# Derek Callahan, Director of Marketing at Cartful (78-person B2B SaaS) — read of marketing-budget-allocation-ai, May 28 2026

> 9 years running growth budgets, currently managing about $420K/year across paid search, social, and email, spending way too much time in spreadsheets justifying channel shifts to the CFO.

## How I got here

Got grilled in a board prep call last week about our Meta vs. LinkedIn spend split. I couldn't give a clean answer, just gut + trailing 30-day ROAS. That stung. Later that night on my commute home I Googled "marketing budget allocation software" looking for something like Northbeam or Triple Whale but for allocation decisions specifically, not just attribution. This page came up somewhere around result 8 or 9. Clicked it thinking it was a real product.

## What I clicked first

The feature list actually. "Predictive reallocation: AI suggests budget moves 2 weeks ahead of saturation" and "Churn early warning: Know when a channel is about to tank and reallocate before you notice lower conversions" -- those two hit the exact nerve. That's the problem. I spend 2 weeks reacting to a channel that already died. So I read the whole thing before I understood what I was actually looking at.

## Where I paused

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

I had to read that twice. So... this isn't a product. It's a pitch deck for a product I would have to build? I scrolled back up and re-read "Try it Live" and "Start Free Trial" and felt confused. Did I misread the hero? Because it opens like a SaaS homepage. It closes like a franchise brochure. That's a jarring gear-shift and I sat with it for a while trying to figure out what I was actually being sold.

## What I distrusted

A few things.

First, "AI knows where your next dollar converts best." That sentence is from the generic AI marketing fluff playbook and it tells me nothing. Every attribution tool says something like this. I've seen this sentence reworded a hundred times.

Second, the Fermi estimates. "$-38,512 Year-1 take-home." Negative. So the honest projection is that the person who builds this will lose money their first year. That's buried in a box styled like a credibility feature. It's admirably disclosed, but it also means the page is scoring a 49/100 on its own rubric and showing projected year-one losses, and still trying to sell me the $99 tier. That math doesn't clear for me as a buyer of the idea.

Third, "1 in 7 meaningful-success odds." I appreciate the honesty but I don't understand the methodology at all. Is this based on market data? Internal failure rates? Vibes? "How scoring works" is linked but I didn't click it. It felt like a footnote asking me to trust without explaining.

## What would convince me

If I were actually in the market to build this (and I'm not sure I am), I'd need to see one real conversation. Not a case study, not a testimonial, just a screenshot of a Slack or email thread where someone who used this framework talked to a real potential customer and what that customer said. Even a "here's a cold email we sent and here's the reply rate" would tell me more than a Fermi estimate. The honesty is interesting. Proof that the honesty leads anywhere is what I'd need.

## What I'd ask in an email reply

1. The feature list -- channel health scoring, drag-drop rebalance, multi-platform sync -- is that a spec document or working code? When you say "$99 for the working code starter" what does that mean in terms of actually deployable?

2. You scored pain intensity at 4/10. I feel this pain every week. Why do you think it's low? Is it because marketers have workarounds they're satisfied with, or because the buyer doesn't think AI can solve it?

3. Who is buying the $99 tier, and what's the next step they actually take after they buy it?

## Verdict: on-the-fence

The unusual honesty (49/100, negative year-one projection, explicit "we have no customers") is the only thing that kept me from closing the tab. But I genuinely couldn't tell for the first 60% of the page that I was buying an idea, not a product, and I'm not sure that ambiguity is serving anyone.

---
*Memo by skeptic persona, generated 2026-05-28. Studio breaks own self-grading loop.*