Corporate Transportation Smart Mobility

The Complete Guide to Employee Transportation Management in 2026

· 24 min read
A corporate commuter shuttle loading employees at a suburban office campus at shift change

Roughly one in six US employees now spends more than 45 minutes getting to work one way (around 17% on the combined 45+ minute tails of the ACS distribution), and the mean has climbed back to 26.8 minutes after a brief pandemic dip (US Census Bureau, American Community Survey 2023). Return-to-office mandates have reversed the hybrid-work thaw: by late 2024, roughly 75% of US workers were required to be in the office regularly, up from 63% in early 2023 (Pew Research Center), averaging 2.78 days in-office per week (FLEX by Fenwick, Q4 2024). Scope 3 Category 7 (employee commuting) is now a disclosure line on two major regulatory sheets: EU CSRD post-Omnibus, and California SB 253. The corporate employee transportation market sits at roughly $38.5B in 2025 and $41.7B in 2026 (The Business Research Company), consistent with estimates from Mordor Intelligence and HTF Market Intelligence.

This piece is written for Operations and Facilities VPs, HR directors, sustainability leads, and mobility program managers at employers in the 200–20,000+ range: the people who own the cost-per-mile, parking load, and retention consequences of how employees get to work. It covers cost anatomy, the decision framework for fixed vs dynamic vs hybrid operating models, a rollout sequence with realistic ranges, a vendor rubric across eight dimensions, and a KPI set that survives board scrutiny. Public transit agency planning, school-bus operations, rideshare consumer apps, and individual-reimbursement payroll mechanics sit outside the scope.

What “employee transportation management software” actually is in 2026

A decade ago, “employee transportation management software” meant a scheduling tool glued to a roster spreadsheet. That has changed. An enterprise-class platform in 2026 combines five functions under one data model: route and trip planning against live rider rosters, dispatcher operations and incident handling, the rider booking app, HRIS- and payroll-integrated eligibility, and analytics feeding both finance dashboards and Scope 3 reporting.

What it is not: a fleet-management system. Fleet software tracks vehicles, fuel cards, and mechanic workflows. A workforce transportation platform tracks riders: who is eligible, who booked, who boarded, who didn’t, and what the ride cost per actual occupied seat-mile. The vehicle telematics layer is an input, not the product. Employers confuse the two during procurement more often than they should. Transit management systems (TMS) used by public agencies are the other adjacent category; they optimize for published schedules, fare collection, and GTFS feeds, while corporate shuttle software optimizes for private eligibility lists, shift patterns, and internal cost-center allocation.

Four functional boundaries are worth naming before you write an RFP. Pre-tax commuter-benefit administration is usually a separate SaaS product (Edenred, Benepass, HealthEquity); some corporate shuttle platforms offer light integration, but none are the pre-tax benefits administrator of record. Ground-transportation brokerage for executive rides (Blacklane, Wheely) is a different service class. Carpool-matching apps (Scoop, Hytch) address a narrower origin pair-matching problem and cap out below ~500 employees in most single-site deployments. Parking-management software handles permits and stall allocation; the better corporate shuttle platforms ingest parking data to size shuttle capacity but don’t replace the permit system.

That platform class is what buyers typing “employee transportation management software” into Google actually need. Ryde’s smart employee commuting platform sits in this category, as do the named vendors profiled later. The category spans fixed trunk-line shuttles (hospitals, manufacturing), dynamic on-demand microtransit (hybrid-work campuses), and multi-mode hybrid networks using fixed corridors for dense origin pairs and dynamic vehicles for the long tail.

The cost anatomy finance teams keep missing

Most corporate-transportation business cases fail at finance review because the author built them around operating cost only: driver hours, fuel, vehicle lease. Those are real, and the smallest of five categories a good employee commute management system touches. The other four are where the return lives.

