PRIVILEGED AND CONFIDENTIAL — ATTORNEY WORK PRODUCT
MEMORANDUM
TO: James Underwood, Lead Associate, Calderwood & Harkness LLP
FROM: Devon Archer, Associate, Calderwood & Harkness LLP
DATE: June 25, 2025
RE: Analysis of RouteSync Dispatch Records and GPS Telematics Data — Delano v. Stonefield Logistics Group Inc., Case No. 3:25-cv-01847-RBC (N.D. Tex.)
CC: Bernard Crawford, Responsible Partner; Lisa Cheng, Paralegal
I. Executive Summary
This memorandum presents a detailed analysis of RouteSync dispatch records and GPS telematics data for (a) the named plaintiff, Marcus Delano, during his tenure at the Fort Worth distribution hub (HUB-01) from March 2021 through September 2024, and (b) a representative sample of 50 drivers across five of SLG's fourteen distribution hubs during the relevant FLSA limitations period (April 30, 2022 through present). The data was extracted by SLG's Information Technology department under the supervision of General Counsel Margaret Hollis and Vice President of Operations Trevor Sandoval, and was compiled and organized with the assistance of paralegal Lisa Cheng. This memorandum is intended to serve as an internal analytical resource for the defense team and is protected by the attorney work product doctrine.
The primary objective of this analysis is to assess whether there is sufficient uniformity or material variation in hours-worked patterns, route deviation frequency, delivery refusal rates, and schedule variability to inform (i) the opposition to plaintiff's motion for conditional certification under 29 U.S.C. § 216(b), which is due August 13, 2025, and (ii) the broader defense that SLG drivers are not similarly situated and exercised meaningful discretion consistent with independent contractor status under the FLSA economic reality test. These two objectives are related but distinct: the conditional certification inquiry focuses on whether putative collective members are "similarly situated" such that their claims can be resolved on a collective basis, while the merits defense focuses on whether individual drivers were properly classified as independent contractors under the multi-factor economic reality framework applied by the Fifth Circuit.
The key conclusion of this analysis is that the data reveals substantial inter-driver and inter-hub variability across all measured dimensions, undermining plaintiff's assertion that SLG exercised uniform control over approximately 1,187 drivers. Average weekly hours varied by more than 11 hours between the highest- and lowest-averaging hubs in the sample, individual route deviation rates spanned from 8% to 61%, delivery acceptance rates ranged from 83% to 99%, and daily start times varied by nearly three hours across the 50-driver sample. These figures reflect a driver population exercising materially different levels of engagement, autonomy, and scheduling discretion — not the uniform, employer-controlled workforce described in the complaint.
The data also shows that many drivers exercised meaningful entrepreneurial discretion — deviating from suggested routes, refusing deliveries, and varying their schedules significantly — consistent with independent contractor status. Importantly, the named plaintiff, Marcus Delano, consistently falls at the high end of hours worked and the low end of route deviation and delivery refusal, making his individual experience an outlier relative to the broader driver population. His experience, while genuine, is not representative of the typical SLG driver and should not serve as the basis for certifying a collective action encompassing all 1,187 drivers across 14 hubs.
This analysis will support the forthcoming Declaration of Dr. Priya Ramanathan of Corbin Risk Advisors LLC, who will conduct a comprehensive statistical analysis of the full driver population, as well as the declarations of Trevor Sandoval (VP of Operations), Karen Lund (Director of Driver Relations), and Dennis Faulkner (Hub Manager, Fort Worth HUB-01). These declarations are currently in preparation and are expected to be finalized by the end of July 2025.
II. Data Sources and Methodology
A. RouteSync Dispatch Platform Data
RouteSync is SLG's proprietary dispatch platform, developed internally and deployed across all 14 distribution hubs beginning in 2018. The platform serves three primary functions: (1) assigning deliveries to drivers based on geographic proximity, delivery type, and driver availability; (2) providing suggested route sequencing — that is, an algorithmically optimized delivery order designed to minimize total drive time; and (3) tracking delivery completion through driver-initiated status updates (accepted, en route, delivered, refused by customer, and returned). RouteSync is the central data repository for all delivery-related activity across the SLG network and, as such, constitutes the most comprehensive source of operational data regarding individual driver engagement and performance.
The following data fields were extracted for purposes of this analysis:
• Driver ID: Unique alphanumeric identifier assigned to each driver upon execution of the Independent Contractor Operating Agreement ("ICOA").
• Hub Assignment: The distribution hub to which the driver is primarily assigned (drivers may occasionally receive cross-hub assignments, which are flagged in the data).
• Delivery Assignment Timestamps: Date and time at which each delivery was offered to the driver through the RouteSync platform.
• Delivery Acceptance/Decline Timestamps: Date and time at which the driver accepted or declined the offered delivery, along with the response code (accepted, declined — scheduling conflict, declined — distance, declined — other).
• Suggested Route Sequence: The algorithmically generated delivery order provided by RouteSync for each delivery day.
• Actual Delivery Sequence: The order in which the driver actually completed deliveries, as recorded by delivery-completion status updates.
• Delivery Completion Timestamps: Date and time at which each delivery was marked as completed, including customer signature confirmation where applicable.
• Delivery Type: Classification of each delivery as either "standard" (packages under 75 lbs., curbside or doorstep) or "white-glove" (packages over 75 lbs. or requiring in-home placement, assembly, or debris removal).
• Per-Delivery Rate: The compensation rate for each delivery, ranging from $14.50 to $38.00 depending on weight, distance from the hub, and delivery type (standard vs. white-glove).
• Customer Delivery Window: The time window within which the customer expected delivery, as communicated through the RouteSync platform.
• Performance Flags: System-generated flags for late deliveries, customer complaints, and acceptance rate warnings.
The data extraction period covered April 30, 2022 through June 15, 2025, encompassing approximately 161 weeks and capturing the full FLSA limitations period applicable to this action. For the named plaintiff Marcus Delano, additional data extending back to March 2021 was extracted to provide a complete picture of his engagement with SLG, although only the data from April 30, 2022 forward falls within the relevant limitations period.
