How Do I Get My Freight Brokers to Grow Margin Per Load?

The job this scorecard is hired to do

Margin per load erodes for one structural reason, and it is not laziness: brokers are compensated to cover freight, so they optimize for the fastest cover, not the fattest spread. A broker staring at a load that needs a truck in four hours will take the first carrier who answers at the posted number, because a covered load pays and an uncovered load is a service failure with a shipper phone call attached. The incentive is doing exactly what you built it to do. The problem is what you built.

The job this instrument is hired to do is narrow: make the economics of a book legible enough that a manager can tell the difference between a busy broker and a profitable one, and make that difference visible to the broker in the same week the behavior happens. Right now most desks cannot do this. The weekly report shows loads moved and revenue booked. Revenue is the most misleading number on a brokerage floor, because a $2,800 linehaul with a $2,610 buy rate looks bigger than an $1,900 linehaul with a $1,520 buy — and the second one paid the lights.

Concretely, the scorecard needs to carry eight or nine lines, each of which describes something a profitable broker actually produces:

Spread per load, in dollars and as a percentage. Both, not one. Percentage alone rewards short cheap freight; absolute dollars alone rewards long hauls with thin points. A broker running 15% on $1,400 loads and one running 9% on $3,200 loads are producing $210 and $288 respectively — you want to see both columns before you decide who is winning.

Rate-per-mile discipline on the buy side. Does this broker consistently source below the lane's prevailing RPM, or do they pay at or above market because they started calling forty minutes before pickup? Late sourcing is the single most common margin leak on a spot desk, and it is measurable: timestamp the load acceptance, timestamp the carrier booking, and look at the gap.

Carrier sourcing depth. Count distinct carriers used per broker per quarter. A broker with eleven carriers in their rotation is a broker with eleven price quotes; a broker with three is a broker who is going to get squeezed the first week one of them gets busy. Depth is a leading indicator of future spread, not a vanity metric.

Contract-versus-spot mix. Contract freight stabilizes the book and usually carries a thinner but far more predictable spread; spot is where the fat margins and the blowups both live. The right mix is a leadership decision, not a broker decision, and the scorecard is where you express it.

Deadhead and empty-mile exposure. Every empty mile a carrier eats gets priced into what they charge you next time. Brokers who pair loads intelligently — who know that the reefer they just dropped in Laredo can pick up two hours away instead of running 300 empty — buy cheaper because their carriers make more.

Detention and accessorial recovery rate. This is free money that most desks leave on the table. Detention, layover, TONU, lumper reimbursement, driver assist: if the broker documents and bills it, it is margin; if they eat it to keep a shipper happy, it is a discount they gave away without authorization. Track recovered dollars against billable events.

Claims and on-time reliability. A broker who hits 18% spread by using the cheapest possible carriers and racks up OS&D claims is not producing margin, they are producing a liability that shows up two quarters later. Score it, weight it, and let a bad claims line drag the composite.

Shipper reorder frequency. The stickiest measure of whether a broker is building a book or renting one. Count how many shippers gave them a second load within thirty days.

Set a weight on each line — leadership decides, not the desk — grade every broker 1-to-5 per line, and the composite is the sum of weight × level. The broker booking 40 loads a week at 9% should score *below* the broker booking 22 at 18%. Until the scorecard says that out loud, your best negotiators are invisible, and invisible top performers eventually go work somewhere they can be seen.

How it fits the RevOps stack

The scorecard is not a standalone artifact. It is a read layer sitting on top of your TMS, and the entire question of whether it works is a question of whether the data reaches it without a human typing anything.

The load record in your TMS already contains nearly everything: customer rate, carrier rate, miles, lane, pickup and delivery timestamps, accessorial line items, the carrier MC number. What it does not contain is the shape you need — per-broker aggregation with weighted grading. That transformation happens either in a BI layer (Looker, Power BI, a warehouse view) or in a purpose-built scoring tool. Either is fine. What is not fine is brokers self-reporting margin into a spreadsheet, because a matrix fed by manual entry rots inside three weeks and a scorecard nobody trusts is strictly worse than no scorecard at all.