Start with parking. The 2024 national median construction cost for a new parking structure was $29,900 per space, up 3.1% year-over-year, with above-ground typical at $21,000–$35,000 per space and underground at $60,000–$120,000 (WGI, Parking Structure Cost Outlook for 2024). That is capex per net-new space; land, lighting, security, and ongoing maintenance ride on top. Donald Shoup’s long-cited figure that 95% of US auto commuters park free at their workplace (Shoup, The High Cost of Free Parking, UCLA, 2005/2011) understates the real cost: it’s free to the employee, not to the employer, who absorbs it. A 1,200-employee site that avoids adding 300 parking stalls has avoided $6.3M to $9M in capex alone, at median.

A second line sits on the employee side. Total cost to own and operate a new vehicle reached $12,297 per year in 2024 ($1,024.71/month), assuming 15,000 annual miles across a 5-year hold; depreciation alone averaged $4,680/year (AAA, Your Driving Costs 2024). A single-vehicle household saves that entire line if a reliable employer shuttle replaces the second car, and HR teams consistently underweight it when pricing shuttle utility for a retention story.

Turnover and absenteeism traceable to commute form the third category. Replacing a departed employee costs 6–9 months of salary, or 50–200% depending on role seniority (SHRM turnover benchmark). Santelli and Grissom, in AERA Open (2024), found each additional 5 minutes on a one-way commute is associated with 0.8 to 1.0 percentage points of added voluntary transfer probability across most of the distribution. Absenteeism costs US employers an estimated $225.8B annually, or roughly $1,685 per employee (CDC Foundation, 2015 methodology, order-of-magnitude only). Circadian Technologies, in Absenteeism: The Bottom-Line Killer, put the figure at $3,600/year for hourly and $2,650/year for salaried, including temporary-staffing cascade costs.

Disengagement is the fourth line. Global workforce engagement sits at 23%, with 62% not engaged and 15% actively disengaged; disengaged employees run 37% higher absenteeism and 18% lower productivity (Gallup, State of the Global Workplace 2024). Gallup puts the global cost of low engagement at $8.9T annually, around 9% of GDP. A single disengaged worker at average US salary costs an employer roughly $16,000 a year. Commute friction feeds this line alongside several other factors; rarely the whole of it, but not zero.

Disclosure overhead rounds out the five. CSRD/ESRS E1 and California SB 253 both defer to GHG Protocol methodology for Scope 3 Category 7. Compliance cost is non-trivial (data collection, assurance, third-party software), and commute emissions usually rank between the second and fourth largest Scope 3 category for office-heavy employers.

Worked example: a 1,200-person manufacturing site in Atlanta, two shifts, 80% drive-alone baseline.

  • Parking: 960 stalls avoided at $29,900 = $28.7M avoided capex (amortized over 30 years at 5% ≈ $1.87M/year).
  • Direct commute subsidy or reimbursement: 960 drivers × $162/month average = $1.87M/year.
  • Absenteeism attributable to commute friction (low estimate, $210/employee × 1,200): $252K/year.
  • Turnover: 15% annual attrition × 1,200 × $45,000 salary × 75% replacement cost × 15% of quits citing commute = $912K/year.
  • Scope 3 Category 7 reporting + assurance: $60K–$120K/year.

Site-level annual commute spend, before the shuttle contract: roughly $4.9M. Most finance teams see the $1.87M subsidy line and miss the other four. That is the conversation an employee shuttle management software vendor actually competes for.

Fixed, dynamic, and hybrid: how to pick an operating model

Three operating models cover the field. Fixed routes run scheduled stops at scheduled times, looping a corridor whether or not every seat is filled. Dynamic microtransit routes vehicles against real bookings; seats and pickup points shift trip by trip. Hybrid networks run fixed trunk lines on dense corridors and dynamic vehicles on the long tail. The smart shuttles category covers fixed and hybrid configurations common at manufacturing and healthcare sites.

Operating model drives 60–70% of per-rider cost and close to 100% of rider experience. Most employers pick wrong the first time because they frame the question around the vehicle, not the demand.