Data extraction was performed by SLG's IT department at the direction of Margaret Hollis, General Counsel, and Trevor Sandoval, Vice President of Operations, pursuant to the litigation hold notice issued on May 2, 2025. The data was provided to Calderwood & Harkness in CSV format and was loaded into a secure analytical database maintained by the firm's litigation support team for processing and review.
B. GPS Telematics Data
GPS telematics data was collected via SLG-provided telematics devices that are mandatory on all driver vehicles pursuant to all versions of the ICOA (v1.0, dated 2018; v2.0, dated August 2020; and v3.0, dated January 2023). The telematics devices are installed in each driver's vehicle at the time of onboarding and remain active throughout the driver's engagement with SLG. The stated purpose of the telematics program, as described in the ICOA, is to support delivery tracking, customer communication, and safety monitoring — the devices are not characterized as employee time-tracking tools, a distinction that may be relevant to the control analysis.
The following data fields were extracted from the telematics system:
• Vehicle Location: Latitude and longitude coordinates recorded at 30-second intervals whenever the vehicle ignition is in the "on" position.
• Ignition On/Off Timestamps: The date and time at which the vehicle ignition was turned on and off, which serves as a proxy for shift start and end times.
• Speed: Vehicle speed at each 30-second data point, used in connection with route reconstruction.
• Idle Time: Periods during which the vehicle ignition is on but the vehicle is stationary for more than five consecutive minutes.
• Route Traveled vs. RouteSync Suggested Route: A comparison of the GPS-derived actual route against the RouteSync-suggested route, enabling calculation of geographic deviation.
GPS data was used to calculate the following derived metrics:
• Actual hours worked per day and per week: Calculated as the elapsed time between the first ignition-on event and the last ignition-off event for each day, less any gaps of 60 minutes or more between ignition-off and subsequent ignition-on events (which are treated as breaks or personal time, consistent with the methodology described in Dr. Ramanathan's work plan).
• Route deviation frequency and magnitude: Determined by comparing the geographic path taken by the driver against the RouteSync-suggested route, with deviations flagged when the actual delivery sequence differed from the suggested sequence by three or more positions.
• Time between deliveries: The elapsed time between delivery completion at one stop and arrival at the next stop, which may serve as an indicator of personal errands, rest breaks, or non-SLG work performed during the delivery day.
• Geographic coverage area: The total geographic footprint of each driver's deliveries, measured as the convex hull of all delivery locations for each week, providing insight into the scope and scale of each driver's operational territory.
C. Sample Selection
The analysis in this memorandum covers two distinct data sets:
Named Plaintiff Marcus Delano. The complete data set for Delano was extracted from both RouteSync and the GPS telematics system, covering his full tenure at HUB-01 from March 2021 through September 2024. Within the relevant FLSA limitations period (April 30, 2022 through September 2024), Delano's data encompasses approximately 126 weeks of activity. This complete data set allows for a thorough evaluation of Delano's individual experience and the extent to which it is representative of the broader driver population.
50-Driver Sample. A sample of 50 drivers was selected from five of SLG's 14 distribution hubs to capture geographic and operational diversity while keeping the preliminary analysis manageable in scope. The five hubs were selected to represent a cross-section of SLG's operations, including its two largest Texas hubs, its largest out-of-state hub, and two smaller hubs with distinct operational profiles:
The selection methodology employed stratified random sampling within each hub to ensure representation across three key dimensions: (1) ICOA version — drivers operating under ICOA v1.0 (2018), v2.0 (August 2020), and v3.0 (January 2023) were proportionally represented; (2) tenure length — the sample includes drivers with tenure ranging from less than one year to more than five years; and (3) delivery type — drivers performing predominantly standard deliveries, predominantly white-glove deliveries, and a mix of both were included. Lisa Cheng assisted in compiling and organizing the data sets, cross-referencing driver IDs across the RouteSync and GPS databases, and verifying the completeness of the extracted records.
It should be noted that Dr. Priya Ramanathan of Corbin Risk Advisors LLC will conduct a broader statistical analysis of the full 1,187-driver population across all 14 hubs. This memorandum provides a preliminary assessment based on the 50-driver sample to guide briefing strategy for the conditional certification opposition. The findings presented here are intended to identify patterns and trends that can be confirmed or refined through Dr. Ramanathan's comprehensive analysis.
D. Analytical Framework
Four primary metrics were analyzed, each selected for its relevance to the central questions in this litigation:
1. Hours-Worked Patterns: Average weekly hours, variability in weekly hours, and distribution of weekly hours across the driver sample. This metric bears on both the similarly-situated inquiry (whether drivers worked comparable hours) and the economic reality test (whether SLG dictated the amount of time drivers devoted to SLG work).
2. Route Deviation Frequency: The percentage of delivery days on which a driver deviated from the RouteSync-suggested delivery sequence by three or more positions. This metric bears directly on the degree-of-control factor, as it measures the extent to which drivers exercised independent judgment in organizing their work rather than following SLG's suggested routing.
3. Delivery Refusal/Acceptance Rates: Each driver's rolling 30-day acceptance rate relative to the 85% minimum acceptance rate threshold in the ICOA. This metric addresses both the degree of control (whether drivers were compelled to accept all assignments) and the opportunity for profit or loss (whether drivers could exercise economic judgment in selecting higher-value deliveries).
4. Schedule Variability: Analysis of daily start times, end times, days worked per week, Saturday participation rates, and multi-day breaks. This metric speaks to whether SLG controlled when drivers worked or whether drivers exercised genuine scheduling discretion.
These four metrics map to key factors in the FLSA economic reality test as applied by the Fifth Circuit, particularly the degree of control exercised by the alleged employer, the worker's opportunity for profit or loss based on managerial skill, and the permanence of the working relationship. See Hopkins v. Cornerstone Am., 545 F.3d 338, 343 (5th Cir. 2008). The metrics also directly address the threshold question of whether the putative collective members are "similarly situated" for purposes of conditional certification under 29 U.S.C. § 216(b), which requires a showing that the proposed collective members are victims of a common policy or plan that violated the FLSA. See Swales v. KLLM Transp. Servs., L.L.C., 985 F.3d 430, 441–43 (5th Cir. 2021).
