FreightMath Alliance

Fall 2026 Meeting Recap

Session-by-session notes from two days of FreightMath at KSM Headquarters in Indianapolis.

Oct 6–7, 2026Indianapolis11 sessions

Eleven sessions covering the platform roadmap, MapLedger benchmarking, segment-level costing, AI in trucking, pricing tools, dedicated freight, and partner tools. Each FreightMath-led recap links to the matching slide in the Fall 2026 presentation deck.

The eleven sessions

01 · Welcome

Chris Henry, FreightMath

On the agenda
  • State of the Alliance: what has happened since Spring, membership and engagement.
  • FreightMath Roadmap Preview: a fast first look at what is new across the platform and what is on the roadmap.
  • The two days ahead: review the agenda and what to expect.

View this session in the Fall 2026 deck →

The platform is built on the following premise: in trucking you cannot be a neutral party. FreightMath has picked a side, and that side is carriers and the non-asset operations that run alongside those fleets. The FreightMath Alliance is a peer group, not a user group, where member feedback sets the roadmap. The FreightMath team is “small but mighty”: an eight-person team with more than 150 years of combined trucking experience.

State of the Alliance: what shipped since Spring

Almost every item is the result of a member request. The rebuilt SpotCheck (about 30 member suggestions) and new TripCheck both went live that morning, a dedicated pricing app sits inside the FreightMath app, the first API endpoints are in beta, and new benchmarking measures are in the dashboard. Members also asked for and got division slicers on the pricing pages and multi-operation dashboards, net fuel cost per company mile, peer to peer power benchmarks, and much more.

BidRight: paid alpha to early access in 20 weeks

BidRight has moved quickly. The slides show a paid alpha announced February 10 (carriers committing and paying before the product was finished), the alpha opening April 10 with more than 12,000 lanes imported in the first week, the first FreightMath-connected clients pricing real bids on May 5, and early access beginning August 28. The slides also list 21 weekly releases from April through September and 22 carriers in early access. Nussbaum Technology’s published results are roughly 20 times faster bid generation and pricing about 5 percent above market across 10 quarters.

The market backdrop

The facilitator’s read on the early-October truckload market: it is tight because trucks left, not because freight came back. The slides put DAT van spot at $3.01 all-in (reefer $3.63, flatbed $3.60), up 28 to 36 percent year over year, with EIA diesel at $6.38, up $2.63 a gallon, and SONAR tender rejections at 13.8 percent. Cass shipments rose 2.1 percent year over year in August, ending a 42-month decline, though Cass was cautious about calling it a turn. The sober point: much of the all-in gain is fuel passing through, van linehaul sits near $2.17 and spot is still below contract, so higher rates are not necessarily higher margins.

External data and the autonomy question

FreightMath scores every move against three independent rate indices: DAT (the market pulse), Carrier First (the rate across all current FreightMath clients, actual carrier-hauled freight normalized by one costing methodology), and Lane Best (the best rate a lane has already proven it will pay). Chris asked for input on which other indexes to add for 2027. He also introduced the FreightMath Autonomy Index, new for this meeting, which tracks the share of the for-hire truckload market running driver-out: low but growing, with a window to make money before the largest autonomous operators begin disintermediating carriers.

The two days ahead

The roadmap is deliberately flexible, reprioritized on member feedback, and AI runs through both days as three working sessions built around what members have built. The ask to the room was specific: bring back feedback on the roadmap, on alternative rate indexes for 2027, and on the AI use cases worth building, and run one SpotCheck before the next morning’s session.


02 · MapLedger: Current State & Benchmarking

Pete Richards, with Chris Henry

On the agenda
  • FreightMath Adjusted Revenue: how we scale costs so the FreightMath OR aligns with the OR from your GL.
  • Benchmarking: the standardized chart of accounts is what makes apples-to-apples benchmarking possible across the Alliance.
  • Best practices & use cases: how members are using MapLedger today and tackling common problems.

View this session in the Fall 2026 deck →

Pete Richards, with Chris Henry, walked through MapLedger’s current state and why it is the foundation supporting everything else in FreightMath. MapLedger’s job is to take a carrier’s raw general ledger and map it into FreightMath’s standardized chart of accounts, so that every carrier’s numbers line up in a common structure.

FreightMath Adjusted Revenue and the standardized chart of accounts

The standardized chart of accounts is what lets FreightMath scale costs so that FreightMath OR (FM OR) aligns with the OR a carrier sees from its own GL. Once every carrier maps to the same structure, benchmarking becomes apples-to-apples rather than a comparison of differently built P&Ls.