DimensionFixed routesDynamic / on-demandHybrid network
Best forSingle-shift, dense-corridor, predictable rostersHybrid-work offices, dispersed origins, T/W/Th peaks200–2,000 employee sites with mixed demand
Rider NPS (typical range)0 to +25+30 to +60+20 to +45
Cost per rider-trip (order of magnitude)$5–$12 at 65% occupancy$12–$28$8–$18
Schedule volatility toleranceLowHighMedium
Vehicle utilizationHighest (if occupancy holds)Lower, but per-rider cost can compete via dynamic poolingMiddle; depends on corridor mix
Break-even density≥ 40 riders per corridor per shift~8–15 riders within a 1.5-mile origin clusterVaries; typically picks off the dense 2–3 corridors and dynamics the rest
WeaknessEmpty seats on T/F, overflow on WCoordination and dispatch overhead; driver-hour utilizationOperational complexity; needs a unified platform

Employer-size breakpoints help, though these are a framework built from vendor playbooks, not a single canonical citation. Under 200 employees, a shuttle program rarely pencils out except in narrow cases: a hard-to-reach site, a healthcare employer with a night shift, a logistics facility with no parking. Most sub-200 employers get more mileage from pre-tax benefits plus carpool matching. Between 200 and 800, dynamic or hybrid usually wins on both cost and NPS, except at single-corridor sites where a fixed pair of morning/evening trips at high occupancy beats everything. From 800 to 2,000, hybrid is the default, with fixed lines on the top 2–4 corridors and dynamic on the tail. Above 2,000 at one site, multi-vendor hybrid networks (like those Microsoft, Google, and Apple run) become viable and often necessary.

Microsoft’s Connector program shows how these mixes evolve. As of the Seattle Times’ 2017 coverage, the Seattle-Redmond network grew from a 2007 pilot into 19 routes operating 53 buses with fleet capacity above 7,000 daily passengers; daily ridership sat around 2,160 round-trip riders, and a 2012 Microsoft Green Blog post reported roughly 60% of lifetime Connector riders had driven solo before switching. Connector remains substantially fixed because Redmond demand is dense and predictable; Microsoft adds flexibility at the edges rather than flipping the model.

RTO demand has moved the calculus. A fixed M–F shuttle with 65% Monday and Friday occupancy and 95% Wednesday overflow is a worse answer than a hybrid network shifting capacity into the midweek peak. If your RTO pattern is T/W/Th-heavy, and the Fenwick data says most are, a pure fixed-route platform bought this year will underperform a hybrid one within two quarters. Several vendors now sell “dynamic top-up” as a product: a fixed backbone with on-demand vehicles spinning up for the Wednesday spike. One caveat on dynamic: rider NPS is almost always higher than for fixed shuttles, but dispatcher workload is the hidden tax. Plan for 1.0–1.3 dispatcher FTE per 500 daily bookings in the first six months before routing models settle.

2026 compliance: what IRS §132, SB 253, and CSRD actually require

The regulatory picture for commute programs in 2026 is genuinely more complex than 18 months ago. Five buckets of rules are live: federal pre-tax benefits, state and city mandates, the SEC climate rule (partial), California SB 253, and EU CSRD post-Omnibus.

Federal pre-tax benefits haven’t changed in structure since TCJA but the dollar amounts move annually. For 2026, the IRS set the monthly exclusion for qualified parking at $340 and the monthly exclusion for commuter highway vehicle and transit passes at $340, both up from $325 in 2025. An employee can take both simultaneously (the buckets are separate). TCJA eliminated the employer deduction for qualified transportation fringe benefits in 2018 but left the employee tax exclusion intact, a point that gets relitigated in benefits procurement more often than it should.