III. Analysis of Named Plaintiff Marcus Delano
A. Hours-Worked Patterns
Marcus Delano operated out of HUB-01, SLG's Fort Worth distribution hub located at 1015 Railhead Parkway, Fort Worth, TX 76106, from March 2021 through September 2024. Delano executed ICOA v2.0 upon commencement of his engagement in March 2021, and subsequently executed ICOA v3.0 in January 2023 when the updated agreement was rolled out across all hubs.
During the relevant FLSA limitations period (April 30, 2022 through September 2024, encompassing approximately 126 weeks), Delano's hours-worked patterns, as derived from GPS ignition on/off data, are summarized as follows:
• Average weekly hours: 58 hours. This figure is consistent with the allegations in plaintiff's complaint, which states that Delano "regularly worked 55–65 hours per week."
• Weekly hours range: A minimum of 38 hours (recorded during the week of July 4, 2023, a holiday week during which delivery volume was reduced) to a maximum of 72 hours (recorded during the week of December 18, 2023, during peak holiday delivery season when SLG experiences its highest annual demand).
• Standard deviation of weekly hours: Approximately 7.2 hours, indicating meaningful week-to-week variability in Delano's time commitment.
• Weeks exceeding 60 hours: 41 of 126 weeks (32.5%).
• Weeks below 50 hours: 18 of 126 weeks (14.3%).
• Weeks between 50 and 60 hours: 67 of 126 weeks (53.2%).
Delano's average weekly gross pay during calendar year 2023 was $1,247, yielding an effective hourly rate of approximately $21.50 per hour ($1,247 ÷ 58 hours). This effective rate exceeds the federal minimum wage of $7.25 per hour and the Texas minimum wage of $7.25 per hour, but plaintiff's claim is premised on the contention that he was an employee entitled to overtime compensation under the FLSA — i.e., that hours in excess of 40 per week should have been compensated at 1.5 times his regular rate. Under plaintiff's theory, his effective hourly rate would be lower when the overtime premium is factored into the calculation.
It is significant that Delano's average of 58 weekly hours is above the sample median of 47.8 hours for the 50-driver sample analyzed in Section IV below, and well above the sample mean of 49.3 hours. This places Delano in the top quartile of the sample distribution. His experience of consistently working long weeks, while genuinely reflected in the data, is not representative of the typical Fort Worth driver — let alone the typical SLG driver across all hubs. The week-to-week variability (standard deviation of 7.2 hours, range of 38 to 72 hours) further indicates that even Delano's own schedule was not rigidly controlled by SLG, but rather fluctuated based on delivery volume, personal choices, and seasonal demand.
A complete week-by-week breakdown of Delano's hours, delivery counts, gross pay, and effective hourly rate is set forth in Appendix A.
B. Route Deviation Frequency
The RouteSync platform provides each driver with a "suggested route" — an algorithmically optimized delivery sequence designed to minimize total drive time based on delivery locations, customer delivery windows, and real-time traffic data. Drivers are not contractually required to follow the exact sequence provided by RouteSync. The ICOA (all versions) states that RouteSync provides "suggested delivery sequencing for the convenience of the Contractor" and that drivers "may complete deliveries in any order consistent with customer delivery windows." This contractual language is significant because it frames RouteSync as an advisory tool rather than a mandatory directive.
Delano's route deviation rate during the relevant period was approximately 22% of delivery days — meaning that on roughly one out of every five delivery days, Delano completed his deliveries in an order that differed from the RouteSync-suggested sequence by three or more positions. For purposes of this analysis, a "deviation" is defined as completing deliveries in an order that differs from the RouteSync-suggested sequence by three or more positions. This threshold was chosen to exclude minor, inconsequential reorderings (e.g., swapping two adjacent stops) and to capture only meaningful exercises of independent judgment about delivery sequencing.
Delano's 22% deviation rate is below the overall sample average of 31.4% (see Section IV.B) and below the HUB-01 average of 27.8%, suggesting that Delano chose to follow RouteSync suggestions more closely than many of his peers at the same hub. This is a notable finding because plaintiff's complaint alleges that SLG "dictated daily delivery routes via RouteSync with limited driver discretion." If RouteSync truly dictated routes, one would expect deviation rates to be uniformly low across all drivers. Instead, the data shows a wide range of deviation behavior, with Delano falling at the more-compliant end of the spectrum — a result of his individual choices rather than SLG-imposed constraints.
Further analysis of GPS data on deviation days reveals that Delano's actual routes on those days were often geographically efficient alternatives to the RouteSync suggestion. For example, on 14 of the deviation days examined in detail, Delano appears to have resequenced deliveries to account for known traffic patterns on Interstate 35W and State Highway 121 that the RouteSync algorithm may not have adequately weighted. This pattern suggests that Delano exercised independent judgment about optimal sequencing on those days — precisely the kind of entrepreneurial discretion that is characteristic of independent contractor status.
C. Delivery Refusal/Acceptance Rate
Under all versions of the ICOA, drivers are subject to a minimum acceptance rate of 85% per rolling 30-day period. A driver whose acceptance rate falls below 85% in any rolling 30-day period is subject to a "performance review," which may include counseling, a formal warning, or, in extreme cases, termination of the ICOA. The 85% threshold is a contractual minimum, not a requirement to accept every offered delivery — it leaves substantial room for drivers to decline up to 15% of offered deliveries in any given 30-day period without consequence.
Delano's delivery acceptance patterns during the relevant period are as follows:
• Average rolling 30-day acceptance rate: 94.1%.
• Acceptance rate range: 87% (lowest rolling 30-day period) to 100% (achieved during multiple rolling periods).
• Total deliveries offered: Approximately 7,840 during the relevant period.
• Total deliveries declined: Approximately 463 (5.9% of offered deliveries).
Delano never fell below the 85% threshold during any rolling 30-day period and thus never triggered a performance review for low acceptance. This fact is relevant but cuts both ways: while it confirms that Delano consistently accepted a high proportion of offered deliveries, it also shows that he maintained a buffer above the 85% threshold, indicating that he had room to decline additional deliveries had he chosen to do so.