Dynamic distribution and smoothing

Two mechanics got attention. Dynamic functions let a single GL account, driver wages for example, distribute across several standard accounts (company driver, owner-operator, lease-purchase) based on the relationship among accounts the carrier has already mapped, so the split follows the data instead of a manual allocation. An adjustments feature on the mapping tab smooths month-to-month spikes, so a large or infrequent cost does not penalize a single period; smoothing runs up to twelve months, either manually to a stated monthly value or as an auto-variance spread across several months.

Benchmarking in the review page

The rebuilt review page consolidates mapped results back up into a P&L view, and benchmarking now sits directly inside it, with a bench column and modals to drill into peer comparisons. The benchmarking page is live. Carriers outside the TCA organization see their comparisons against other FreightMath clients; TCA members can see both the FreightMath set and their own groups within TCA. The stated goal is clarity on two things at once: how a carrier is doing versus peers, and how much opportunity is on the table.

From history to forecast

The throughline is discipline: a carrier has to understand its historical performance precisely, on a common structure, before it can forecast or benchmark anything. Deeper value comes from moving past summary GL figures into the granular, transactional data underneath. Members also discussed using contextual filters such as length of haul or geographic region so comparisons stay meaningful.


03 · Segment-Level Costing

Dave Dunst, FreightMath

On the agenda
  • How it works, end to end: a full walk through now that SLC is live and members have seen the results.
  • Driver pay: how compensation is distributed down to the individual segment.
  • Allocation of cost buckets: how each bucket is allocated to segments.
  • Deadhead models: choosing a deadhead model (pre, split, post).
  • Forecasting: how segments are costed before the GL month closes.

View this session in the Fall 2026 deck →

Dave Dunst gave the full walk through on segment-level costing now that it is live and members have seen results. SLC is activity-based costing taken down to the individual segment, and it rests on three components: a standardized GL (via MapLedger), orders, and segments. Orders capture customer intent; segments capture how the work was executed, which is what determines cost. Every point-to-point movement, loaded or empty, carries time and miles, both of which drive the allocation.

Standardized miles, times, and guardrails

Miles and times are standardized per segment so costs are comparable across carriers. A guardrail algorithm corrects impossible TMS timestamps before they distort anything. Hours-of-service logic scales break time as a segment approaches eight hours, with a 30-minute break fully accounted for at eight hours, and the 10-hour break fully accounted for at 11 hours. The algorithm ensures short segments are not overstated. Segment hours include pickup time, drive time, and mandatory stops, not just drive time.

Deadhead models

Empty miles and time are assigned with one of three models: pre (empty before pickup charged to the following load), post (empty after delivery charged to the delivered order), and split (empty divided between the preceding and succeeding orders). Members shared live practice: one runs split for fairness even though their TMS (McLeod) defaults to post; another moved from pre to split, noting the overall difference was small but the impact by account was significant. The plan is to set the chosen model as a client database-level default. That preferred model will flow directly into BidRight.

Driver pay, trailers, and the cost buckets

Driver pay is distributed down to the segment across mixed pay types (mile, hour, differentials, guarantees): pay tied to a dispatch goes to that dispatch’s segments, weekly pay not tied to a dispatch spreads across all segments the driver ran that week, and safety bonuses and PTO are held out of segment allocation but reconciled to the GL. Any segments with no pay receive an estimated amount based on driver type (company, owner-operator, lease-purchase). Pay is pre-distributed, then the variance to the GL total is computed separately for owner-operators, company drivers, and lease-purchase drivers and redistributed per mile.

A new option estimates the most recent week’s driver pay to get metrics out faster before payroll finalizes, trading some final accuracy for speed. That option estimates pay by individual driver ID, then falls back to driver type if no pay history exists.

Trailer cost follows presence on the segment; pools are costed in trailer-days (10 trailers for 20 days is 200 trailer-days) based on trailers committed to a customer, with a pool’s total split across the orders originating from or delivering to it. Segments starting at a pool and ending at another pool take 50 percent of their cost from each pool. Remaining trailer cost spreads across the general fleet of trailers. Power-only moves can exclude company trailer cost, and LCVs and doubles distribute revenue and cost to both trailers.

Mileage-distributed buckets (purchased transportation, fuel, DEF, IFTA, lumpers, insurance) spread by billable miles at the order level; reefer cost runs pickup to delivery and scales by temperature tier; tolls default to a modeled share from origin and destination ZIPs unless the carrier supplies actual toll data.