State and city commuter-benefit mandates layer on top of the federal baseline. New Jersey was the first statewide mandate in 2019 (20+ employees), joining NYC (2016, 20+), DC (Transportation Benefits Equity Amendment Act 2020, 20+; a separate 2016 pre-tax commuter ordinance also applies at 20+ employees), the Bay Area BAAQMD (2014, 50+), Illinois/Chicago RTA (2024, 50+), Seattle (2020, 20+), and Philadelphia (50+).

JurisdictionThresholdYearPenalty range
New Jersey (statewide)20+ employees2019$100–$250 first offense
New York City20+ employees2016$50–$2,000 first offense
Washington, DC20+ employees2020 (Transportation Benefits Equity Act)~$100 per employee per month under the Clean Air Compliance Fee
Bay Area (BAAQMD)50+ employees2014Varies by district
Illinois / Chicago RTA region50+ employees2024 (Illinois Transportation Benefits Program Act)Compliance order; fines on non-response
Seattle20+ employees2020Graduated
Philadelphia50+ employees2022Warning then fine

On the climate side, the SEC’s March 2024 climate disclosure rule dropped Scope 3 before adoption and remains subject to judicial stay regardless. The same does not hold in California or Europe.

California SB 253 requires US-based entities with over $1B in annual revenue doing business in California to disclose Scope 1 and 2 emissions starting in 2026 (first deadline August 10, 2026) and Scope 3 from 2027. CARB’s framework phases assurance (limited by 2027, reasonable by 2030) and applies good-faith discretion for initial Scope 3 disclosures, with Category 7 (employee commuting) named among the initial categories reporters need to estimate. California nexus plus $1B in revenue puts you in scope regardless of where your headquarters sits.

EU CSRD, after the February 2026 Omnibus I package, applies to EU companies with more than 1,000 employees and more than €450M turnover, a substantial narrowing from the original thresholds. Wave 1 companies already in the regime continue through FY2025 / 2026 reports. ESRS E1 requires disclosure of gross Scope 3 emissions across all material categories, including employee commuting where it passes the materiality test. Most office-heavy employers find commuting to be material.

GHG Protocol defines three approved methods for Category 7 (Chapter 7 technical guidance): fuel-based, distance-based, and average-data. Distance-based is the most common for corporate reporting; it requires a commute survey plus EPA or equivalent emission factors. EPA’s 2024 GHG Emission Factors Hub is the US canonical reference, with a typical passenger vehicle at roughly 400 grams CO2 per mile. APTA’s published methodology puts the average US transit system at about 0.45 pounds CO2 per passenger-mile across all modes; bus-specific figures run closer to 0.64 lb, light rail closer to 0.36 lb. A 20-seat diesel shuttle at 28% occupancy runs roughly 53% below a single-occupancy sedan on a per-passenger-mile basis; at 100% occupancy on the same vehicle, the reduction is in the range of 80%+. Treat those figures as directional and recalculate against your own fleet spec before reporting.

Two practical implications. If you’re a US employer with EU subsidiaries above the post-Omnibus threshold, or do substantial business in California, you have Scope 3 Category 7 obligations even if the SEC rule never activates. And the cleanest path to auditable commute data is a platform with a per-rider distance ledger; survey-based estimates carry assurance risk, platform-captured rider-miles do not.

The realistic 60–90 day rollout (and when you can compress it)

Most corporate shuttle software rollouts take between 60 and 90 days from kickoff to full-network go-live, though the range holds wide. The phases are stable across vendors; what changes is how much of the work has already been done before the SOW is signed.

Weeks 1–2: commute-origin data gathering and survey. Pull home ZIP codes from HRIS (not self-entered survey addresses, which drift and bias), build an origin heat map with shift patterns overlaid, and run a targeted survey for travel-time and modal preference. This phase most often runs long because HR owns the HRIS and ops owns the shuttle, and the two calendars rarely align.

Weeks 3–4: route and corridor design. With origin distribution in hand, the platform’s routing engine or a qualified operations lead lays out candidate corridors against the shift calendar. Question to resolve at the end of week 4: which two or three corridors have the density for fixed trunk lines, and what does the dynamic tail look like for remaining origins. Vendor workshops belong in this window, not earlier.