Plaintiff's complaint alleges that Delano "had no ability to decline deliveries." The data directly contradicts this allegation. Delano declined approximately 5.9% of offered deliveries across the relevant period — a meaningful number representing nearly 463 individual delivery refusals. While Delano may characterize the 85% minimum as effectively compulsory acceptance, the data shows that he exercised the right to decline deliveries on numerous occasions and faced no adverse consequences for doing so. Moreover, as discussed in Section IV.C, other drivers in the sample exercised this right even more aggressively, with acceptance rates as low as 83%.
D. Schedule Variability
The ICOA establishes a daily delivery window of 7:00 AM to 8:00 PM, Monday through Saturday, during which drivers may perform deliveries. This window defines the outer boundaries of when deliveries can be made to customers, but it does not require drivers to work during the entirety of this window or to work on every available day. Drivers are free to choose their start times, end times, and which days they work, subject to the minimum acceptance rate threshold discussed above.
Delano's GPS ignition data reveals the following scheduling patterns during the relevant period:
• Average start time: 7:12 AM. Delano consistently began his delivery days shortly after the opening of the delivery window, suggesting a preference for early starts.
• Average end time (Monday–Friday): 5:48 PM. Delano typically concluded his weekday deliveries well before the 8:00 PM close of the delivery window.
• Average end time (Saturdays): 3:15 PM. On Saturdays, Delano worked shorter days on average, finishing more than four and a half hours before the close of the delivery window.
• Saturday participation rate: Approximately 78% of available Saturdays. Delano chose not to work on approximately 22% of available Saturdays, indicating some degree of scheduling discretion.
• Multi-day breaks: Delano took multi-day breaks (defined as three or more consecutive days off) on 7 occasions during the relevant period. None of these absences required formal approval from SLG per the records reviewed, and no disciplinary action was taken in connection with any of them.
The scheduling data reveals that while Delano worked long hours on average, he exercised meaningful discretion in determining when and how much he worked. His 22% Saturday non-participation rate and 7 multi-day breaks demonstrate that he was not required to work a fixed schedule. The 5:48 PM average weekday end time — more than two hours before the close of the delivery window — suggests that Delano chose to conclude his days at a time of his choosing rather than working until the delivery window closed.
E. Subcontracting Activity
Cross-referencing with the subcontracting analysis currently being prepared in a separate memorandum (anticipated completion date: June 30, 2025), Delano made zero subcontracting requests during his entire tenure with SLG. This is consistent with his allegation in the complaint that subcontracting was "effectively prohibited" and that the contractual right to subcontract was illusory.
However, this is an individual fact specific to Delano's experience. Other drivers in the 50-driver sample did submit subcontracting requests through the RouteSync platform, and some of those requests were approved. The subcontracting analysis memorandum will address this issue in greater detail, including the frequency of subcontracting requests, approval rates, and any differences across hubs and ICOA versions. For present purposes, it is sufficient to note that Delano's failure to utilize the subcontracting provision does not establish that the provision was unavailable to the broader driver population — it may simply reflect Delano's individual preference not to engage subcontractors.
IV. Analysis of 50-Driver Sample Across Five Hubs
A. Hours-Worked Patterns — Inter-Driver and Inter-Hub Variation
The hours-worked analysis of the 50-driver sample reveals substantial variation both across hubs and among individual drivers within the same hub. The overall sample statistics for the relevant FLSA limitations period are as follows:
• Mean weekly hours: 49.3 hours.
• Median weekly hours: 47.8 hours.
• Standard deviation: 11.6 hours.
• Range: 28 hours (lowest individual driver average, recorded by a driver at HUB-09 New Orleans) to 63 hours (highest individual driver average, recorded by a driver at HUB-03 Houston).
The hub-level breakdown of average weekly hours is summarized in the following table:
Several findings merit emphasis. First, there is significant inter-hub variation in average weekly hours, with an 11.2-hour spread between the highest hub average (Houston at 53.8 hours) and the lowest hub average (New Orleans at 42.6 hours). This spread represents more than a full standard deviation and demonstrates that the time commitment associated with being an SLG driver varied materially depending on geographic location and local market conditions. Second, there is substantial intra-hub variation, with standard deviations ranging from 7.9 hours (Oklahoma City) to 12.4 hours (New Orleans). The New Orleans hub, in particular, exhibited the widest dispersion of weekly hours, with individual driver averages spanning from 28 to 60 hours per week — a 32-hour gap within a single hub.
Third, Delano's average of 58 hours per week is well above the 50-driver sample mean of 49.3 hours, placing him in approximately the 85th percentile of the sample distribution. This finding is critical to the conditional certification analysis because it demonstrates that plaintiff's experience of working long weeks is not representative of the broader driver population. A collective action premised on the assumption that SLG drivers uniformly worked 55–65 hours per week would be inconsistent with the data, which shows that a substantial number of drivers — particularly at the Oklahoma City and New Orleans hubs — worked well under 50 hours per week on average.
The implication for the conditional certification opposition is clear: plaintiff's allegation that SLG drivers "regularly worked 55–65 hours per week" describes Delano's individual experience but does not accurately characterize the driver population as a whole. The material variation in weekly hours across and within hubs precludes a finding that drivers were "similarly situated" with respect to the fundamental question of how many hours they worked.
Complete driver-level summary statistics are set forth in Appendix B.
B. Route Deviation Frequency — Inter-Driver and Inter-Hub Variation
The overall sample route deviation rate was 31.4% of delivery days — meaning that, on average, drivers deviated from the RouteSync-suggested delivery sequence on nearly one out of every three delivery days. This figure represents a meaningful exercise of discretion across the driver population, as it indicates that RouteSync suggestions were treated as advisory rather than mandatory by a significant portion of drivers on a significant portion of their working days.
Hub-level deviation rates are as follows:
The hub-level data shows a 10.3-percentage-point spread between the lowest deviation rate (Fort Worth at 27.8%) and the highest (Oklahoma City at 38.1%). This spread reflects the influence of local geography and operational conditions: Oklahoma City drivers, who tend to cover longer distances between delivery stops in a more dispersed metropolitan area, apparently found it more frequently necessary or advantageous to resequence their deliveries.