Overhead, forecasting, and OR by tractor

Fixed overhead, non-driver compensation, and technology use a hybrid split: half allocated evenly per order, half distributed by share of hours. Overhead forecasting moved from calendar days to a percentage-of-weekdays-completed basis on a standardized workday calendar, so holiday week forecasts are more accurate.

Fuel is moving from a baseline DOE forecast toward a segment-level model that ties fuel rates to ZIP-code pairs for geographically accurate rates. Clients will have the option to choose their fuel cost distribution model (rack, POD, DOE average, retail).


04 · FreightMath Platform Improvements: Dashboard & Spot Check / Trip Check

Adam Smith & Jordan Nelson, FreightMath

On the agenda
  • Dashboards: the current state of the FreightMath dashboard, including the new features and pages added since Spring.
  • Spot Check & Trip Check: the apps that put load- and trip-level OR in front of the pricing and dispatch decision, worked hands-on in the FreightMath app at app.freightmath.com.

View Dashboard Updates in the deck →View the Pricing Suite in the deck →

Adam Smith and Jordan Nelson split the session: Adam on the dashboard refresh, Jordan on the rebuilt pricing suite. The thread is a move off Excel and Power BI toward a web platform, app.freightmath.com, FreightMath’s first proper web application, which removes the need for client Power BI Pro licenses. SpotCheck is the first app on it.

Dashboard: pricing opportunity, lane by lane

The Pricing Opportunity page (renamed from Price Impact) ranks every Bill To’s lanes against three rate indices: Lane Best, Carrier First, and DAT. It shows the dollar gap per load and tags a confidence tier (high when two or more references sit above the current rate). An adjustable OR Target recalculates which lanes clear the bar.

The Load Detail page was redesigned so an order opens into four card groups (load info, geography, revenue, cost) with the deadhead model and Core OR, FM OR, and IB/OB OR all on one screen. New measures included net fuel cost per total company mile, peer per-power benchmarks, three-way velocity (total, transit, loaded), and revenue per power mile.

A new Logistics Dashboard shows rolling four-week tiles, each compared to the prior period and the same period last year.

The pricing suite: SpotCheck rebuilt

SpotCheck now runs 4,000 to 6,000 times a month across FreightMath clients, and it was rebuilt from a Power Apps tool into a web application: every external connection (miles, tolls, fuel, market rates, cost model) now runs in parallel instead of in sequence, so a quote returns far faster. The method is unchanged — start with the target margin and back into the rate: revenue needed equals modeled cost divided by target OR, on the carrier’s own costs.

SpotCheck now prices against three distinct targets: Core OR (a load’s standalone one-way profitability, for a single lane on its own), FM OR (Core plus inbound and outbound area economics, for ongoing business), and FM OR2 (Core plus outbound, inbound excluded, for broker and one-off loads where the truck is already in position). Round trips are native in one entry, and every cost assumption from MapLedger is visible and editable.

TripCheck and the API

TripCheck is new: build the first order like a SpotCheck, then add orders and deadheads to see the profitability of a whole trip or a driver’s week. The key difference is deliberate — TripCheck ignores network inbound and outbound economics and prices only the orders entered, so it judges a tour on its own merits.

The FreightMath API is in beta: one endpoint returns a client’s enriched data for the last six months, another evaluates a potential load with the same methodology as SpotCheck, and an MCP connector is in development so agents can connect directly. Jordan takes API access requests.

Fuel by Bill To

With diesel moving quickly, the Fuel by Bill To page shows the EIA weekly rate next to the expected surcharge change under the rule of one cent per mile for every 7.5-cent-per-gallon change, staggered a week, with a green check or caution flag per customer and a slicer for Breakthrough Fuel customers (who must be flagged in network configuration). Net fuel cost per total company mile is a recovery measure — fuel cost net of surcharge — to watch as diesel rises.

Fuel strategy under rising diesel

The session also treated fuel as a margin problem. KSMTA estimated the run-up at about $4,000 per truck per month in additional cash carrying cost versus a year ago, before any time value of money, and noted the math rests on assumptions about haul and MPG. Breakthrough Fuel programs drew criticism as opaque. KSMTA put the deviation of Breakthrough tables from DOE baselines at roughly 6 to 12 percent, varying by client, and advised carriers to ask for the formula for each customer and to check whether a customer has multiple Breakthrough tables.

In a show of hands, roughly 30 to 40 percent of the room had gone back to shippers for an emergency fuel surcharge. One carrier said Breakthrough customers on a zero peg were the most resistant, while customers with a higher peg were more open and a few added a premium on their own. On tactics, members reported pairing fuel-optimization software with driver accountability and locking fuel cards to specific locations or volumes.

