instant liquidity for sellers. So why make people choose between two confusing products to get it? Today you pick between Simple Cash and Cash Now, More Later, and the fees swing so wildly between them ($99k vs $21k) that it reads like gouging. It isn't: it's the same costs in different buckets, plus the cost of fronting you cash, plus a premium for who carries the price risk. Cash Now, More Later already has the right interface: a single liquidity slider. The fix is to fold Simple Cash into that same bar. One bar = how much cash you want now. • More liquidity now → less upside later. Make that obvious, it's already true: every extra dollar upfront raises the fee and lowers your ceiling, dollar for dollar. • The top ~4% of the bar = Simple Cash: 100% guaranteed, zero contingency. Give it its own visual, because it's a real endpoint, the guaranteed amount jumps up and the operation changes (Opendoor buys vs. facilitates). Then two honest lines at the bottom: • Cost of advancing your cash now, the part of the fee that actually scales with the slider (today it's buried in one "service fee") • Guaranteed vs. contingent on the sale, a disclosure of what's certain, NOT a fee. Never show the risk as a deduction; that only scares people. One control, one mental model, and the "scary 12% fee" explains itself. You see exactly what you're trading: certainty for upside. Just an idea, I think there’s room to think more deeply about this. From what I’ve seen, many users reacting to Opendoor’s ‘high’ fees are mostly reacting to confusion, not the actual economics. @morganb

Cash now, more later is really strong. They pay you $100,000 or $250,000 upfront so you can cover the down payment on the next home you’re buying while your current one sells, and the reduction in the second payment is actually very small compared to the advance. A smart seller is going to aim for this middle ground: a large guaranteed upfront payment and the rest in the second payout.

This gets even more interesting when someone who wants to move has already found a home they absolutely want to buy, they love it and don’t want to miss the opportunity, but they don’t yet have the cash to move forward. Opendoor gives you that liquidity so you can buy it right away. Maybe the bank wouldn’t have approved the loan because you didn’t have enough cash on hand, but now they will.

It looks like yesterday Opendoor cleared a bunch of homes from the website’s map, leaving only around 250 active listings, maybe because they’re changing something in the backend. For now, the delisting module will be disabled until things stabilize again, everything else keeps working as usual.

IDK what you all think, but right now the Opendoor desktop listings page is basically 50% map and 50% list. Personally, a small CSS tweak to something like 68% map and 32% list feels better to me. More space for the map nodes, less visual compression, and if the user is interested in a home they can still see all the details on the right. This is a quick test I did by tweaking the CSS from the console.

Everyone talks about how uncertainty hurts the home seller. Nobody talks about how it destroys the agent. Based on $400K avg transaction homes: Traditional agent reality: → 6-8 deals worked per year → 3-4 actually close (50% fall-through) → $12K commission × 3-4 closes = $36K-$48K → Months of unpaid work on every deal that collapses Opendoor-powered agent reality: → 40 deals per year → Every single one closes. Guaranteed. → $8K × 40 deals = $320K → 14 days per deal instead of 5 months The agent working 6 deals and closing 3 isn't lazy. They're trapped in a model where half their work generates zero income. That's the real story of traditional real estate. And it's the problem Opendoor actually solves. The next brilliant move? Agent-focused brokerage firms built entirely around Opendoor as their liquidity engine. Not individual agents using it occasionally. Full companies whose entire model is volume guaranteed closes. Same synergy as a financial broker using a market maker. The agent brings the client. Opendoor provides the liquidity. Everyone wins. This distribution layer doesn't exist yet at scale (realscout maybe the first). Whoever builds it first wins the decade. $OPEN

New chart added to the Opendoor Data Explorer tracker: Cohort Price Retention Curve Each line = one month of new listings. As Opendoor sells through each cohort, you can track: 📊 % of that month's inventory already sold (X axis) 📉 How much the avg listing price has dropped to close those sales (Y axis) Apr '26 stands out, steeper price drop vs Mar '26 at the same % sold, suggesting more aggressive repricing was needed to move that inventory. As more months are added, we'll build a clearer historical picture, making it easier to spot stronger vs weaker cohorts and track how Opendoor's repricing behavior evolves over time. ⚠️ This tracks listing price retention only — not contribution margin. A 3% drop from listing ≠ a loss.

Less than a year ago, Opendoor didn’t have homes around Chicago, Detroit, etc. Their inventory keeps expanding and covering more areas across the US. It’s only a matter of time before they gain experience and operational depth in these new markets. Still, it fascinates me how fast everything is changing in just a few months.

If you are an Opendoor shareholder, I have an ask. Proxy advisors at ISS and Glass Lewis have recommended shareholders to vote against me at our Annual Meeting. I don’t take this personally. This is the fifth time in my career these same people have told people to vote against my team. These proxy advisors have built no companies and are not meaningful shareholders of OPEN. They're a checkbox industry charging fees to tell other people what to do with shares that aren't theirs. Usually most companies can’t do anything about this since many institutional shareholders will just vote the way ISS tells them to. But Opendoor has the Open Army! It is important that we stand up against this separation of management from shareholders. If you are so inclined, help tilt the world in favor of shareholders and away from bureaucrats. Find out how (ask your broker, check your emails) and vote your shares. Our board is excellent. We are back on mission and we are winning. Don't outsource your vote. Read the proxy. Vote your shares.

Today we're launching a new Opendoor app on iOS You can browse homes, see prices and home details, book a tour in a few taps, and manage everything in one place Try it out and let me know what else you'd like to see us build into the app

Operational Q2 ends very close to the $800M tracked. A percentage of homes are sold privately and/or separately through local agents and therefore do not appear in Opendoor’s public listings. The adjusted estimated revenue for Q2 is currently around $920M . This is the first fully tracked quarter since the beginning, and the final revenue is very likely to land between $850–950M, exactly in line with their latest guidance (around $900M). As escrow times decrease with the introduction of Opendoor’s mortgage system, the operational calendar should compress, reducing the number of days of lag versus the original quarter. For now, we’re already starting operational Q3.

279 new listings in the last 5 days. Last week, during the slowdown caused by the platform transition, there were only 114 listings for the entire week. They’ve already recovered quite a bit, and we still have the weekend ahead (when they usually don’t add many new homes)

there wil be signs.. something big is cooking 🚀

Wrapping up my first week at Opendoor. When I left Shopify a few weeks ago, I had planned to take the summer off with the family. The plans changed drastically when I got the call from @nejatian. The team is too good and the opportunity for impact unparalleled. LFG

Only about 13% of the homes listed on Opendoor right now have been on the market for more than 90 days. The rotation capacity is incredibly high in a weak macro environment.