Analyst: Start with the number everyone gets wrong. What did the deck say take rate would be, and where did it settle?
Expert: Deck said fifteen. That was the Series C deck, and I'm fairly sure it got copied into the extension deck after, untouched. Where it actually settled, blended, in my last full year — eleven and a bit. Eleven four. Maybe eleven six. Don't hold me to the decimal.
Analyst: That's gross take rate.
Expert: That's gross. That's the slide number. Take out shipper rebates, which we accrued monthly and always underaccrued, the claims we ate, the detention we chose not to bill because nobody wanted that fight with a top account, and you're nearer nine. Eight nine, nine one. That's the number I'd use if I were you.
Analyst: So why the gap? Was fifteen ever real?
Expert: It was real for a slice of the book. Small shippers. Someone doing four or five loads a month, no procurement function, ringing us Wednesday because their regular guy didn't have a truck for Thursday. We were making eighteen, twenty, sometimes twenty-two points on that freight. Problem is that slice was six percent of gross volume, and the fifteen assumed mix stayed where it sat in year two.
Analyst: Hold on. Six percent of volume — what percent of gross profit?
Expert: More. Fourteen, fifteen. Which is the argument I used to make, and I was wrong about what it meant.
Analyst: Come back to that. So the mix didn't stay.
Expert: It can't. You raise a big round, you get told to grow sixty-odd percent a year, and you cannot grow sixty percent a year selling to people who ship five loads a month. So you go win the big accounts, and big accounts employ people whose whole job is knowing what a fair spread looks like. Our top twenty sat at six to eight. Couple of them at four. Three point something on one, if you count the rebate honestly. I was in the room for that one and I voted for it, because the volume filled lanes where we were thin. That's a real reason. It's just not fifteen percent.
Analyst: So blended take rate falls mechanically as you succeed.
Expert: Yes. And this used to make me want to put my head through a table. Every quarter we'd show cohort take rate, and it was flat. That was true. Blended fell anyway, every quarter, because the mix kept moving underneath it. Both things true at once, and only one of them ever got said out loud.
Analyst: How much of your gross profit sat in the top twenty?
Expert: Top ten was forty-one percent. Top twenty, call it fifty-six, fifty-seven. That's from the March board pack, so I'm confident on it.
Analyst: Let me move you to retention. That's the other place these models hide things. Net revenue retention?
Expert: Headline was a hundred and eighteen.
Analyst: And excluding the top ten accounts?