David Roush closed with the harder-nosed view: identify the customers where fuel is not being recovered and take it to them directly, because for now it is a negotiation with customers.


05 · The 5 Strategic Mistakes That Sink Change in Trucking

Jim Stockeland, Fractional COO, Your C-Suite, with Abby Ryan, President, Crawford Trucking, and Max Herbst, Vice President of Finance, Leonard's Express

On the agenda
  • Pitfall 1: starting with the tool instead of the target.
  • Pitfall 2: nobody owns it (role confusion).
  • Pitfall 3: mistaking go-live for the finish line.
  • Pitfall 4: keeping the old way running “just to be safe.”
  • Pitfall 5: asking a maxed-out team to run the business and lead the change.

This panel approached change management from the leadership seat, with fractional executives and operating carriers who have run real initiatives. The frame was blunt: most change efforts in trucking fail for predictable reasons, and each has a fix. Jim Stockeland (Fractional COO, Your C-Suite), Abby Ryan (President, Crawford Trucking), and Max Herbst (Vice President of Finance, Leonard’s Express) worked through five pitfalls with their own examples.

The five pitfalls

  • Starting with the tool instead of the target. Change without a clear ROI does not work. Leonard’s Express was spending about $70,000 a month on a third-party driver-recruiting system with weak ROI and an unseated-tractor count over 10 percent; moving recruiting to internal marketing and HR cut recruiting expense by nearly 70 percent while holding class sizes.
  • Nobody owns it. Breakdowns come from blurring authority (who decides) and accountability (who owns the outcome), and from multiple decision-makers with no final say. Crawford Trucking’s presidential succession — a ten-year path accelerated with outside help — worked because a firm date was set and decisions were presented as a unified front. Ownership needs direct conversation, not email, and blockers either align or step aside.
  • Mistaking go-live for the finish line. Adoption is a marathon. One carrier described a clean switch to a new telematics platform on a one-month deadline with sustained leadership support, and treated the first 90 days as the real test, using structured roundtables to surface resistance with data. Early adopters need continued attention or momentum dies.
  • Keeping the old way running “just to be safe.” Parallel systems (two TMSs, or spreadsheets beside a new tool) create double work and block full ROI. Once the new system is live, commit; optimizers in particular only pay off at high compliance, not partial use.
  • Asking a maxed-out team to run the business and lead the change. A team at capacity cannot do both. Crawford Trucking added a vice president during its transition and grew 13 percent.

The checklist

The panel closed with five questions to pressure-test any initiative: do we know the ROI, does someone own it, are we measuring adoption rather than just the go-live date, did we stop running the old system, and does the team have the capacity to lead this.


06 · Beyond AI Fatigue: Why Trusted Data Comes First

Moderated by Eric Jacobson, CEO, Tivitri, with Adam Blanchard, CEO, Double Diamond, and David Knight, Data Engineer, Argo Analytics

On the agenda
  • AI fatigue: why so many AI initiatives stall in fleet operations.
  • Operational problems become data problems: why the work routes back to data.
  • Clean, connected data: building trustworthy telematics and operational data.
  • Master data and AI: applying AI in ways leaders can trust.

Moderated by Eric Jacobson, with Adam Blanchard and David Knight, this panel took on AI fatigue directly: every fleet is being pitched AI, and most are tired of it. A confidence poll on large language models landed in the middle of a one-to-five scale, with accuracy the recurring worry. The panel’s claim was that the fatigue is misplaced, because the real work is not AI, it is data.

Operational problems are data problems

Blanchard described realizing his first instinct, deploying AI agents, was backwards. The underlying issue was inconsistency between his TMS and ERP (Microsoft Business Central), so the fix was to build a database that aggregates and integrates the TMS, the ERP, and telematics (Samsara) first. David Knight’s point reinforced it: historical data is the raw material for prediction and is gone forever if it is not captured, so carriers must control and retain their own data rather than trust a vendor to expose it.

Why clean data makes AI usable

The panel was direct that AI produces inconsistent output from unorganized data, and will give different answers from the same raw data without business context. Getting data into one place, cleaning it, and applying explicit business rules and a semantic layer is what turns AI from a hammer into a scalpel. The foundational work, making all data accessible and automatic, is roughly 80 percent of the effort; the AI itself is the cherry on top. The output is a set of master datasets modeling core entities (truck, customer, driver) as a single source of truth, which can then be used for things like testing retention and turnover hypotheses.

