Supply is fading. Your concession math isn’t.

Multifamily deliveries hit their smallest quarterly volume in four years in Q1 2026 — just 75,205 units, down roughly 30% year-over-year. The pipeline behind that is thinner still: starts reached their lowest quarterly level since 2011, at 55,000 units, a 73% drop from the early-2022 peak. The supply wave is cresting. But nearly 1.3 million apartments remain in lease-up nationally, equal to 6.9% of total inventory — nearly double the pre-pandemic norm. That overhang means the recovery isn’t uniform, and operators who pull concessions off the national headline rather than their own submarket’s absorption data will feel it.

The operators who time this correctly stand to recover 200–400 basis points of effective rent growth before the market broadly agrees the cycle has turned. The ones who wait for consensus will give that margin away.

Also: The AI deployment gap is wider than you think, stack sprawl is costing more than one CTO’s patience, and this week’s move gives you a prompt to run your active concessions through a sanity check today.

The Radar

The 28% AI deployment gap: Awareness of AI in property operations is near-universal — 45% of property professionals say they understand AI’s potential, but only 28% have actually deployed it. That gap gets uglier upstream: JLL’s 2025 Global Real Estate Technology Survey of 1,500+ senior decision-makers found 88% of owners are running AI pilots, yet only 5% have hit most of their program goals. The root cause isn’t the tools — it’s that only 28% of firms have a formal AI training program in place, so pilots stall at the demo stage and never touch a live workflow. What it means: If you’re managing 200–1,000 units, the edge right now isn’t buying a smarter tool — it’s being one of the few operators who actually trains your team to run it.

2026 deliveries tracking down ~one-third: Market-rate apartment deliveries are projected to total approximately 323,000 units in 2026 — down roughly one-third from the 477,000-unit annual average of 2023–2025. It’s not just a forecast anymore: supply is clearly tapering, with Q1 2026 deliveries already falling roughly 30% year-over-year and, per RealPage, hitting the smallest quarterly volume in four years at just 75,205 units. The pipeline behind that is even thinner: Q1 2026 multifamily starts reached their lowest quarterly level since 2011, at just 55,000 units — a 73% drop from the early-2022 peak. The catch: markets like Austin and Phoenix are still sitting on years of overhang, and property owners may only begin pulling back on concessions as absorption stabilizes and lease-up conditions gradually improve. What it means: Shrinking supply is the headline, but if your portfolio is in a Sun Belt market where concessions are still running deep, don’t reprice off the national narrative — wait for your local absorption data to confirm the turn before pulling discounts off the table.

Stack sprawl is operators’ #2 priority to fix in 2026: A Multifamily Insiders flash poll of 45 operators ranked tech consolidation second only to AI as a 2026 priority — and the reason is pain, not strategy. RPM Living’s CTO Scott Pechersky is cutting from more than 40 platforms down to roughly 15, saying the payoff is “better training, better adoption, and a better handle on what’s working.” Venture-backed point solutions told operators to trust the roadmap; many never delivered, and as proptech funding dries up, some of those vendors are disappearing or freezing development — leaving operators holding maintenance contracts on thin tools. The tradeoff is real: consolidating to a platform often means trading depth in a specific function for cohesion across all of them. What it means: Before your next renewal cycle, audit every point solution in your stack for actual usage and integration health — if a tool isn’t talking cleanly to your PMS, you’re paying twice for the same data problem.

The Breakdown

The supply wave is cresting — and operators who treat concessions as a permanent cost of doing business will give away margin they’re about to earn back. The window to start pulling concessions is narrowing, but it hasn’t closed, and pulling too early in the wrong market is just as costly as pulling too late.

The supply math has genuinely shifted. Deliveries fell roughly 30% year-over-year in Q1 2026, and construction activity dropped to its lowest level since 2016. Further out, annual supply declined approximately 25% in 2025 to roughly 523,000 units and is forecast to contract by an additional 36% in 2026 to approximately 333,000 units — the lowest annual delivery total since 2014. That is a structural shift, not a seasonal dip. The question is no longer whether supply is falling. It’s whether your submarket is ahead of the national curve or still digesting the tail end of the pipeline.

