# Marcus Okafor, VP of Sales at Cargomatic -- read of Sales Training AI, May 11 2026

> 11 years in B2B SaaS sales, currently managing an 8-person inside sales team at a 55-person logistics software company. We run Aircall, Gong, Salesforce, and Outreach. My biggest recurring headache: I hire decent reps, then watch them flounder for five months while I pray the pipeline doesn't crater.

## How I got here

Searched "AI call coaching for small sales teams" on Google because I was waiting for my daughter's soccer practice to end and I had 20 minutes to kill. This came up third or fourth. The headline -- "your reps fumble more than they practice" -- was blunt enough that I clicked instead of scrolling past.

## What I clicked first

The hero line got me: "Your best closer could train others, but instead she's booked solid taking accounts herself." That's not a made-up problem. That is Tuesday at my company. Literally what happened last week when I tried to shadow Kezia on two calls and she had to cancel one because an existing account called in hot.

So I kept reading.

## Where I paused

The pricing section. No per-rep licensing is genuinely interesting. Most tools in this space (looking at you, Gong) will nickel-and-dime you the second you hire rep number six. "$2,499 for up to 20 reps" is a number I can actually defend in a budget conversation without a spreadsheet. I stopped and re-read that section twice.

Also paused at "Calls are transcribed locally. Your recordings never leave your infrastructure." That's a sentence that would matter to our legal counsel, who blocked us from using one competitor last year over data residency concerns.

## What I distrusted

Three things, in order of how bad they are.

First: the stats. "40% Faster onboarding." "3.2x More reps reaching full productivity in year one." "18% Average close rate improvement in month three." These are not sourced. There's no "based on a study of X customers" or even a case study name attached. I've been sold software with stats like these before. They never match reality. They feel assembled, not measured.

Second: "9 out of 10 reps prefer real-time guidance to post-call reviews." Whose survey? Their own beta? A panel? My reps have complained to me that Gong's post-call summaries feel surveillance-y. Real-time ear coaching from an AI is a bigger ask, not a smaller one. That 90% number needs a source or it's meaningless.

Third -- and this is the big one -- I scrolled all the way to the bottom of the page and found this:

> "Honest disclosure: we don't have live customers on this idea yet. We shipped the strategy package; you ship the customer conversations."

So this is not a product. This is a product studio's research package for sale. The page I just read -- with the demo request buttons, the pricing tiers, the FAQ answering "how long does setup take?" (fifteen minutes!) -- is marketing copy for a product that does not exist yet. The whole thing is a pitch deck dressed as a landing page. The stats I distrusted in point one are especially hollow now.

## What would convince me

If the product actually existed and had customers, I'd want one thing: a before-and-after ramp curve from one real company. Not a percentage. A graph. "Team of 7 reps, average time to first closed deal went from 14 weeks to 9 weeks, here are the rep names and they'll take a 15-minute reference call." That's it. That's the whole ask.

The compliance claim -- local transcription, never leaves your infrastructure -- also needs documentation. A one-pager I can hand to legal, not a bullet point on a landing page.

## What I'd ask in an email reply

1. The page says setup is 15 minutes and the first coached call happens the same day. How does playbook training work in that window? Am I just describing my process in a text field and the AI improvises, or is there actually a human onboarding session where someone helps me define what a good call looks like?

2. We use Gong already for call recording and post-call review. Would this replace Gong, sit next to it, or conflict with it? The page lists dialer integrations but doesn't mention existing call intelligence tools.

3. The page says "depending on your location, we manage consent, notification, and recording requirements." In California, we have specific two-party consent obligations. Has this been tested in a California inside sales environment or am I the first one?

## Verdict: dismissive

Not because the idea is bad -- the idea is genuinely good and the pain is real. But this is a product studio selling research on an unbuilt product, and the page is presenting itself as a live SaaS with pricing and a demo request flow. The moment I read "we don't have live customers on this idea yet," any trust built by the hero copy evaporated. If they'd led with "we're building this, here's the concept, want to be a design partner?" I might have replied. The bait-and-switch framing is the problem, not the idea.

---
*Memo by skeptic persona, generated 2026-05-11. Studio breaks own self-grading loop.*