Build and own

Blanchard’s strategic argument was to build and own rather than subscribe: owning a centralized database reduces vendor leverage at renewal, makes switching software far easier, and insulates against fly-by-night vendors disappearing. The consolidation work also forced internal process fixes, in his case correcting thousands of reconciliation issues and enabling a one-day month-end close. He now buys only software with open APIs for read and write-back. A small carrier in the room confirmed the approach was affordable as a one-time capital expense, with later integrations cheaper, and the recommendation was to start with one high-ROI, manual, time-consuming process.


07 · How Carriers & FreightMath Are Using AI

Facilitated by Chris Henry, FreightMath

On the agenda
  • Part 1 — Agent Challenge report-back: members show the agents they built since Spring and what they saved; live use cases from spot-pricing bots and daily fuel-optimization maps to driver-risk monitoring and institutional memory; vendor AI versus internal build.
  • Part 2 — Where AI is moving OR: a working framework to quantify the return, governance and data handling that keeps driver, customer, and employee data out of un-consented models, and one automation to put in place before Spring 2027.
  • Part 3 — Closing the loop: what each member takes home and what FreightMath should build.

View Part 1 in the deck →View Part 2 in the deck →View Part 3 in the deck →

The AI track ran as three working sessions across the two days: Part 1 (Tuesday) was the six-months-on report-back on the Spring challenge to put an LLM to work, wins and concerns round the room; Part 2 (Wednesday) pushed past personal use into the business, guardrails, and what members would hand to AI next; Part 3 (Wednesday) closed the loop on what each member would take home and what FreightMath should build.

Recruiting and retention

The recruiting use cases ranged from AI-generated marketing content driving leads, to AI agents that prescreen candidates against qualifications (with disclosure practices varying), to a drivers-first referral program built on QR-coded cards with bounties for accepted referrals.

Members were candid that responsive human recruiters often beat AI call-and-text tools on first-contact speed, and one flagged a coming surge of automated inbound calls from consumer AI assistants. On retention, one carrier built an AI repository from internal communications, retention-call notes, and TMW data to flag flight risks and recommend actions; another uses call transcription with sentiment labeling as an early indicator.

The non-AI tactic that drew the most agreement: a CEO personally calling new recruits a week before orientation to improve show rates. Sign-on bonuses are being used selectively, by location and need.

Web presence for AI search

A distinct theme was AI Engine Optimization — optimizing to appear in AI-generated answers rather than only traditional search results. Members described upgrading sites and building robust FAQ pages so AI systems can scrape and surface them, since AI answers tend to name only a few carriers, and the facilitator encouraged structuring content (open repositories, markdown files) for machine readability. Local SEO, video, and Google reviews were still cited as heavy ranking factors.

Maintenance and fleet decisions

On maintenance, one operator sends personalized, tone-adjusted emails to managers about overdue preventive maintenance, improving compliance, and connected maintenance data to the shop system to ask preemptive-replacement questions. Another moved maintenance to a granular cost-bucket model, assigning overhead and facility costs directly to the maintenance division to reveal true cost per unit, and urged auditing shop expense out of general facility costs. On replacement decisions, members agreed a full data model (MPG, downtime, repair history) beats mileage alone, with examples of keeping a reliable high-mileage truck while retiring newer problem units.

Detention, vendor risk, and owning your strategy

One of the more concrete builds: one carrier put cameras on flatbed trailers and uses AI image analysis to track in and out times for detention automatically, compiling video as proof for claims, at accuracy in the low 90-percent range. KSMTA added that data speed is what makes this kind of automation work, and warned against long-term hardware contracts because the hardware dates quickly.

That led into tech-vendor volatility: the facilitator used a VC-funded AI vendor that pivoted out of trucking as a cautionary tale, and pressed the point that carriers must own their technology strategy rather than be led by vendors, with optimizers increasingly open-source and buildable in-house on shared, platform-agnostic tools.


08 · BidRight

Austin Knobloch, Director of Product Development, Nussbaum Technology

On the agenda
  • Where BidRight is today: bid import to commitments, FreightMath built in.
  • Use of AI: Auto Rate with reasoning, contract ingestion, bid insights.
  • Where BidRight is going: selections from the roadmap.

Austin Knobloch presented BidRight, the bid automation and intelligence tool Nussbaum Technology built as a KSM partner, drawing on lessons from Nussbaum Transportation. About a year old, the platform has processed more than 608,000 lanes across 700-plus bids for 20 clients and was named to the FreightWaves FreightTech 100.

