# The Hardest Lease Structures to Automate — and the Edge Cases That Break Generic Tools Blog | LeasePilot [Blog](/blog)CRE Expertise # The Hardest Lease Structures to Automate — and the Edge Cases That Break Generic Tools The specific CRE lease structures that break generic document automation — percentage rent breakpoints, multi-building CAM, co-tenancy, options, TI amortization — and why they need a deterministic engine, not a template. ![Lior Kedmi](/_next/image?url=%2Fleadership%2Flior-kedmi.jpg&w=3840&q=75&dpl=dpl_DCLYQhNPd3GNPNYxbbiBBmuuBWjA) Lior Kedmi CTO June 21, 202610 min readCopy link TL;DR Generic automation assumes a shallow conditional tree. Commercial leases don't have one. Here are the seven structures that break simpler tools — tiered percentage rent, multi-building CAM, co-tenancy, options, TI amortization, date cascades, the connected document set — and the deterministic logic each one actually needs. § 01 ## [Where the Demo Ends and the Portfolio Begins](#where-the-demo-ends-and-the-portfolio-begins) Every automation demo handles a standard deal. Fixed base rent, a flat escalation, one tenant, one building. The clauses drop into place and the draft looks clean. Then you bring your own leases. The anchor with a co-tenancy trigger. The retail deal with three percentage-rent tiers. The office lease whose commencement date floats until the certificate of occupancy lands. This is where most tools stall. The reason is structural. Generic document automation assumes a shallow conditional tree — a handful of if/then branches sitting on top of a fixed template. Commercial leases don't have a shallow tree. One deal term can rewrite language in a dozen sections, drive a calculation that has to actually compute, and change a schedule inside an exhibit three documents away. I argued the general case in [Supporting Any Complexity in Lease Documents](/blog/supporting-any-complexity-in-lease-documents). This post is the concrete version. Below are seven structures that break simpler tools, and the specific mechanism that breaks them. > **DISCLAIMER:** The example provisions here are illustrations only, not legal advice. Your clauses should be drafted by your own counsel for your specific deal. § 02 ## [Four Ways a Lease Term Propagates](#four-ways-a-lease-term-propagates) Before the examples, the four mechanisms. Every hard structure below trips at least one of them. **Deep conditional logic.** One deal term changes language in many places at once, not one. Choosing a guaranty type edits the signature block, the definitions, and a remedy clause together. **Calculations that resolve.** A provision isn't just words. It's a rent schedule, a breakpoint, an amortization table that has to produce the right number and print it into the document. **Cross-references that stay consistent.** Section 3.2 names a figure that Exhibit B has to match. Move one, and the other has to move with it, every time. **Inter-document synchronization.** The lease, the work letter, the commencement agreement, and the estoppel all cite the same dates and dollar figures. They have to agree. Generic tools handle the first mechanism at shallow depth and mostly punt on the other three. That gap is the whole story. § 03 ## [1\. Percentage Rent With Tiered Breakpoints](#1-percentage-rent-with-tiered-breakpoints) Percentage rent — additional rent a retail tenant pays once sales cross a threshold — sounds like one multiplication. It isn't. The breakpoint can be \_natural\_ (base rent divided by the percentage rate) or \_artificial\_ (a negotiated sales figure). It can be a single tier or several, each with its own rate. The first partial year may prorate, or not. The breakpoint may adjust with CPI, or hold flat. Each choice changes the number, and each changes the exhibit that reports it. A natural breakpoint recomputes the moment base rent moves. An artificial one holds until someone renegotiates. Tiers stack: 4% above the breakpoint, 3% above a second threshold, 2% above a third. A template with merge fields can print a breakpoint. It can't decide which breakpoint definition applies, recompute the natural figure when base rent shifts, and reconcile the exhibit against the body clause. That's a calculation plus a cross-reference plus a conditional, firing together. I go deeper on the traps in [Percentage Rent Drafting Traps for Retail Landlords](/blog/percentage-rent-drafting-traps-retail-landlords). § 04 ## [2\. Multi-Building, Multi-Tenant CAM](#2-multi-building-multi-tenant-cam) Common Area Maintenance (CAM) — the tenant's share of operating the shared parts of a property — is the most disputed provision in commercial leasing for a reason. The pool defines the math, and the pool is rarely simple. In a multi-building park, which buildings share which costs? Parking-lot expense might spread across the whole property; HVAC might stay inside one building. Each tenant's pro-rata share — its slice of the pool — depends on which pool you're measuring against. Then the modifiers stack. A \_gross-up\_ restates variable costs as if the building were fully occupied, so a half-empty property doesn't undercharge the tenants who are there. A cap limits year-over-year growth, and caps come in flavors: cumulative, compounding, on controllable costs only. Base-year exclusions carve out the first year as a reference. Every one of those is a conditional that reshapes the clause and a calculation that has to land the same way in the reconciliation. Change the cap type and the language changes, the number changes, and the exhibit changes. A tool that hardcodes one gross-up method quietly produces the wrong figure for every deal that differs. See [CAM Reconciliation: The Most Disputed Provision](/blog/cam-reconciliation-most-disputed-provision) for how much rides on getting it right. § 05 ## [3\. Co-Tenancy Clauses](#3-co-tenancy-clauses) Co-tenancy protects a retail tenant when the property loses the anchors or occupancy that drew them there. It's the clearest case of one condition rewriting the lease in many places. There are two triggers. \_Opening co-tenancy\_ governs whether the tenant has to open at all if named anchors aren't operating on day one. \_Ongoing co-tenancy\_ governs remedies if occupancy later drops below a threshold. Each carries its own consequence: reduced rent, a switch to a percentage-only rent, a right to go dark, or a kick-out — the right to terminate. The logic branches hard. If Anchor A goes dark, remedy X. If A and B both go dark, remedy Y. If total occupancy falls below a set percentage, the remedy escalates again, and a cure period starts a clock. Those branches touch the rent clause, the remedies section, the definitions, and the termination provision at once. This is deep conditional logic in its purest form. A shallow if/then can't carry a remedy ladder that reshapes four clauses and starts a date calculation. Encode it wrong and the tenant reads a right you didn't intend to grant. § 06 ## [4\. Options and Rights That Change Many Sections](#4-options-and-rights-that-change-many-sections) Renewal options, expansion options, a right of first offer (ROFO) or first refusal (ROFR) — these look like standalone clauses. They aren't. The presence of one option edits language across the whole lease. Grant a renewal option and you add option-rent mechanics, a notice window measured backward from expiration, and a term definition that now has to account for the extension everywhere it appears. Grant an expansion option and the premises definition, the pro-rata share, and the CAM pool all become conditional on exercise. A ROFO rewrites how the landlord markets adjacent space and how notice cascades. The hard part is presence-driven insertion. The ROFO clause has to appear exactly when the deal terms say it exists, in every place it touches, and vanish cleanly when it doesn't — with the cross-references still intact. Miss one dependent reference and the lease contradicts itself. I cover the mechanics in [Options: ROFO and ROFR Provisions](/blog/options-rofo-rofr-provisions). § 07 ## [5\. TI Allowance Amortization and Reconciliation](#5-ti-allowance-amortization-and-reconciliation) A tenant improvement (TI) allowance — landlord money toward building out the space — is a calculation wearing the clothes of a clause. If the landlord amortizes the allowance into rent, you need a schedule: the allowance amount, the rate, the term, the resulting monthly add-on. That schedule has to compute, print, and match the base-rent table it rides alongside. If the tenant overspends, the excess flows back through as additional rent. If the allowance disburses in phases against construction milestones, each phase carries its own documentation condition. Now connect it. The allowance figure in the lease has to equal the figure in the work letter. The amortization has to reconcile against the rent schedule in the body. The commencement agreement later confirms the final number once the buildout is measured. That's a calculation, a cross-reference, and inter-document synchronization in a single provision. A merge field can drop a dollar amount. It can't amortize it, reconcile it against the rent table, and keep it equal to the work letter. [The economics landlords get wrong on TI allowances](/blog/tenant-improvement-allowances-economics-landlords-get-wrong) walks through where the money actually leaks. § 08 ## [6\. Date-Driven Cascades](#6-date-driven-cascades) Dates aren't fields. They're triggers, and in a lease they cascade. Consider a commencement date defined as the earlier of ninety days after the certificate of occupancy or the tenant's substantial completion, but never before the first of the following month. That single definition compares two inputs, adds ninety days, applies a month adjustment, and lands a date. Then the date moves everything downstream. The rent schedule shifts. The free-rent period repositions. Every escalation step recalculates against the new anchor. Option-notice windows measured from expiration slide. The critical-date schedule — the