More striking than the hub-level variation, however, is the individual driver variation. Individual driver deviation rates within the 50-driver sample ranged from 8% (one Fort Worth driver who closely followed RouteSync suggestions on nearly all delivery days) to 61% (one Oklahoma City driver who routinely resequenced deliveries on a majority of delivery days). This 53-percentage-point range is among the most powerful data points in this analysis. It demonstrates that SLG did not enforce rigid route compliance and that drivers exercised widely varying degrees of discretion in organizing their delivery sequences. If SLG were truly "dictating" routes as plaintiff alleges, one would expect route adherence rates to be uniformly high across all drivers. Instead, the data reveals a distribution consistent with genuine driver choice: some drivers preferred to follow RouteSync suggestions, others routinely deviated, and most fell somewhere in between.
Delano's 22% deviation rate is below both the overall sample mean of 31.4% and the HUB-01 mean of 27.8%, making him an outlier in the direction of greater adherence to RouteSync suggestions — that is, less autonomy. His individual experience of relatively high route compliance does not reflect the typical driver's exercise of discretion. To the contrary, the data shows that the majority of drivers deviated from suggested routes more frequently than Delano did, undermining his claim that RouteSync functioned as a mandatory routing system rather than a suggestion tool.
Driver-level deviation rates are set forth in Appendix C.
C. Delivery Refusal/Acceptance Rates — Inter-Driver and Inter-Hub Variation
The delivery acceptance rate analysis examines drivers' rolling 30-day acceptance rates relative to the 85% minimum threshold established in the ICOA. The overall sample statistics are as follows:
• Overall sample average rolling 30-day acceptance rate: 91.2%.
• Individual driver acceptance rate range: 83% (one Houston driver) to 99% (one Dallas driver).
Hub-level average acceptance rates are summarized below:
The hub-level data shows a 4.4-percentage-point spread between the highest average acceptance rate (Fort Worth at 93.0%) and the lowest (New Orleans at 88.6%). While this spread is narrower than the variation observed in hours worked or route deviation, it nonetheless reflects meaningful differences in driver behavior across hubs. The New Orleans and Houston hubs, in particular, showed lower average acceptance rates, suggesting that drivers at those locations exercised greater selectivity in choosing which deliveries to accept.
The individual driver range of 83% to 99% is more revealing. Seven of the 50 sampled drivers (14%) fell below the 85% threshold in at least one rolling 30-day period during the relevant period. The outcomes for these seven drivers varied considerably:
• Four drivers received informal counseling from their hub manager — typically a phone call or brief in-person conversation encouraging the driver to increase acceptance rates. No formal documentation was generated for these interactions.
• Two drivers had no documented follow-up of any kind, despite falling below the 85% threshold. This suggests that enforcement of the threshold was inconsistent and that some hub managers treated the 85% minimum as a soft guideline rather than a rigid requirement.
• One driver received a formal written warning, which was placed in the driver's file. This driver's acceptance rate had fallen to 83% during the relevant rolling period, and the driver had previously received informal counseling on two occasions.
The inconsistency in enforcement outcomes is itself a significant finding. If SLG exercised the kind of uniform, employer-like control alleged in the complaint, one would expect consistent enforcement of the 85% threshold across all hubs. Instead, the data shows that hub managers had substantial discretion in how — and whether — they responded to below-threshold acceptance rates. This inconsistency is inconsistent with an employment relationship and more consistent with the kind of arm's-length commercial relationship that characterizes independent contractor engagements.
The acceptance rate data also demonstrates that drivers had a genuine ability to decline deliveries. Across the 50-driver sample, the average driver declined approximately 8.8% of offered deliveries (100% minus the 91.2% average acceptance rate). Some drivers declined up to 17% of offered deliveries (the Houston driver with the 83% acceptance rate) without termination of their ICOA. This level of delivery selectivity is inconsistent with plaintiff's claim that drivers "had no ability to decline deliveries" and supports the defense that drivers exercised meaningful economic discretion in choosing which deliveries to accept — for example, by favoring higher-rate white-glove deliveries over lower-rate standard deliveries.
Driver-level acceptance rate data is set forth in Appendix D.
D. Schedule Variability — Inter-Driver and Inter-Hub Variation
The schedule variability analysis examines four dimensions of driver scheduling: daily start times, daily end times, average days worked per week, and Saturday participation rates. The data reveals wide variation across all four dimensions, demonstrating that SLG did not impose a uniform work schedule on its drivers.
Start Times. Average daily start times across the 50-driver sample ranged from 6:45 AM to 9:30 AM — a span of nearly three hours. Some drivers consistently began their days before the 7:00 AM opening of the delivery window (GPS ignition data shows vehicle movement as early as 6:30 AM in some cases, likely reflecting pre-delivery vehicle inspections and drive time to the first delivery location), while others did not begin deliveries until well into the mid-morning. The variation in start times was particularly pronounced at HUB-09 (New Orleans), where individual driver average start times ranged from 6:50 AM to 9:30 AM — a spread of 2 hours and 40 minutes within a single hub.
End Times. Average daily end times ranged from 2:30 PM to 7:15 PM — a span of nearly five hours. This range is especially significant because it shows that drivers exercised substantial discretion in determining when to conclude their delivery days. A driver who routinely finished at 2:30 PM was working a fundamentally different schedule than a driver who routinely worked until 7:15 PM, despite both operating under the same ICOA delivery window of 7:00 AM to 8:00 PM. No driver in the sample averaged an end time at or near the 8:00 PM close of the delivery window, indicating that the window functioned as an outer boundary for customer delivery purposes rather than as a mandated work schedule.
Days Worked Per Week. Average days worked per week ranged from 4.2 days per week (one New Orleans driver) to 6.0 days per week (several Houston and Fort Worth drivers who worked virtually every Saturday). The overall sample average was approximately 5.1 days per week, with a standard deviation of 0.7 days. The New Orleans hub showed the greatest variability in days per week, consistent with the broader pattern of schedule flexibility at that location.
Saturday Participation Rates. Saturday participation rates ranged from 42% at the New Orleans hub (sample average) to 81% at the Fort Worth hub (sample average), with significant intra-hub variation as well. Some individual drivers worked nearly every available Saturday, while others worked fewer than half. The variation in Saturday participation is particularly relevant because it demonstrates that Saturday work was genuinely optional: drivers who chose to work fewer Saturdays did so without documented consequence.