The workflow, demoed live

Austin ran a sample RFP end to end. BidRight ingests the shipper’s spreadsheet, auto-mapping some columns and letting the user map the rest so nothing is lost, then normalizes data such as trailer types and ZIP codes. An Interest stage suggests which lanes to bid based on network preferences; an Auto Rate feature proposes rates from user inputs (market temperature, target OR), with a Rate Logic column explaining the rationale and AI reading the uploaded bid documents for high-level insight. Rates export back into the shipper’s original Excel format for submission, with some fields such as transit time still entered by hand.

FreightMath inside the bid

Profitability data was built in early, so each lane carries FreightMath historical OR and a modeled OR, and rates can work backward from a target OR, with rate per mile normalizing lanes for comparison. The tool reads DAT RateView via manual upload today, with API integration planned. It handles Breakthrough Fuel and markups but does not yet support all-in rates or percentage-based fuel surcharges, both on the list.

A Teams concept supports multi-division bidding, so an asset team can price first and a logistics team follow within the same bid, triggered by trailer type, and a table view with pivot-like grouping plus a review-by-area flow helps align rates across similar origins. Contract versus spot can be distinguished only if the client tags it in their TMS.

Commitments and the 21-week release run

The headline new feature is Commitments: a user publishes awarded bids as commitments, and BidRight overlays actually hauled loads through FreightMath to check whether promised volumes are being met, by both carrier and shipper.

BidRight has had 21 weekly releases from April through September, including Auto Rate with reasoning, AI Bid Insight (July 31), the AI contract reader (September 4), commitment tracking (September 4), and commitments matched to actual loads (September 18). Multi-round bidding lets users pick a break-back period, and a snapshot feature shows what inputs changed between rounds. Members are entering historical bids and awards to build the data foundation for commitment analysis.

Roadmap

Austin outlined custom market areas for visualization, FreightMath data refreshes moving from weekly toward near real-time by 2027, APIs starting with commitments (documented but not yet built), and AI features under consideration: a Q&A interface for lane targeting, collaboration tools, light CRM and revenue projection, and agent-assisted bid processing. Longtime users in the room endorsed it on measurable efficiency, citing less staffing for pricing and faster bid iterations. The team also asked for feedback on which additional rate indexes carriers want supported.


09 · Securing Profitable Dedicated Freight

Brad Heisterkamp, FreightMath

On the agenda
  • Advanced dedicated pricing techniques: pricing the fleet, the driver, and the schedule instead of the lane, so the rate covers what dedicated actually costs to run.
  • Marketing to dedicated opportunities: finding and positioning for the shippers whose freight fits a dedicated model, and making the case with FreightMath data.
  • Contractual guardrails: the terms that protect the business once the freight is won — volume commitments, rate adjusters, and exit provisions.

View this session in the Fall 2026 deck →

Brad Heisterkamp walked the advanced dedicated pricing model, built to fill the gap that one-way tools (FreightMath, BidRight) leave: dedicated is contractual and multi-year, priced once and run for years, with a fleet and driver mix that often differs from the one-way network. Brad is VP at KSM Transport Advisors, a former FreightMath client, with more than 20 years of truckload leadership at Ascend, NFI, Transco Lines, and CRST. Over roughly 50 iterations the tool now runs one workflow end to end: cost it, model it, propose it, and generate the contract. It includes a basic capacity optimizer and a guided modeler, though Brad was clear it is not yet a deep optimizer for complex routing.

A demo opportunity, input by input

He priced a sample opportunity live: a new dedicated fleet out of an Indianapolis DC, eight trucks to start with room to grow, a three-year term with an annual true-up, a Q1 2027 go-live, a mix of drop-and-hook dry van plus a few reefers, and a customer that wanted the rate split into a fixed piece and a variable piece rather than one all-in number. The model took 11 inputs from four departments: six pulled straight from the GL (regional and spot driver pay, fuel cost per mile, cost per hire, insurance per mile, overhead per truck) and five adjusted for the deal (volume and lane commitment, the rate-expectation ceiling, a near-zero deadhead assumption for the shuttle runs, a stretched time-to-staff ramp for a new hiring market, and new tractor and trailer quotes rather than existing book value). That intake, not a spreadsheet, is where every dedicated deal starts.

Capacity planning and the guided modeler

The walkthrough starts from a lane list (template-supported) with automated field mapping and manual overrides, then a miniature optimizer estimates trucks and hours of service, turning a naive one-truck-per-lane plan into a tighter fleet (an example moved 22 trucks to 14) using baseline buffers such as a 15-minute yard time and a 30-minute stop. Sensitivity tests probe speed reductions, time-window slack, added days, and mid-week volume peaks, with risk flags on lanes that cannot be completed in a day. From there a 19-step modeler captures commercial terms, margin targets and contingency, cost of capital, externally estimated tolls, equipment policy, wage bands and off-shift differentials, temp-labor planning, and driver-transfer assumptions when displacing an incumbent.

