The Audit Is Graded By The People Being Audited

If you've ever sat through a carrier evaluation, here's a question worth asking: where did the data you used to evaluate the alternatives actually come from?

Most of the time, the honest answer is uncomfortable. The rate cards came from the carrier sales reps. The benchmarks came from comparison tools that aggregate carrier-reported data. The volume profiles came from carrier scorecards built on invoices the carrier cut. The transit time commitments came from carrier service guides.

Even the questions in your RFP were probably written based on what carriers told you to ask.

That's not a carrier evaluation. That's an audit graded by the people being audited.

Why this matters more in 2026 than it did three years ago

For most of the last decade, this didn't really hurt shippers. The parcel market was competitive enough that even if your data was carrier-shaped, the carriers had to compete for your business. UPS and FedEx pushed against each other on price. USPS pulled volume on the low end. Regional carriers were small enough to be opportunistic but not strategic.

That balance is gone.

UPS and FedEx are both reporting record per-parcel revenue while volume drops. They're using long-term contracts with steep early-termination penalties to lock in shippers. They're optimizing networks around higher-margin segments and letting low-margin volume go — which is corporate-speak for "we're choosing which customers to serve well." FedEx publicly announced earlier this year that it's pulling back from general ecommerce volume entirely.

This isn't a market where you can wait six months for an RFP to produce a defensible answer. The carriers are pricing dynamically, contracting aggressively, and pulling levers your team doesn't have visibility into. By the time your evaluation finishes, the conditions you evaluated against have changed.

And the data you used to run the evaluation came from the carriers doing the changing.

What "carrier-shaped data" actually looks like in practice

Three concrete examples:

Rate benchmarks. Most shippers benchmark their rates against industry comparisons or peer data sourced from third parties. But the third parties are aggregating data that originally came from carrier invoices and carrier reporting. The "benchmark" is a smoothed version of what carriers are charging — which means it's a benchmark of carrier pricing power, not a benchmark of what's possible.

Service-level expectations. Your transit time targets are mostly inherited from what your current carriers commit to. If UPS Ground commits to two-day delivery on a lane, your team builds delivery promises around two-day on that lane. You never check whether a regional carrier on the same lane could do it in one — because the question isn't on the form.

Volume distribution. Your team probably has a clear picture of which carriers handle what percentage of your volume. What they likely don't have is a picture of which carriers should handle that volume given your actual ZIP-by-ZIP, weight-by-weight shipment profile. That second question requires a different analysis — one that doesn't take any carrier's word for what they're good at.

In every case, the data exists. It's sitting in your own invoice records. It's just not being used in a way that escapes the carriers' framing.

The fix is structural, not procedural

It would be easy to read this and conclude that the answer is "do a better RFP" or "hire a consultant." Both of those are common responses. Neither of them solves the underlying problem.

A better RFP still asks the carriers to provide the answers. A consultant still builds the analysis on data the carriers shaped, just with a higher hourly rate attached.

The structural fix is different: evaluate carriers using only data that you control, that you generated, that the carriers can't influence after the fact. Your shipment history is the only dataset that meets that bar. It captures every package you've actually shipped, every ZIP you've actually delivered to, every weight and dimension you've actually moved. It's the only honest data in your decision.

Most shippers have this data. Almost none of them have used it as the foundation of a carrier evaluation, because the tools to do that have historically been expensive, slow, and built by — well, the people being audited.

That's changing. There's a small set of approaches emerging that flip the sequence: start with the shipper's data, untouched, and use it to evaluate every available alternative in days, not months. The output is a carrier recommendation built on facts the carriers can't shape — which means for the first time, the audit isn't graded by the people being audited.

What to do in the meantime

If you're a shipper, the most useful thing you can do this quarter is run a quiet inventory of where your evaluation data is actually coming from. Look at every input feeding your carrier conversations and ask: who generated this number? Did it come from us, or did it come from the carrier?

That exercise alone will tell you how much of your strategy you actually own.

I'll be writing more on this over the next few weeks. The shape of carrier strategy is changing fast, and the shippers paying attention right now are setting themselves up for a meaningfully different position by Q3.

— Anthony