Long & McQuade Gear Rental

Today, renting from L&M means browsing a category page, finding items one at a time, and submitting a quote request that staff resolve manually. There is no real-time availability, no way to build a multi-item kit in one session, and no visibility into which branch actually has stock. For a first-time renter picking up a single mic, that friction is tolerable. For an event professional assembling a full PA rig across multiple days and potentially multiple branches, it is a dealbreaker.

Peer-to-peer platforms like ShareGrid and KitSplit have solved per-item availability calendars and multi-item insurance stacking - but they are peer-to-peer marketplaces, not a single retailer’s own inventory. The gap nobody has closed is real-time availability, multi-item kit curation, and cross-branch fulfillment from one company’s own stores, combined into a single self-serve flow.

The first-time renter enters through Browse and Search - a guided category filter with date-range availability, pulling real L&M product data into a card grid. The event professional enters through GearGenie - a deterministic kit builder that takes event type, guest count, venue type, and dates, then returns a recommended gear list.

Both paths converge at the Cart, where the cross-location fulfillment resolver does the work nobody else has built - detecting when items in a single order span multiple branches and offering consolidation. From there, Checkout surfaces deposit terms and delivery options inline, and the flow fans out to Pickup and Return (mobile-first condition check-in) and Order History (repeat-kit reordering for pros).

The recommendation engine takes four inputs: event type, guest count, venue (indoor, outdoor, or mixed), and rental dates. It runs deterministic rule matching - not a black-box LLM - against two data layers: four named kit templates that L&M staff already quote manually (Standard PA, Panel/Conference, Recording/Hybrid Meeting, Presentation + Projector), and a heuristic rules table covering scenarios those templates do not (weddings, DJ sets, outdoor festivals).

When a request matches a named template closely, GearGenie returns it directly - fast and predictable. When it does not, the engine looks up target wattage, speaker counts, mic counts, and mixer channel minimums from the heuristic rules, then matches those specs against real in-stock items from L&M’s inventory to assemble a live gear list. Every recommendation links back to actual catalog data. Nothing is invented.

Most recommendation systems present every result with equal confidence. GearGenie does the opposite: the UI changes based on how the result was produced. A template match gets a green badge and a named package; a heuristic assembly gets a different label and a tier disclosure; a no-match gets an honest fallback. The user can always tell whether they are looking at a known combination or a computed suggestion.

The resolution logic runs in strict priority order: template match first, heuristic fallback second, no-match last. When a guest count falls between two heuristic tiers, the engine defaults to the nearer or larger tier and surfaces which tier it matched. When an item in the suggested list has no confirmed price, it displays “Price to be confirmed” on that line item rather than showing $0 or silently excluding it from the total. The design principle is the same throughout: if the system is uncertain, say so - do not paper over the gap.

When a user adds items that are stocked at different L&M locations, the cart detects the split and surfaces it immediately: “Your kit will be split across 2 locations - Toronto (Bloor St.) and Toronto (Yorkdale) - you will need two pickups.” A Consolidate button lets the user attempt to source everything from one branch, subject to availability.

This is not an edge case. For an event pro renting a full PA setup, cross-branch fulfillment is the norm. The competitive teardown found that neither ShareGrid nor KitSplit has solved this because they are peer-to-peer - items come from different owners, not different branches of the same retailer. L&M’s multi-branch inventory is the asset; the cart is the feature that makes it usable.

On L&M’s current site, critical rental information - ID requirements, deposit amounts for cash payments, hold fees for advance reservations - lives on a separate FAQ page. A renter discovers these terms after they have already committed to the rental, not before. The redesign moves this information to two places where it belongs: the Gear Detail drawer (inline with the item the user is considering) and the Checkout screen (inline with the payment form).

Pricing is shown as day, week, and month tiers by default. When the user has already set a date range upstream in Browse or GearGenie, the detail drawer resolves to an actual total for that exact rental window instead of showing only generic rates.

GearGenie’s three trust states let the system be honest about how much inference went into a recommendation rather than presenting everything with equal confidence. The cross-location fulfillment resolver solves a real logistics gap that peer-to-peer competitors structurally cannot address - their items come from different owners, not different branches of the same retailer.

The decision to use deterministic rule matching over an LLM for the recommendation logic was validated in testing: users trusted the output more when they could see the underlying rule (“6-8W/guest, indoor, 300-400 guests”) than when they received an opaque “AI-generated” suggestion. The “Why this list” disclosure, initially designed as a pro-user feature, turned out to matter for first-time renters as well - it answered the question “can I trust this?” before it was asked.

What is next: the mobile-first Pickup and Return condition check-in flow (photo capture for pre/post-rental damage documentation), visual polish applying iOS’s Liquid Glass design language, expanded heuristic rules for outdoor events and festivals, and real-time inventory sync for live stock status.