The costing model and revenue strategy

Costs roll up with color-coded alerts when inputs deviate from baselines (illustrative defaults included roughly $2,350 monthly lease or depreciation, $0.15 a mile maintenance, trailers near $350 at a 2:1 ratio, a $72,500 wage, recruiting around $2,600 per driver, and a placeholder 25 percent benefit load). Overhead is held in a dedicated silo rather than burdened with one-way overhead, with a 6 percent of revenue baseline, and Brad noted labor typically runs 40 to 50 percent of dedicated revenue. Pricing balances fixed and variable to insulate margin: carrier-favorable mixes run 50 to 70 percent fixed, shipper-favorable 40 to 50 percent. Fuel and DEF mirror the billing basis (DEF near 87.7 percent of a diesel gallon’s cost), and CPI caps are stress-tested across the term. Revenue share was framed as a poor carrier’s dedicated: the shipper pays round-trip rates and takes a backhaul credit (80, 85, or 90 percent) while the carrier keeps a 10 to 20 percent spread, which only works with disciplined backhaul execution; BidRight can quantify the backhaul potential.

Contract protections and the cafeteria builder

Brad spent real time on the terms that protect dedicated once it is won: minimum volumes, hours, or billable amounts; guaranteed volume and first right of refusal; access rights such as parking, which he called a frequent flashpoint; mileage guarantees; exit clauses and equipment buybacks on termination without cause; wage carve-outs for fast pass-through of market wage increases; and a 10-percent-plus swing clause on insurance or equipment that triggers good-faith renegotiation. A cafeteria-style builder assembles an MSA or addendum from pre-approved blocks (rev-share rules, guarantees, KPIs, termination, indemnification, scope, rates, term), and the proposal tool turns rate inputs into a customer-facing PDF with scope, charge types, annual pricing, and signature lines. One caution he stressed: state the valid pricing period explicitly so a proposal does not accidentally lock 14 to 16 months of pricing.


10 · Future of FreightMath

Moderated by Chris Henry, with the FreightMath team

On the agenda
  • Telematics: integrating telematics to replace inferred miles with actual miles, targeted for next year, for more precise hours and cost.
  • Machine learning: the normalization and prediction layer, including hours-of-service normalization that estimates reload potential beyond what the raw data shows.
  • Insights: a triage-style start page that surfaces prioritized action items on login, dismissible and logged, so the platform tells you what to act on.
  • Fuel cost: lane- and segment-level fuel costing with a per-segment fuel index for far more geographically accurate cost.
  • Dedicated pricing: the advanced dedicated pricing wizard and cafeteria-style contract generator.
  • Freight vs. ops: separating pricing quality from operational execution, so a lane reads as what actually happened against what could have happened under standard operations.
  • Control Room app: broker/shipper, new area control, trailer pools, strategic footprint, custom network logic, charges and more.

View this session in the Fall 2026 deck →

The FreightMath team closed the meeting by laying its roadmap candidates on screen, grouped by build effort, and asking each member carrier to vote on which to build first. The exercise doubled as a demonstration: Chris Henry fed the live vote to an AI tool and had it sequence the list into a first-cut build order on the spot. Driver and Truck P&L and FreightMath Actions came out on top, with LLMs on a carrier’s own data as the feature that drew the most discussion.

How the winners were chosen

Chris Henry moderated. The team put its candidate features on screen, grouped by build effort (low, medium, and high lift), and asked each member carrier to call out its top three. Chris fed the tallied votes to an AI tool and asked it to sequence the list, weighting by hands raised and by lift, to produce a first-cut order: low counts as one, medium two, high three, with four features to a window and anything with no votes sitting out. Chris was explicit that the output was a starting point for the conversation, not a commitment.

Next up: Driver and Truck P&L

The top vote. Members wanted profitability pushed down from the division level to the individual tractor and driver, with the ability to group results by driver manager. The team said the first version would carry settlement, fuel, revenue, and tolls to the tractor level, with everything else blended, since transactional maintenance cost is not yet available per tractor.

Next up: FreightMath Actions

The other top vote. This reimagines the insights email into pointed, prioritized actions rather than a wall of pages, and pulls cost data into the notifications, not just price and lane opportunities. The goal the team described: communicate in natural language the handful of things a carrier should focus on, including when the operating ratio is lagging on cost rather than on price. It also ties to the persona work, so a finance user, a CSR, a pricing lead, and an executive each see the actions that fit their role.