The catch: lease-up backlog is masking the recovery. New deliveries down 30% sounds like relief, but nearly 1.3 million apartments remain in lease-up nationally, equal to 6.9% of total inventory — far above the pre-pandemic norm of roughly 700,000 units, or 4.7% of stock. That overhang is doing real damage. Concessions in newly delivered buildings are encouraging trade-ups into newer, higher-quality spaces, pressuring occupancy in older, stabilized stock. If your community is mid-vintage Class B, the competition isn’t just the new lease-up across the street — it’s the effective rent gap that lease-up’s concessions create. RealPage data shows Class C units continued to see markedly higher discounts, averaging 23.4% in April, compared to 13.2% for Class A and 14.5% for Class B. Class and vintage matter enormously here.

Geography determines your actual timeline. The national averages obscure a stark divide. Austin deliveries are projected to decline 47% in 2026, Denver supply is expected to be cut by more than half, and Phoenix faces an additional 40% decline following an 18% reduction in 2025. Those markets are closer to the inflection. Contrast that with Miami and Charlotte, which lead nationally with more than 8% of existing inventory still under construction. CBRE confirms the split: net absorption exceeded new construction completions nationally for the first time since Q2 2025, and average monthly rent ticked up 0.2% year-over-year to $2,217 in Q1 2026 — a fragile positive, not a green light. The timeline to achieve positive asking rent growth has been pushed to late 2026 for many high-supply markets.

What to actually do. Don’t pull concessions on a national narrative — pull them on your submarket’s deliveries schedule. Check how many units remain in lease-up within a one-mile competitive set, and track whether that number is falling month over month. Atlanta and Orlando, where construction pipelines have cooled sharply and are running below pre-pandemic trends, appear to be among the first markets for some stabilization. If you’re in a market like that, start testing concession rollbacks on renewals first — renewal rates are already running at 57% of all leasing activity, up from 51% in 2015 — where you have leverage. On new leases, shorten free-rent periods before eliminating them entirely; you’ll see the demand signal before you’ve given up the full incentive. The operators who time this right will recover 200–400 basis points of effective rent growth before the market broadly agrees the cycle has turned.

This Week’s Move

Pull your active concessions — every unit carrying free rent or a reduced deposit — and run them against your current vacancy rate and your submarket’s delivery calendar for the next 90 days. Supply is clearly tapering: deliveries fell roughly 30% year-over-year, and construction activity just hit its lowest level since 2016. That means the competitive pressure justifying each concession you granted six months ago may no longer exist in your submarket today. The move is to put your concession schedule into a large-language-model chat (ChatGPT, Claude, whatever you have) with this prompt:

I operate an apartment community. Below is a list of my current active concessions
(unit #, concession type, dollar value, date granted, and current lease expiration).
My current vacancy rate is [X%]. My submarket had [N] new units deliver in Q1-Q2 and
has [N] units remaining in the pipeline through Q3. Identify which concessions I can
likely phase out at next renewal without materially risking occupancy, and flag any
where the cost-to-carry exceeds what a $50-$75 rent increase at renewal would recover
within 12 months.

Concessions remain prevalent in supply-heavy submarkets as owners prioritize occupancy over rent growth — but “prevalent” is not the same as “necessary at your address.” After 18 months of muted rent growth driven by elevated new supply, rents are beginning to accelerate. If you’re still pricing like it’s peak supply season, you’re leaving money on the table that the market is now willing to give back.

Worth a Look

NAA × AppFolio, 2026 Property Management Benchmark Report: The clearest current read on the AI deployment gap — AI adoption among property managers rose from 21% in 2024 to 34% in 2025, yet only 8% of operators have fully automated a single workflow — and on the platform consolidation pressure, with 45% of operators signaling plans to consolidate their tech stacks. The full data set is worth an hour if you’re making any tooling decisions this year. appfolio.com/newsroom/property-manager-benchmark-survey-2026

EliseAI, 2025 State of AI in Multifamily: A vendor-produced report, so read it skeptically — but the competitive-pressure angle is worth pressure-testing: 77% of operators using AI report moderate-to-significant reductions in operating expenses, and 78% of respondents admitted they have already lost new business opportunities to AI-enabled competitors. eliseai.com/resources/the-state-of-ai-in-multifamily

MMCG Invest, U.S. Multifamily Market Outlook 2026 (March 2026): The supply math in one place — annual supply declined approximately 25% in 2025 to roughly 523,000 units and is forecast to contract by an additional 36% in 2026 to approximately 333,000 units, the lowest annual delivery total since 2014 — with market-by-market breakdowns that help operators distinguish where tightening concessions is already justified from where it still isn’t. mmcginvest.com/post/u-s-multi-family-market-outlook-2026

The operators who recover margin this cycle won’t be the ones who waited for the market to announce it was safe — they’ll be the ones who read their own absorption data and moved first.

— Josh