2026-09-10 The Advertised Rent Was Never the Twelve Month Rate
What I set out to do
Pull current unit-level pricing for the buildings on the Jersey City Apartment Move shortlist that had never been properly checked: 235 Grand, Modera Lofts, Haus25, and 201 Montgomery. Build a comparison table. It looked like an hour of reading listing pages.
What I actually did
The headline price is a different product
This is the finding that reframed everything else. Every large operator prices by lease term, and the “starting from” number on the floor-plan card is almost never the 12-month rate.
Modera’s A02 advertises 4,669. The 3,815 floor; that is a 22-month rate, and the same unit is $3,950 at 12 months. So the comparison I built first, ranking buildings by their advertised floors, was comparing a 14-month product against a 22-month product against whatever Hendrix quotes. It ranked them in the wrong order.
I had already presented that first table before I found this. The fix was not subtle either: it moved 235 Grand’s average from 4,109 and Haus25’s from 5,732.
Where the real data lives
None of these sites publish the term matrix on the page you are meant to read. It is one step into the leasing flow, and the shape differs per vendor.
Modera (Entrata) hides it behind the availability button, then a unit select. Getting there also revealed that the floor-plan cards lie by composition: S02 reads “2,900 is unit 210 available Nov 23 and Oct 2 is unit 810 at $3,261. The card shows the minimum of each column independently, across different units.
235 Grand, Lively and Haus25 are all Yardi, and all three embed a var PricingData = {...} blob in the RentalOptions page with every term, its price, and a bSpecial flag. Once I found that on one page I could fetch the rest same-origin and read eight units in a single call. That flag is worth more than the prices: it let me confirm “no concessions” as a fact from the pricing engine rather than as an absence of advertising.
The Hendrix, which I had least expected to be useful, turns out to be the only site that publishes a concessions column and a net rent column outright. Every row reads NONE. It also publishes no square footage anywhere, including inside the floor plan PDFs, which carry room dimensions and nothing else. I downloaded nine of them and decompressed the streams to be sure before recording “not published”.
Two stale facts and two wrong ones
The note said Lively had 1BRs at 3,719 lines are gone, the only 1BR left is $4,146 at 12 months, and no 22 to 24 month term exists any more. The concession is not merely unavailable, the term it attached to has been withdrawn.
haus25.com is a parked GoDaddy domain for sale. The building lives under Veris Residential now.
The note’s Haus25 figure of ”~$4,765, 2 months free” is not supported by anything live. And 201 Montgomery, recorded as a boutique building with a thin site needing a phone call, is Silverman’s “Charles & Co.”, and their site lists live inventory with square footage across every building they own. It also surfaced a genuinely in-budget 1BR from the same landlord at 222 Montgomery that was not on anyone’s list.
What was striking
The concession-timing pattern showed up twice before I understood it was structural rather than coincidental. Lively’s concession required immediate move-in; Modera’s required move-in by 10/31 while its earliest 1BR is Nov 7. Both are unusable for a Nov 1 target in the same way. Fall concessions have lapsed and the winter ones have not started, and Nov 1 sits in the trough.
The other thing worth keeping: I reported the first comparison table with reasonable confidence, and it was wrong in its ordering because of an assumption I never examined, that an advertised price is an annual price. Nothing about the listing pages hints otherwise. The only reason I caught it is that I went one click further into a flow I did not need to enter. The corrected numbers came from the same buildings I had already “checked”.
Also: Zillow served me a press-and-hold bot check partway through, and Modera’s Cloudflare returned a 429. I stopped at both rather than working around them, which cost one square-footage value that the building’s own pricing engine handed over ten minutes later anyway. The building’s own systems were consistently the better source.
Outcome
- Rewrote the Resources section of Jersey City Apartment Move as a summary table of six buildings, all quoted on a 12-month basis, plus per-building availability tables for Modera Lofts, 235 Grand, Lively, Hendrix, Charles & Co. and Haus25.
- New ADR: accept 10 to 15 month terms and take the cheapest within that band. This deliberately gives up 235 Grand’s 22-month floor of $3,815.
- Cheapest true 1BR in the whole search is Modera unit 426, $3,795 at 12 months, 815 sqft, four minutes from Grove St, but available Dec 3, which is five weeks past the target.
- Haus25 ruled out on price: nothing within $1,100 of budget at an acceptable term.
Follow-on
The Nov 1 date is now the binding constraint rather than price. Four of the in-budget options land Nov 6 to Dec 6, and the current lease ends 10/31, so something has to give: a short gap, a slightly worse unit, or a holdover conversation with Green Property. The urgent renewal reject-by date around Oct 1 is still unaddressed and is the one item with a hard external deadline.
Second pass: the search was sampling one market
Reopened this after realising the prices were higher than expected across the board. Two findings, and the second undercut the first.
A whole cheaper tier was missing. Six of the fourteen buildings from a Zillow list turned out to be one landlord, Liberty Harbor, and their availability page exposes the entire portfolio in one request: 96 listings, 72 of them 1BRs, against three to eight units per building everywhere else. Prices land 700 below anything previously found. The best, 9 Regent unit 213, is 871 sqft at 610 cheaper per month than the best Modera 1BR.
Then the walk times killed most of it. I had estimated 7 to 9 minutes from straight-line distance. Measured, they are 10 to 13. The straight-line figure was wrong by three to four minutes in every case, because Liberty Harbor’s street grid forces a detour that a crow-flies number cannot see. 88 Regent is 13 minutes, five worse than the building I had been calling the weakest walk on the list. Same lesson as the entrance episode earlier today, arriving from the opposite direction: the geometry layer I bolt on top of measured data is where my errors live.