Three integration details determine whether this survives contact with a real desk.

Latency matters more than precision. A scorecard refreshed nightly changes behavior. One refreshed monthly does not, because by the time a broker sees that their sourcing discipline slipped, they have booked another eighty loads the same way. Nightly is the floor; intraday on spread is better.

The comp engine and the scorecard must read the same field. If your commission calculation uses gross revenue from the accounting system and your scorecard uses spread from the TMS extract, they will disagree, brokers will notice within a pay period, and you will spend the next month adjudicating disputes instead of coaching. Pick one source for margin and make both systems read it.

Someone owns the weights. This is a RevOps function, not an IT function. The weights are a strategy expressed numerically, and they need to change when the market changes — which means the person who owns them needs the authority to change them and the visibility to know when.

Adjacent teams get pulled in whether you plan for it or not. Carrier sales cares about sourcing depth because it justifies their onboarding pipeline. Accounting cares about accessorial recovery because it changes DSO. Customer success cares about reorder rate because that is their number wearing a different hat. Building this as a shared object rather than a sales-manager artifact is what keeps it alive past the first quarter.

Pricing, engagement models, and typical ranges

You can build this at four cost tiers, and the honest advice is to start lower than you think you need to.

Tier zero — spreadsheet, $0. List the KPIs, set the weights, grade 1-to-5, let one formula roll the composite. This is where most brokerages should genuinely start, because the argument you have with your leadership team about what deserves 30% weight versus 5% is the actual value of the exercise, and software will not have that argument for you. The cost is downstream: someone has to update it, and the first genuinely busy week nobody does. Budget four to six hours a week of a manager's time, and expect it to go stale within a quarter unless one person's job description explicitly includes it.

Tier one — BI on top of your TMS, low hundreds per month plus build time. Power BI, Looker Studio, Metabase. You are paying for licenses and a few weeks of someone's time to model the load table. This is the best cost-to-durability ratio for a desk of ten to forty brokers, because once the extract is scheduled, the scorecard maintains itself. The trap is treating it as a one-time build — the model needs an owner when the TMS schema changes.

Tier two — scorecard and coaching platforms, roughly $15 to $50 per user per month depending on tier and negotiation. Tools in this category — Ambition, Spinify, Hoopla among them — construct weighted multi-metric scorecards, push them to floor displays and Slack, and hang a coaching cadence off them. They differ meaningfully in emphasis: some are genuinely built around weighted multi-KPI grading, others are motivation and recognition layers where the weighting is thin and the leaderboard is the point. Ask specifically whether the tool supports multiple weighted metrics rolling to a composite, or just multiple leaderboards side by side, because vendors describe both the same way.

Tier three — incentive compensation management, custom pricing, generally the largest line item. CaptivateIQ, Xactly, and similar platforms exist to run multi-component commission plans that would break a spreadsheet — tiered spread payouts, differentiated rates on contract versus spot, clawbacks on claims, dispute workflows so a broker can formally contest a mis-credited load. If your entire margin strategy is going to be enforced through comp, this is where it lives. Pricing is quote-based and scales with headcount and plan complexity; the administrative overhead is real, and reaching for it before your plans are genuinely complicated buys you a system to maintain rather than a problem solved.

There is a fifth category worth naming even though it is not a scorecard: conversation intelligence. Tools like Gong score the call itself, which surfaces the thing the margin data cannot — whether a broker even attempted to negotiate the buy rate, or accepted the first number. That behavioral signal leads the spread number by weeks. It is a complement to the scorecard, not a substitute, and it only makes sense on desks where brokers negotiate by phone at volume.

A note on sequencing the spend. The common failure is buying tier three to fix a tier zero problem. If nobody on your floor can articulate what a profitable load looks like, a comp engine will pay them precisely and confidently for the wrong thing. Prove the model cheap, then buy automation.

How to evaluate and shortlist

Six questions, asked in this order, will eliminate most of the field before you sit through a demo.

Does it read spread per load directly from the TMS? Not revenue. Not "we can integrate." Ask for the specific connector or API path, and ask whether carrier cost comes through on the same record as customer rate. If a vendor's integration pulls revenue and expects you to load cost separately, you have just bought a reconciliation project.