Weeks 5–6: pilot corridor go-live. One corridor, two to four vehicles, 50–200 daily riders. The pilot validates load, timing, driver roster, and dispatch workflow (demand validation is what the survey was for). Expect OTR to sit in the 80s at first and climb into the 90s as routes settle. Rider NPS should clear +20 by end of week 6 or something is misaligned.

Weeks 7–10: full-network rollout. HRIS integration goes live, push notification flows are tested, SSO is enforced, and payroll code mapping for any cost-share is wired up. A common trap: teams leave HRIS integration for “phase 2” and end up managing eligibility in a spreadsheet for six months.

Weeks 11–12+: KPI review at 30 / 60 / 90 days post-full-rollout. First 30 days are noisy; 60-day numbers are directional; 90-day numbers are the ones to benchmark against. Route adjustments in the 6–8 week range are normal and should be planned for, not treated as scope creep.

Compression is possible. Ryde has implemented programs in 20 business days for sites where HRIS data is clean, origin concentration is high, and existing shuttle operator relationships can be onboarded to the platform rather than rebuilt.  That pace requires operational maturity most first-time buyers don’t have; budget for 60–90 days and treat faster as upside. Compression levers are consistent: HRIS integration pre-negotiated with IT; a named operator relationship in place, or the platform running software-only on the employer’s fleet; a single pilot corridor agreed before kickoff; and an executive sponsor who can unblock SSO and procurement in hours.

Do the survey and the HRIS data pull in parallel and reconcile in week 2. Name the KPI targets (cost per rider-trip, occupancy, rider NPS) before pilot go-live. Teams that wait until “we have data” end up benchmarking against the pilot itself.

A vendor rubric: eight dimensions, scored 1 to 5

A vendor evaluation rubric beats a feature checklist because feature parity is widespread; what separates platforms is how the features fit together under operational load. Score each of these dimensions 1–5, weighted by your own priorities.

  1. Functional fit. Does the platform support fixed, dynamic, and hybrid in one data model, or are they separate products? Multi-site? Shift-pattern support (three-shift manufacturing, rolling schedules, seasonal peaks)? A 5 covers all three models natively; a 3 supports one well and bolts on the others; a 1 is single-model only.
  2. Integrations. HRIS (Workday, BambooHR, SAP SuccessFactors), payroll, SSO and identity (Okta, Azure AD), dispatch, and telematics. 5 = pre-built connectors for all major HRIS and SSO providers; 3 = SFTP + custom work; 1 = spreadsheet upload as the eligibility system of record.
  3. Optimization engine. Constrained vehicle routing (CVRP), time-window variants (VRPTW), re-optimization latency for dynamic bookings. Ask vendors for their re-optimization latency target at peak load and their approach to vehicle rebalancing.
  4. Rider app quality. Booking flow, ETA stability, push notifications, offline handling, accessibility (screen readers, color contrast, large-text support). Rider NPS is downstream of this more than anything else. Ask for the app; don’t trust screenshots. For the experience layer, examine how vendors like Ryde approach rider experience.
  5. Operations support. Dispatcher workflow, incident escalation, exception handling (“my shuttle didn’t show up”), 24/7 support SLA with named coverage. Ask for the dispatcher UI, not the admin dashboard.
  6. Pricing model. Per-seat, per-mile, per-trip, or platform + service unbundled? Which line scales linearly with your ridership and which is fixed? Hidden-cost risks: rider-app licensing per-active-user, dispatcher licensing, reporting tier gates.
  7. Reporting and analytics. KPI dashboards, cost-per-rider, occupancy, Scope 3 exports mapped to GHG Protocol Category 7 methods. Can the platform produce audit-ready Scope 3 data for CSRD and SB 253?
  8. Security and compliance. ISO 27001 and/or SOC 2 Type 2, GDPR for EU employees, regional safety regulations (e.g., women-safety protocols in India, required last-drop escort procedures), role-based access control. ISO 27001 and SOC 2 have significant but not complete control overlap (published mappings put the figure anywhere from 53% to 95% depending on scope), so one is generally sufficient, though European enterprise buyers frequently require ISO 27001 specifically.