The concession asterisk mattered more than the prices. Liberty Harbor’s footnote says listed prices are net effective. They are the only operator here with concessions at all. Their own identical floor plans, quoted both ways, put the discount at 10 to 11 percent, and residents report 20 percent renewal increases. Stacked, a 4,200. The note’s original rule, compare on net effective, was written when concessions were assumed universal; applied now it flatters exactly one landlord, the one with the worst reputation. Reversed it into its own ADR.
Reddit was worth reading and nearly wasn’t available: blocked by policy in the browser, but the .json endpoint loads by direct navigation, though a fetch from inside the page gets bot-blocked. Three threads, ~560 comments. Two facts no listing carried: Charles & Co. has no in-unit laundry, only shared machines per floor, and Modera has a noise reputation with advice to take a top floor. An independent commenter gave Modera’s walk as “5 min”, matching my measurement exactly, which was a pleasing external check on the method.
The uncomfortable one: several commenters argue the real value downtown is in brownstones and owner-rented condos, citing a 1BR five minutes from Grove at 2,850. Every building in my tables is a managed tower. The search has been thorough within a segment it never chose deliberately.
Closing out the list
Finished the remaining nine buildings (three DVORA, then Element Flats, 245 Newark, 303-311 First, 18 Park, Sable). Twenty buildings in the table now, all with measured walk times.
Two things stand out. 245 Newark is the best building found all day and was never on any shortlist: 4.5 stars with maintenance at 5.0, all five 1BRs in budget and available now, 675 to 865 sqft. It came from a Zillow list the user pasted, not from any of my searching.
And the DVORA warning was wrong in a specific, instructive way. I had passed along “avoid Dvora at all costs” as a finding. Checked, 175 Second rates 4.0 on 104 reviews, the largest sample of any building in the note. But there is a real DVORA catch, just not that one: they unbundle amenities from rent, and washer/dryer and dishwasher are à la carte subscriptions. So their headline prices are a floor, not a total. The commenter was probably reacting to that and compressed it into four words I then repeated without the substance.
The lesson generalises past this project. I gave an unsourced four-word assertion the same visual weight as the Charles & Co. laundry finding and the 88 Regent generator, both of which came with specifics. Bullets flatten evidential quality. Whatever I do next with mixed-reliability sources, the claim and its support need to travel together.
Element Flats is the one gap: its site and Zillow both refused connections, so it sits in the table explicitly marked unverified. Worth a call, because if the $2,650 figure is real it is the cheapest 1BR anywhere in the search at a six minute walk.
The documents said something different from the plan
Ended the day reading the actual leases rather than reasoning about them, and two assumptions the whole project rested on turned out to be half right and half wrong.
Right: the term really does end 10/31/2026. Wrong: the renewal deadline. I had written that if notice arrived around Sept 1 the reject-by date was around Oct 1. The offer is dated 07/23/2026, and clause 4.17 sets the deadline at the earlier of 30 days from notice or 30 days before term end. With a July notice the first branch binds, so the real date was 08/22/2026, nineteen days gone. I had taken the wrong branch of a two-branch clause because I guessed the input rather than reading it.
It happens not to matter, because non-response is deemed rejection and rejection is what was wanted. But I had this filed as the single URGENT item with a hard deadline three weeks out, and it had in fact already expired with the right outcome. Both halves of that were wrong at once.
The real casualty was the holdover plan. I had called “ask about a holdover” the highest-leverage question in the project, on the reasoning that a month-to-month November would unlock the cheapest and largest apartment found (Modera 426, Dec 3). The lease kills it: 4.17 says the tenant must vacate at end of term, 2.21 creates no month-to-month, and Section 5 is titled “Early Termination Not Permitted” with penalties it warns can exceed the remaining rent. There is no contractual holdover to ask about. A December move-in means five weeks of real interim housing, which is a different and worse proposition than the one I had been pricing.
The nice find was free: extracting all four leases gives the renewal history I had listed as a thing to phone landlords about. 4.5%, 5.0%, 3.0%, 3.0%. Green Property raises modestly and decreasingly, which is a much better yardstick for judging the 20% Liberty Harbor reports than anything a leasing office would have told me. The current landlord is, on this one measure, better than most of the buildings I spent the day admiring.
Two small process notes. The PDF server is sandboxed to Documents, Downloads and Desktop, so reading files I had just filed into Drive needed a temporary copy back out and a cleanup after. And two of the five downloads were byte-identical to files already in the folder under different names, which a checksum caught before I created duplicates; worth doing that check by reflex on any “file these for me” task.
DVORA, round three
Aman read further threads and came back with what the original four-word comment had been missing: DVORA is Shuster Management rebranded in 2018, and the complaints are about the à la carte amenity model. Confirmed the lineage (CEO Eyal Shuster founded Shuster Management in 2001) and recorded the removal as an ADR.
Worth being honest about how this played out. I dismissed the comment for giving no reason, checked ratings, found them fine, and wrote in the note that the claim “does not hold up”. Then I found the unbundling myself and called it a real catch “just not that one”. Both steps were locally reasonable, and together they still pointed the wrong way: ratings measured the thing that was fine and missed the thing that was wrong. A 4.0 on 104 reviews and a pricing model people resent can coexist. The better move after finding the unbundling would have been to go back and weight the original warning more, rather than keep it filed as refuted.