Multi-Day Breaks. Multi-day breaks, defined as three or more consecutive days off without formal approval, were observed for 38 of the 50 sampled drivers (76%). The number of such breaks per driver during the relevant period ranged from 2 to 14 instances. No evidence of SLG disciplinary action was found in connection with any of these multi-day absences. Some multi-day breaks coincided with holidays or seasonal demand lulls, but many occurred during ordinary business periods, suggesting that drivers were exercising genuine scheduling autonomy.
The schedule variability data is among the strongest evidence supporting the defense position. The wide variability in start times (nearly 3-hour range), end times (nearly 5-hour range), days per week (4.2 to 6.0), and Saturday participation (42% to 81% across hubs, with even wider individual variation) demonstrates that SLG did not impose a uniform work schedule on its drivers. The 7:00 AM to 8:00 PM delivery window established in the ICOA functioned as an outer boundary for customer-facing delivery activities, not as a mandatory work schedule. Drivers exercised substantial discretion in choosing when to start, when to stop, which days to work, and when to take extended breaks — all hallmarks of independent contractor status.
Driver-level schedule variability data is set forth in Appendix E.
E. Hub-Specific Operational Differences Observed in Data
Beyond the aggregate statistical findings, the data reveals qualitative operational differences across the five sampled hubs that further undermine the premise of a uniform, one-size-fits-all employment relationship.
Houston (HUB-03) showed a notably higher proportion of white-glove deliveries, accounting for approximately 28% of total deliveries compared to the system-wide average of approximately 12%. White-glove deliveries command higher per-delivery rates (typically $28.00 to $38.00 versus $14.50 to $22.00 for standard deliveries) but require significantly more time per delivery due to in-home placement, assembly, and debris removal requirements. As a result, Houston drivers showed higher average weekly hours but also higher average gross weekly pay relative to drivers at other hubs. The concentration of white-glove work at the Houston hub creates a fundamentally different economic profile for Houston drivers — one that cannot be meaningfully compared to drivers at hubs with predominantly standard delivery operations.
Oklahoma City (HUB-06) featured a higher proportion of regional freight deliveries, with longer average drive distances between delivery stops. The dispersed geography of the Oklahoma City metropolitan area and surrounding communities resulted in correspondingly higher route deviation rates, as drivers exercised greater independent judgment in routing decisions necessitated by the geographic spread of their delivery territories. Oklahoma City drivers also showed a higher incidence of time gaps between deliveries exceeding 30 minutes, which may reflect both the longer drive distances and the possibility of non-SLG commercial activity during the delivery day.
New Orleans (HUB-09) exhibited the greatest schedule variability among the five sampled hubs, with the lowest average weekly hours (42.6), the lowest Saturday participation rate (42%), and the widest dispersion of individual driver averages across all metrics. This pattern may reflect seasonal demand fluctuations inherent to the New Orleans market, as well as a driver population with more variable engagement levels — some drivers appear to have treated SLG work as a primary income source, while others appear to have treated it as supplemental or part-time work. The New Orleans data is particularly powerful evidence against conditional certification because it demonstrates that even within a single hub, drivers' relationships with SLG varied enormously.
Fort Worth (HUB-01) and Dallas (HUB-02) were the most similar to each other among the five sampled hubs, which is unsurprising given their geographic proximity within the Dallas-Fort Worth metropolitan area. However, even these two closely situated hubs showed statistically meaningful differences in route deviation rates (27.8% vs. 29.5%), acceptance rates (93.0% vs. 91.8%), and Saturday participation rates (81% vs. 73%). If two hubs located approximately 35 miles apart in the same metropolitan area show meaningful operational differences, the variation across SLG's full 14-hub network — spanning Texas, Oklahoma, Louisiana, and Arkansas — is likely to be considerably greater.
These hub-level differences support the argument that a single collective action definition spanning all 14 SLG hubs is overbroad. The operational realities experienced by a white-glove delivery driver in Houston are materially different from those experienced by a regional freight driver in Oklahoma City or a part-time standard delivery driver in New Orleans. A collective action that lumps all of these drivers together as "similarly situated" would obscure the very differences that are central to the classification analysis.
V. Comparison to Plaintiff's Allegations
A. Allegations vs. Data — Named Plaintiff
The following chart compares plaintiff's key allegations in the complaint to the data findings for Marcus Delano:
Allegation: Delano "regularly worked 55–65 hours per week." Data Finding: Delano's average weekly hours of 58 falls within the alleged range, confirming the general accuracy of this allegation as to the average. However, the data reveals a wide weekly range of 38 to 72 hours, with a standard deviation of 7.2 hours. Delano worked below 50 hours in 18 of 126 weeks (14.3%) and above 60 hours in 41 of 126 weeks (32.5%). This variability is inconsistent with a rigid, employer-controlled schedule and instead reflects a work pattern influenced by seasonal demand, personal choice, and other variable factors. The allegation, while not false as to the average, paints an incomplete picture by omitting the substantial week-to-week fluctuation.
Allegation: SLG "dictated daily delivery routes via RouteSync with limited driver discretion." Data Finding: Delano deviated from RouteSync-suggested routes on 22% of delivery days during the relevant period. The broader 50-driver sample deviated at a 31.4% rate. These figures demonstrate that RouteSync functioned as a suggestion tool, not a mandate. The ICOA itself characterizes RouteSync routing as "suggested delivery sequencing," and the data confirms that drivers — including Delano himself — frequently exercised independent judgment in organizing their delivery sequences. Notably, GPS analysis of Delano's deviation days shows that his alternative routes were often geographically efficient, reflecting the exercise of professional judgment about optimal routing rather than mere noncompliance.
Allegation: SLG "prohibited subcontracting." Data Finding: Delano made zero subcontracting requests during his tenure. While this is consistent with his subjective belief that subcontracting was prohibited, it does not establish an objective prohibition. Other drivers in the sample did submit subcontracting requests through RouteSync, and some of those requests were approved. The absence of subcontracting activity by Delano reflects his individual choice, not a company-wide prohibition. This issue will be addressed in greater detail in the separate subcontracting analysis memorandum anticipated for June 30, 2025.