list of deadlines the asset manager lives by — rewrites itself. A date picker stores a value. It can't hold a formula that recomputes a dozen dependent provisions when the input changes. When commencement slips two weeks, a deterministic engine reflows the escalations and exhibits together; a template leaves you to catch the ones you forgot. [Rent Escalation Clauses: The Complete Guide](/blog/rent-escalation-clauses-complete-guide) shows how tightly the escalation math couples to the commencement anchor. § 09 ## [7\. The Connected Document Set](#7-the-connected-document-set) A lease rarely travels alone. It moves with exhibits, a work letter, an SNDA (subordination, non-disturbance, and attornment — the agreement that orders the tenant's rights against the lender's), a commencement agreement, and later an estoppel (a signed confirmation of the lease's key facts). These documents share facts. The premises square footage, the commencement date, the TI figure, the rent schedule — each appears in several of them, and each has to agree across all of them. When a deal term changes late, the change has to reach every document that cites it. This is the failure mode that survives even careful drafting: the lease gets updated, the work letter doesn't, and nobody notices until closing. The provisions look right in isolation and contradict each other across the set. I unpack this in [The Hidden Complexity of Ancillary Documents](/blog/hidden-complexity-ancillary-documents). Synchronization is the mechanism a template can't offer, because a template is one document. You need a model where the deal terms sit above the documents and every document reads from them. § 10 ## [The Pattern, in One Table](#the-pattern-in-one-table) Structure Why it breaks generic tools What deterministic automation needs Tiered percentage rent Breakpoint type and tiers change both the number and the exhibit Calculation engine + body-to-exhibit cross-reference Multi-building CAM Pool, gross-up, and cap type reshape clause and figure together Conditional logic + reconciliation math Co-tenancy One trigger drives a remedy ladder across four clauses Deep branching + date-triggered cures Options and rights Presence edits language and references site-wide Presence-driven insertion + reference integrity TI amortization Allowance must compute and match the work letter Calculation + inter-document sync Date cascades Commencement shift ripples through schedules and exhibits Dates as formulas, not fields Connected document set Shared facts drift out of sync across documents Deal terms above the documents § 11 ## [Why Deterministic, Not Probabilistic](#why-deterministic-not-probabilistic) Read the right-hand column again. Every entry is a guarantee: the same input resolves the same way, every time. That's what a deterministic engine gives you. Enter the deal terms once; the correct language, the calculations, and the cross-references resolve identically on every draft. This is exactly where a probabilistic model can't stand in. Ask a large language model to draft the co-tenancy remedy and it will produce fluent, plausible text. It can't promise the ROFO clause appears in all six places the condition demands. It can't guarantee Exhibit B's breakpoint schedule equals the figure in Section 3.2, or that the work letter's TI number survives a late change to the lease. Plausible isn't the same as consistent, and lease math has to be consistent. The gap isn't that generic tools are weak. It's that they assume a simpler conditional structure than commercial real estate has. During onboarding we encode your breakpoint logic, your CAM pools, your co-tenancy ladders, and your date formulas into the platform, so the hard structures stop being the manual residue and start being part of the draft. For the fuller argument on why generic automation never took hold in CRE, see [Why Generic Document Automation Failed CRE Legal Teams](/blog/why-generic-document-automation-failed-cre). The test for any tool is the same one I'd apply to a new hire: hand it your hardest lease from last year. If the answer is "simplify your terms," the tool is the constraint. If the answer is "let's encode the logic," you're looking at a system built for the leases you actually sign. § Adjacent reading ## More from the ledger [§ 01APR 18, 2024 CRE Expertise ### Rent Escalation Clauses: The Complete Guide to Getting Them Right LeasePilot Team11 MIN READ Read →](/blog/rent-escalation-clauses-complete-guide) [§ 02MAR 03, 2026 CRE Expertise ### The Complete Guide to Commercial Lease Clause Drafting LeasePilot Team10 MIN READ Read →](/blog/complete-guide-lease-clause-drafting) [§ 03JAN 22, 2025 CRE Expertise ### The 7 Most Expensive Clause Mistakes in Commercial Leases David Saltman9 MIN READ Read →](/blog/expensive-clause-mistakes-commercial-leases) § See it in practice ## Reading about it is one thing. Watching it happen is another. See LeasePilot draft a lease in your team’s own templates, with your clauses and your defaults. [Schedule a Demo](/demo)