Can I change the weights myself, tonight, without a support ticket? This is the single most predictive question. The whole value of the instrument is that when the dry van spot market collapses into a rate war, or fuel spikes, or a headhaul lane flips backhaul-heavy, you re-weight the matrix and the desk re-prices the next morning. A tool where weights are a professional-services change request is a tool that will be six weeks behind your market permanently.

Can each broker see their own levels and the exact gap to the next tier? Legibility to the individual is where behavior change actually lives. A scorecard that only managers see is a management report, not a behavior instrument. Ask to see the rep view specifically, not the manager rollup, and check whether it renders on a phone — brokers are not at their desk all day.

Does the composite feed the commission calculation, or does something else? If they are separate systems, ask how disagreements get resolved. If the answer is "manually," price that labor in.

What happens when a broker disputes a load's margin? There will be miscoded accessorials, wrongly attributed loads, and split-credit arguments. A tool with a formal inquiry workflow saves your managers from being the appeals court. A tool without one means every dispute lands in a Slack DM.

Can I run it on one pod first? Pilot on a single desk for a full quarter, including at least one market swing. A margin lift that survives one rate shock is the only evidence worth scaling. If the vendor's minimum is a floor-wide rollout, that is a pricing structure telling you something about their confidence.

Two things to check that buyers routinely skip. First, historical backfill: can you load two years of past loads so the scorecard has a baseline the day you turn it on? A scorecard with no history cannot show trend, and trend is what makes a coaching conversation land. Second, agent-model support: if you run a franchise or agent network rather than W-2 brokers, confirm the tool can handle agents who set their own margin and take a revenue split, because a lot of comp tooling assumes a salaried rep on a quota.

Buyer decision framework

The decision reduces to where the teeth are. Some desks fail because nobody can see the scoreboard; those desks need visibility. Some desks see it fine and ignore it because the paycheck says otherwise; those desks need comp. Buying the wrong one is the expensive mistake.

A few adjacent situations worth naming, because they change the answer.

If your problem is capacity, not negotiation, no scorecard fixes it. A broker who cannot find a truck is not going to negotiate a better rate on a truck that does not exist. Score sourcing depth, but invest in carrier sales and carrier onboarding — the margin lever there is upstream.

If you run a hybrid agent model, the composite still works, but the weights differ per agent tier and you will need the scorecard to support that. Agents on a 70/30 split already feel the margin directly; what they usually need is claims and accessorial discipline, not spread coaching.

If you are LTL-heavy rather than truckload, the whole shape changes: margin lives in class accuracy, reweigh and reclass exposure, and carrier discount tier utilization rather than in per-load negotiation. Same method, entirely different KPI lines. Do not port a truckload matrix onto an LTL desk.

If the constraint is shipper-side pricing, meaning your sell rates are set by contract and the broker genuinely cannot move them, then the only remaining lever is the buy side and accessorial recovery. Re-weight accordingly and stop grading brokers on a number they cannot influence — that is the fastest way to make the whole instrument lose credibility.

Roll it out this way: publish the matrix so the entire desk sees the ranking, run four to six weeks in read-only mode where scores are visible but pay has not changed yet, then flip the comp switch. The read-only window is what converts skeptics, because a strong negotiator visibly out-ranking a high-volume peer does more persuading than any manager speech. Expect pushback in the first two weeks; it is trading the easy dopamine of a high load count for the harder work of protecting spread, and that trade feels bad before it pays.

Related questions

Should margin per load be a percentage or a dollar figure?

Track both. Percentage rewards short cheap freight and penalizes long hauls unfairly; absolute dollars rewards length regardless of efficiency. Weight both lines in the composite so brokers cannot game one at the other's expense, and set the ratio based on your lane mix.

How long before a scorecard changes broker behavior?

Expect four to eight weeks for visible movement once scores are published and pay follows the composite. Run the first four to six weeks in read-only mode so brokers can see where they stand before money is on the line. Full behavior change typically lands within a quarter.

What weight should spread per load carry?