Allegation: Drivers "had no ability to negotiate per-delivery rates." Data Finding: This allegation is confirmed by the data insofar as per-delivery rates were set by SLG on a standardized basis ranging from $14.50 to $38.00, depending on weight, distance from the hub, and delivery type. Drivers did not negotiate individual rates for specific deliveries. However, this finding must be placed in context. While rate-setting by the alleged employer is one factor in the economic reality test, it is not dispositive. More importantly, the data shows that drivers exercised economic discretion by choosing which deliveries to accept and which to decline. Drivers who strategically favored higher-rate white-glove deliveries over lower-rate standard deliveries could meaningfully influence their earnings — an exercise of entrepreneurial judgment that is consistent with independent contractor status, even in the absence of rate negotiation.
B. Allegations vs. Data — Similarly Situated Question
Plaintiff seeks to represent a collective of approximately 1,187 drivers across all 14 SLG distribution hubs from April 30, 2022 to present, alleging that all such drivers were "similarly situated" in that they were misclassified as independent contractors and subjected to uniform policies and practices that deprived them of overtime compensation. The data from the 50-driver, 5-hub sample seriously undermines this allegation.
Even within this limited sample, the data reveals the following variation:
• An 11.2-hour gap in average weekly hours between the highest hub average (Houston at 53.8 hours) and the lowest hub average (New Orleans at 42.6 hours), with individual driver averages spanning a 35-hour range (28 to 63 hours).
• A 53-percentage-point range in individual route deviation rates (8% to 61%), demonstrating that drivers exercised vastly different levels of routing autonomy under the same contractual framework.
• A 16-percentage-point range in individual acceptance rates (83% to 99%), with inconsistent enforcement of the 85% minimum threshold across hubs.
• A near 3-hour range in average daily start times (6:45 AM to 9:30 AM) and a nearly 5-hour range in average daily end times (2:30 PM to 7:15 PM), demonstrating that drivers set fundamentally different daily schedules.
• Saturday participation rates ranging from 42% (New Orleans sample average) to 81% (Fort Worth sample average), with even wider individual variation.
These variations are not trivial — they go to the heart of the classification inquiry. The degree of control SLG exercised over each driver, the amount of time each driver devoted to SLG work, the extent to which each driver exercised entrepreneurial discretion, and the economic profile of each driver's engagement all varied materially across the sample. A collective action that seeks to resolve these intensely individualized questions on a group-wide basis is inconsistent with the data.
Moreover, these variations are likely to be amplified when Dr. Ramanathan's full statistical analysis covers all 1,187 drivers across all 14 hubs — including the nine hubs not represented in this preliminary sample (San Antonio, Austin, Tulsa, Baton Rouge, Shreveport, Little Rock, El Paso, Lubbock, and Corpus Christi). Each of these additional hubs may present its own unique operational characteristics, delivery mix, geographic profile, and driver engagement patterns, further compounding the inter-hub variability documented here.
VI. Implications for Conditional Certification Opposition
A. Supporting "Not Similarly Situated" Argument
The data analysis presented in this memorandum provides substantial evidentiary support for the argument that SLG's drivers are not similarly situated for purposes of FLSA conditional certification under 29 U.S.C. § 216(b). In Swales v. KLLM Transport Services, L.L.C., 985 F.3d 430, 441–43 (5th Cir. 2021), the Fifth Circuit rejected the lenient, two-step Lusardi approach to conditional certification and instead directed district courts to rigorously analyze whether putative collective members are "similarly situated" at the outset — including by considering the merits of the case to the extent they bear on the similarly-situated inquiry. Under the Swales framework, the court should consider "all available evidence" to determine whether the proposed collective members are similarly situated, including evidence of individualized defenses and factual variations among putative members.
The data supports the "not similarly situated" argument on the following grounds:
• Material variation in hours worked: Average weekly hours varied by more than 11 hours across hubs and by 35 hours across individual drivers. Drivers' time commitments were not uniform, and the factors influencing each driver's hours — seasonal demand, personal preference, family obligations, engagement with other delivery platforms, and local market conditions — are inherently individualized.
• Material variation in route adherence: Route deviation rates varied by more than 53 percentage points across individual drivers, demonstrating that the degree of autonomy each driver exercised in organizing deliveries was a function of individual choice, not uniform company policy. Resolving the "degree of control" question for the collective as a whole would be impossible without examining each driver's actual route behavior.
• Material variation in delivery acceptance: Acceptance rates varied by 16 percentage points across individual drivers, and enforcement of the 85% threshold was inconsistent across hubs. Whether any given driver was "compelled" to accept deliveries is an individualized inquiry that cannot be resolved on a collective basis.
• Material variation in scheduling: Start times, end times, days worked per week, and Saturday participation all varied widely, demonstrating that drivers set their own schedules within the broad parameters of the ICOA delivery window. Whether SLG exercised scheduling control over any given driver requires individualized analysis.
• Hub-level operational differences: Delivery mix (standard vs. white-glove), geographic scope, demand patterns, and management practices varied across hubs, creating materially different working conditions at each location. A collective action spanning all 14 hubs would encompass drivers subject to fundamentally different operational realities.
• ICOA version differences: The three ICOA versions (v1.0, 2018; v2.0, August 2020; v3.0, January 2023) imposed different contractual frameworks, and the data reflects that operational practices likewise varied over time. Drivers operating under different ICOA versions were subject to different contractual terms, further undermining the premise of uniformity.
B. Supporting Independent Contractor Defense (Economic Reality Factors)
Beyond the conditional certification inquiry, the data also provides meaningful support for the merits defense that SLG's drivers were properly classified as independent contractors under the FLSA economic reality test. The Fifth Circuit applies a multi-factor test that considers, among other things, the degree of control exercised by the alleged employer, the worker's opportunity for profit or loss, the worker's investment in equipment and materials, the skill required, the permanence of the relationship, and the extent to which the work performed is an integral part of the alleged employer's business. See Hopkins v. Cornerstone Am., 545 F.3d 338, 343 (5th Cir. 2008).