On the named vendor field as of 2026: Zeelo’s public website puts its corporate-customer count at around 200+, with named accounts including Amazon, Tesco, and Fidelity International. MoveInSync leads in India with deep female-employee safety compliance support. Routematic covers shift transport, shuttle, rental travel desk, and parking across APAC. In North America, TripShot leans toward university and hospital fixed routes, RideCo is microtransit and paratransit-centric, Liftango’s differentiator is built-in Scope 3 reporting, and SHARE Mobility positions itself on on-time performance as a headline metric. WeDriveU is the largest US contract-operator relevant to big-tech shuttle programs; Optibus is transit-adjacent and worth evaluating for public-transit style fixed networks.

This list is an overview, not a ranking. Multi-shift manufacturing will load dimensions 1, 2, and 5 heavily; a tech-campus hybrid program leans on 3, 4, and 7; a regulated-market deployment (India, EU, Bay Area municipalities) weights 8 above most others.

The seven KPIs that survive board scrutiny

Metrics proliferate in early-stage shuttle programs. The board wants seven. Build the dashboard around these.

1. Cost per rider-trip. Total operating cost divided by completed trips: (driver hours × loaded hourly + fuel + vehicle amortization + platform license) / completed trips. Variation across operating models is the signal: a fixed-route program at $6 per rider-trip at 70% occupancy is healthy; at $14 per rider-trip at 38% occupancy, something is wrong with route design or corridor choice.

2. Cost per occupied seat-mile. operating cost / (vehicle-miles × average occupancy). This normalizes for vehicle size and route length. A 14-seat shuttle at 70% occupancy and a 56-seat charter bus at 45% occupancy look very different on cost per rider-trip but often converge here.

3. Load factor (seat occupancy). Actual occupancy divided by vehicle seat capacity, averaged across peak-service windows. Target 60–80% at peak. Below 45%, the route design needs revisiting; above 90%, riders are standing or getting bumped, which kills NPS.

4. On-time rate (OTR). Trips arriving within a defined tolerance window (often ±5 minutes for fixed, ±3 minutes for dynamic pickup ETAs) divided by total trips. A 95% threshold is a commonly stated corporate-shuttle SLA target; expect 85–90% in the first 90 days while routes settle.

5. Rider NPS. Administered via a low-friction in-app prompt at month 1, month 3, and month 6 post-rollout. The transportation sector at large runs low (public NPS benchmarks for rental cars and ground transport frequently sit below zero), so corporate shuttles should set higher expectations because the rider is the employer’s own workforce. A well-run program should clear +30 within two quarters; pure dynamic programs can reach +50 to +70 if dispatch is clean.

6. Scope 3 Category 7 reduction. Year-over-year reduction in kg CO2e per employee, calculated against a baseline that uses the same methodology year-to-year. Distance-based (GHG Protocol) is the default. A shuttle program replacing single-occupancy-vehicle trips at a 60% modal shift typically delivers 30–50% reduction in year one; year-two gains come from occupancy tuning, not new routes.

7. Commute-attributable voluntary attrition. Year-over-year change in voluntary quits where commute is cited as a contributing factor, captured through the exit-interview process. Softer than the others and subject to interview-design bias, but it is what the CFO will ask about when ROI is challenged. Pair with a counterfactual metric (attrition rate among shuttle-eligible vs non-eligible cohorts at the same site) to strengthen the causal story.

Three metrics that look useful and aren’t, or not as the headline number. Raw ridership (use rider-trips per eligible employee instead). Fuel savings (volatile with market prices; let it drop into the Scope 3 line). Vehicle utilization as a percentage of service hours (measures the vehicle, not the rider).