Following: LLMs on your data

The feature that drew the most discussion, slotted for the phase after the first set lands. Members asked to interact with their data in natural language, and the team confirmed this is planned inside the new web application, with API access so carriers can build their own uses on top of the order, SpotCheck, and benchmarking data. The team framed the division of labor it intends to hold to: the model asks and phrases the questions, and FreightMath does the math, so reported numbers trace back to the platform rather than being generated by the language model.

After that: calendar month intervals

Calendar month intervals answers the finance users’ most frequent request: P&L on calendar-month blocks instead of four-week rolling periods, to line forecast up against actual. The team confirmed a normalization already exists for carriers on 4-4-5 or 4-5-4 accounting calendars, as long as the carrier supplies its accounting periods in advance.

Backlog: revisit at Spring 2027

Lower-vote items held for later: telematics, forecasting and budgeting within MapLedger, decision automation with BidRight, freight versus ops, dedicated and non-OTR profitability, and HOS normalization. These stay on the board rather than off it, and the team flagged them to revisit at the Spring 2027 meeting.


11 · MileMaker: The Latest in Mileage & Routing

Mark Lukenbill, MileMaker

On the agenda
  • Where the miles do differ, it is real: MileMaker updates its maps four times a year, so a changed weight restriction or road closure shows up as a genuine, defensible difference.
  • Why it matters: miles and time are the denominators in nearly every transportation metric, and routing you can count on is what keeps those numbers accurate.

Chris Henry introduced Mark Lukenbill of MileMaker, the mileage engine KSM uses for mileage, transit time, toll cost, and routing. Lukenbill walked its history and structure: Rand McNally is about 170 years old, and roughly five years ago it split into three units — publishing (the road atlases), hardware (cab devices and cameras), and MileMaker (commercial mileage). MileMaker runs as its own business, TMS-agnostic, and does not require a Rand McNally device.

Map updates and why the miles differ

MileMaker pushes about four map updates a year. A dedicated GIS team fields DOT notifications (road reclassifications, new construction, closures), validates them, and batches them into the map. There are two guides, 19 and 20, with every update now going to Guide 20. Mileage differences between providers are real and come from methodology (how lanes and off-ramps are measured), which is the point the agenda flagged: where the miles differ, it is a genuine, defensible difference.

How carriers access it

MileMaker is sold at a single flat price with no add-ons, and that price covers every access method. On-prem IntelliRoute is still supported, including AS400 environments, with no plan to sunset it. There is a modern API, a web platform that visualizes bridge heights, weight restrictions, weather, and road conditions, and an Excel add-in built as an RFP module: point and click lanes and it returns miles, duration, and total cost, up to 10,000 lanes at a time. Coverage is full North America, including Canada and Mexico. The platform supports single sign-on and, Lukenbill stressed, live human support with response times measured in minutes. MileMaker has also built its own MCP server and an in-product AI agent, so a user can ask for the miles between two cities in plain language and get an answer in about four seconds without building a route.

Routing options and what is in development

Beyond a single shortest-mile answer, MileMaker offers household-goods, quickest, practical, and avoid-tolls routing, plus custom avoids (ferries, U-turns, border crossings, school zones, specific roads) and custom locations. Lukenbill floated a least-cost model that would balance tolls, fuel, and mileage, which the engineering team considers feasible but has not scheduled. Two items are in active development: an hours-of-service module that factors driver availability into optimization, and route tracing that connects telematics to overlay the planned route against the actual GPS track, so a carrier can see where planned and driven miles diverged and investigate. Batch routing is already supported, and Lukenbill raised co-building the trace feature with KSM.

Pricing and the PC Miler comparison

Lukenbill positioned MileMaker as typically 30 percent or more below PC Miler for equivalent use, with the flat price bundling the API, web app, Excel add-in, MCP server, and on-prem access, and no per-API-call fees. Commercial terms are flexible, Capex or Opex, priced on forecasted usage such as annual shipments rather than a rigid schedule. He offered a free 30-day proof of concept for any carrier that wants to test MileMaker against PC Miler. On the competitive side he pointed to PC Miler contract terms worth reading closely, including 90-day auto-renewal clauses with built-in price increases, and to the pattern of giving shippers the product free while carriers pay enterprise licenses. Because shippers often dictate which mileage provider a carrier must use, his ask to the room was to raise MileMaker directly with their shippers.

Implementation and integrations

MileMaker integrates with most major TMS platforms, and switching is often as simple as generating and pasting an API key, a quick step that depends on the TMS. A recent milestone Lukenbill called out is an Oracle OTM integration, with the contract signed in December 2025.