Degree of control. The route deviation data and schedule variability data indicate that SLG did not exercise the degree of control typical of an employment relationship. Drivers deviated from RouteSync-suggested routes on an average of 31.4% of delivery days, with individual deviation rates ranging from 8% to 61%. Start times spanned a nearly 3-hour range, end times spanned a nearly 5-hour range, and days worked per week ranged from 4.2 to 6.0. These patterns are inconsistent with the kind of pervasive day-to-day control that characterizes an employment relationship and are more consistent with the arm's-length oversight exercised by a company over its independent contractors. RouteSync provided suggestions — not mandates — and drivers exercised meaningful discretion in how, when, and in what order they performed their work.
Opportunity for profit or loss. The delivery refusal data and schedule variability data show that drivers could and did make economic choices that affected their earnings. Drivers who accepted a higher proportion of white-glove deliveries at premium rates earned more per delivery than those who accepted primarily standard deliveries. Drivers who worked more days per week and more hours per day earned more than those who chose lighter schedules. Drivers who declined lower-rate deliveries in favor of waiting for higher-rate assignments exercised the kind of entrepreneurial judgment that is characteristic of independent contractor status. The 83%-to-99% range in acceptance rates demonstrates that drivers had genuine economic discretion — they were not passive recipients of assignments but active economic decision-makers.
These data-driven findings should be woven into the forthcoming declarations of Trevor Sandoval (VP of Operations), Karen Lund (Director of Driver Relations), and Dennis Faulkner (Hub Manager, Fort Worth HUB-01). Each declaration should reference specific data points from this analysis to support the declarant's testimony about the operational realities of the SLG driver relationship. The findings will also be statistically formalized in Dr. Ramanathan's expert declaration, which will provide the rigorous quantitative analysis necessary to withstand a Daubert challenge.
C. Arbitration Clause Intersection
An additional consideration bearing on the conditional certification analysis is the arbitration clause and collective action waiver contained in ICOA v3.0 (January 2023). Approximately 640 of the 1,187 drivers in the proposed collective signed ICOA v3.0, which includes a mandatory arbitration clause requiring that all disputes arising out of or relating to the ICOA be resolved through binding individual arbitration, with an express waiver of the right to participate in collective or class actions. SLG's motion to compel individual arbitration for these 640 drivers is currently pending before Judge Caldwell.
The data sample analyzed in this memorandum includes drivers operating under all three ICOA versions — v1.0, v2.0, and v3.0. If the motion to compel arbitration is granted for the approximately 640 v3.0 signatories, the remaining potential collective would comprise approximately 547 drivers operating under ICOA v1.0 and v2.0. Importantly, the inter-hub and inter-driver variability documented in this memorandum would still apply to this reduced collective. Even among the v1.0 and v2.0 driver populations, the data shows substantial variation in hours worked, route deviation, acceptance rates, and scheduling — the same fundamental obstacles to a "similarly situated" finding would persist regardless of the arbitration outcome.
Conversely, if the motion to compel arbitration is denied (or if only some v3.0 drivers are compelled to arbitrate based on challenges to the formation or enforceability of the arbitration clause), the full variability findings of this memorandum will be applicable to the broader collective of approximately 1,187 drivers. Either way, the data supports the defense position.
VII. Data Limitations and Caveats
This analysis is subject to several limitations that should be kept in mind when relying on its findings:
1. Sample scope. This analysis covers 50 drivers across 5 of SLG's 14 hubs. The remaining 9 hubs — San Antonio, Austin, Tulsa, Baton Rouge, Shreveport, Little Rock, El Paso, Lubbock, and Corpus Christi — are not included in this preliminary review. These hubs collectively account for approximately 524 of the 1,187 drivers in the proposed collective and may exhibit operational characteristics, delivery mixes, geographic profiles, and driver engagement patterns that differ from the five sampled hubs. Dr. Ramanathan's full statistical analysis will cover the complete 14-hub, 1,187-driver population and will provide a comprehensive picture of inter-hub and inter-driver variability across the entire SLG network.
2. GPS data as a proxy for hours worked. GPS ignition on/off timestamps are used as a proxy for hours worked, but they may not perfectly capture all work time. Specifically, pre-trip vehicle inspections, post-delivery administrative tasks (such as completing delivery logs or responding to customer inquiries), and other work performed without the vehicle ignition in the "on" position would not be reflected in the GPS data. To the extent that drivers performed meaningful unpaid work with their vehicles off, the GPS-derived hours figures may understate actual hours worked. Conversely, if drivers left their vehicles running during extended personal breaks or errands, the GPS data may overstate hours worked. The methodology described in Section II.B (excluding gaps of 60 minutes or more between ignition-off and ignition-on events) is designed to mitigate the latter issue but may not fully eliminate it.
3. Limitations of RouteSync data. RouteSync data reflects delivery assignments and completions recorded through the platform, but it does not capture all communications between hub managers and drivers. Phone calls, text messages, in-person conversations, and other informal communications that may bear on the degree of control exercised by SLG are not reflected in the RouteSync data. To the extent that hub managers issued informal directives or exercised control through channels outside of RouteSync, the data analyzed in this memorandum may understate the degree of SLG's actual control. The declarations of hub managers and drivers will be important supplements to the data-driven analysis.
4. Informal delivery refusals. Delivery refusal data reflects only explicit declines recorded through the RouteSync platform. If drivers informally communicated refusals to hub managers by phone, text, or in person — and those refusals were processed by the hub manager without being recorded as driver-initiated declines in RouteSync — the data may understate the actual frequency of delivery refusals. This limitation could cut either way: it could mean that drivers exercised even more delivery selectivity than the data suggests, or it could mean that hub managers redirected assignments in response to driver resistance in a manner that would not appear in the formal acceptance/decline records.
5. Data integrity. SLG's IT department has confirmed that there are no known data gaps in the RouteSync or GPS records for the relevant period. However, minor data transmission interruptions — estimated at less than 0.5% of all data points — may exist due to telematics hardware issues, cellular connectivity problems in rural delivery areas, or software glitches. These interruptions are unlikely to materially affect the aggregate statistics presented in this memorandum but should be disclosed in connection with Dr. Ramanathan's expert report to forestall any challenge to data completeness.
Details regarding data extraction methodology, query specifications, data cleaning steps, and validation procedures