A pattern the best-run programs share: KPI targets are set at pilot kickoff, not at quarter-end. Rider NPS by day 90, OTR by day 60, cost per rider-trip at steady state; write them down before the pilot goes live. That single discipline separates programs with ROI stories from programs with ridership stories.

Four rollout pitfalls that kill ROI

Programs that underperform almost always trip on one of four failures, and all four are avoidable.

Pitfall one: designing routes on first-draft survey data. Surveys capture stated preference; real rider behavior differs. Teams that lock in routes at week 3 based on raw survey responses, without a validation check (origin ZIPs cross-checked against HRIS, or a “ghost ride” measuring real travel time), almost always rework half the network in month two. Better: use the survey to identify candidate corridors, then validate against HRIS and traffic-time actuals before committing vehicles.

Pitfall two: buying fixed-route software when your pattern is hybrid. RTO schedules are T/W/Th-heavy for a growing majority of employers. A purely fixed-route platform cannot reallocate capacity from Monday and Friday into Wednesday. Programs built on fixed-only platforms in 2023–2024 and now carrying hybrid workloads are the largest replatforming cohort in the vendor market. If your workforce is even partially hybrid, insist on native hybrid support in the contract, not as a “phase 2 module.”

Pitfall three: leaving HRIS integration for later. Manual roster management looks fine at 200 riders. At 1,500, it is the slowest-to-fix problem in the program. Every onboarding, termination, and cost-center change lands on an ops admin with a spreadsheet; eligibility errors pile up; billing disputes reach the CFO. Insist on live HRIS connectors (Workday, BambooHR, SAP SuccessFactors) during vendor evaluation, with SCIM or equivalent provisioning. A vendor who says “we handle it through CSV uploads” is signaling either immaturity or a pricing trick.

Pitfall four: underinvesting in the rider app. A shuttle can be on-time, well-routed, and safely operated, and still hit +5 NPS because the app is bad: unstable ETAs, unreliable push notifications, a login flow that breaks on SSO, an accessibility gap. Rider NPS tracks app quality more than any other factor. Have three people from different roles (night-shift worker, hybrid office worker, sight-impaired user if available) actually use the app for a week on a demo tenant.

Two pitfalls worth naming but ranking below these four: over-promising on-time rate in the first 90 days (85–88% is realistic early, not the 95% the marketing deck promised), and designing the KPI dashboard after launch instead of before. A final pitfall on dynamic-capable platforms: running without a hard budget cap for pilot-phase dispatcher hours. The ops lead who owns the shuttle budget often doesn’t own the dispatcher hours. Get that line into the budget narrative at kickoff.

What to do next

Two decisions drive 80% of the outcome: which operating model fits your workforce, and how you weight the vendor dimensions against your reality. Get those right and the rest is execution.

Three concrete actions for this week. Pull the parking-cost line from facilities, operating cost plus avoided-capex for any spaces you’d otherwise need to add; that is the biggest hidden category in the cost anatomy, covered further in the upcoming reducing corporate transportation costs post. Run a commute-duration distribution from HRIS home ZIP codes against site addresses; a one-day exercise that tells you whether your workforce has the origin density for fixed corridors. Draft a five-vendor shortlist, score it against the rubric, and put at least three vendors through a scored demo in six weeks; a vendor comparison for 2026 can anchor that shortlist. Named customer case studies calibrate what “good” looks like.

The 2026 environment is not kind to static shuttle programs. Commute times are climbing, RTO mandates have reinstated midweek peaks that break M–F networks, Scope 3 disclosure is a board-level line at more employers, and the cost stack keeps shifting. Employers still running healthy commute programs in 2028 will be those treating the program as a data system first and a vehicle operation second.

If your team is scoping a new program or replatforming an existing one, contact Ryde and we’ll walk through the rubric against your site profile.

Sources