There is no universal answer, but spread is usually the heaviest single line — often two to three times any other. The real discipline is re-weighting when the market shifts: heavier on carrier negotiation in a volatile spot market, heavier on retention when your contract book is exposed.

Does this work for a five-person brokerage?

Yes, and it scales down cleanly. Three to five KPI lines are enough to separate who builds margin from who moves cheap freight. Start in a spreadsheet, argue out the weights, and only buy tooling once the manual upkeep becomes the bottleneck.

How do I score a broker who inherited a bad book?

Grade trend alongside level. A broker taking a 6% book to 11% is outperforming one holding a static 14%. Add a delta column to the composite so improvement is visible, otherwise inherited-book brokers will disengage from the scorecard entirely.

FAQ

What is a weighted multi-KPI scorecard for freight brokers?

It grades each broker across several profit-driving metrics simultaneously — spread per load, rate-per-mile discipline, carrier sourcing depth, accessorial recovery, and shipper retention among them — rather than crowning whoever books the most loads. Each KPI carries a leadership-assigned weight and a 1-to-5 maturity level, and the composite equals the sum of weight times level. The result reflects the true economics of a broker's book rather than raw activity, which is why a high-volume, thin-margin broker can and should rank below a lower-volume negotiator.

How do I set the weights without guessing?

Start from your P&L. Look at where margin actually leaked over the last four quarters — was it buy-side rates, uncollected detention, claims, or churned shippers? — and weight the lines that map to your biggest leak the heaviest. Then sit with leadership and sanity-check that the weights describe the desk you want in twelve months, not just the fire you are fighting today. Revisit quarterly at minimum, and immediately after any significant market shift.

Will brokers push back on moving away from revenue-only scoring?

Some will, especially high-volume brokers who have been top of the leaderboard for years and are about to drop. Two things convert them: publishing the matrix so the grading is fully transparent, and running a read-only period before pay changes. Once a strong negotiator visibly out-earns a high-volume peer in front of the whole desk, the argument ends. The brokers who never convert are usually the ones whose entire method was volume, and that is useful information too.

How often should the weights change?

As often as the freight market forces it — weekly or monthly is normal in a choppy rate environment, quarterly in a stable one. The point is to keep the desk aimed at the current profit driver. Announce every change with a one-line reason, because unexplained weight changes read as moving the goalposts and destroy trust in the instrument faster than anything else.

What do I do about a broker high on volume and low on spread?

The composite drops them below where load count alone would place them, and the low lines point directly at the fix: earlier carrier sourcing, deeper carrier rotation, better lane selection, and actually billing detention instead of eating it. Coach the specific line, not the composite. Because pay follows the composite, volume alone stops being a viable path to top earnings, which usually resolves the behavior faster than the coaching does.

Can conversation data improve margin, or is it just call recording?

It genuinely helps on phone-heavy desks. The margin data tells you a broker's spread is thin; the call data tells you whether they attempted to negotiate the buy rate or accepted the first quote. That distinction determines whether the fix is training or lane assignment. It is a complement to the scorecard, not a replacement, and it is only worth the spend where negotiation actually happens by voice at volume.

Sources

• https://www.fmcsa.dot.gov/registration/become-broker-freight
• https://www.dat.com/industry-trends/trendlines
• https://www.freightwaves.com/
• https://www.jbhunt.com/
• https://www.ttnews.com/
• https://www.bts.gov/topics/freight-transportation
• https://www.tiaonline.org/
• https://www.salesforce.com/products/platform/
• https://www.gong.io/
• https://www.captivateiq.com/

Related on PULSE

• [How Do I Know Where, When, and How Many People to Schedule at Each of My Multi-Unit Retail Locations?](/knowledge/tl0001)
• [How Do I Figure Out How Many People to Schedule Each Day and at What Times for My Single Store?](/knowledge/tl0002)
• [How Do I Know How Many Cooks and Servers to Schedule Each Shift at My Pizza Restaurant?](/knowledge/tl0003)
• [How Many Salespeople Should I Schedule Each Day on My Furniture Store Floor?](/knowledge/tl0004)
• [How Do I Decide How Many Reps to Schedule at Each Store in My Mattress Retail Chain?](/knowledge/tl0005)