In this solo presentation, Baron Belalov introduces the "Kill Criteria Matrix," a lightweight, 40‑minute framework designed to prevent organizations from continuing to fund ineffective marketing initiatives. He explains that the core problem is not data quality but political – stakeholders argue over what "success" means after results are known. The solution is to agree in writing, before launch, on three outcome tiers – favorable, acceptable, and unfavorable – each linked to a clear, revenue‑connected metric (e.g., ROAS). He provides a concrete example: a 90‑day, $30k‑per‑month paid‑channel test with ROAS thresholds >5× (favorable), 3‑5× (acceptable), <3× (unfavorable). By pre‑signing these tiers, the post‑test decision is automatic, eliminating goal‑post moving. Belalov addresses common objections (how to set thresholds for unknown channels) and offers two answers: articulate the implicit expectation and anchor thresholds to existing channel performance or cash‑payback constraints. He also outlines failure modes – mismatched evaluation windows (e.g., SEO vs. paid), wrong metrics (cost per lead vs. revenue‑linked), and lack of senior sign‑off – and advises how to avoid them. The talk ends with a call to subscribe and a plug for his marketing leadership services.
Chapters
The most useful thing I've ever done for a client's marketing budget cost nothing and took about 40 minutes. I call it the Kill Criteria Matrix. Let me explain how it works. Think about how the argument usually goes. You spend 90 days and some money on something, the results come in, and then a meeting happens where people argue about whether it worked. Marketing says it's early and the trend is positive. Look at the leading indicators. Sales says the leads were just all bad. Finance says the number is what the number is, and everybody is arguing from a different standard because nobody wrote it down. The meeting isn't a disagreement about the data; it's actually a disagreement about what good. was supposed to mean, happening after everybody already knows the outcome, which is actually the worst possible time to have that discussion. So here's the practice: 3 tiers agreed in writing before anything launches: favorable, acceptable, and unfavorable. Favorable means it outperformed, the thesis was right, So increase the allocation. Acceptable means it cleared the bar, so hold and fix the weakest input. Unfavorable means it missed the floor we agreed to, so stop and reallocate. Maybe note what you learned while it's still fresh. That bottom row that you see is the important one. That's the kill criteria, and the whole exercise is really about being Willing to write it down while you're still optimistic. But let me make it concrete because in the abstract this sounds like paperwork. Let's say you're testing a new paid channel. 90 days, $30,000 a month, so $90,000 in total. Favorable, the return on ad spend was above 5x. Acceptable, the return on ad spend was between 3 to 5x. And then unfavorable, the return on ad spend was below 3x. X. Written down on day 0, signed off by whoever controls the budget. Now here's what that buys you. On day 91, nobody is negotiating with the evidence. There's no meeting where somebody argues that 2.5x is basically 3x and the 4th quarter is seasonally weak anyway. The number is the number, the decision was pre-made, and everybody moves on. People mistake this for an analytics exercise, but it actually isn't. It's actually political. What you're removing is the option to redefine success after the fact. That's not a data problem, it's a human one, and it is far and away the most common reason marketing budgets survive for years while producing, you know, very little. It's not incompetence, guys. It's not dishonesty. It's just the very natural human tendency to move the goalposts a little bit when you're standing next to them holding the result. Now the objection I always get: you can't predict what a new channel will do. How am I supposed to set a threshold for something I've never run? 2 answers. First, you're setting one anyway. Every time you approve spend, you have some implicit idea of what would make it worth it. All I'm asking you is to say it out loud where other people can hear it. If you genuinely have no idea what would make it worth $90,000, that's worth knowing before you spend the $90,000. Second, you can anchor it. What is your best existing channel return? What's your current blended average? What payback period does your cash position tolerate? Those give you a floor. It doesn't have to be precise, it just has to exist. A few things that make this fail so you can avoid them: the window might be too short. 90 days is reasonable for most paid channels, but it's completely wrong for content, SEO, or brand, where you'd be killing something at exactly the moment it starts working. Match the window to how the thing actually behaves. The second thing is having the wrong metric. If your unfavorable tier is written in cost per lead, you'll kill a channel that produces expensive leads that close at triple the rate and keep the one producing cheap leads that go nowhere. Set the threshold on something connected to revenue. And finally, if nobody senior signed it, right? If the person who controls the budget didn't agree to the tiers, They'll simply overrule the outcome, and then you've done the paperwork and kept the politics, which is the worst of both. The 40 minutes is the easy part. The hard part is the day you have to honor it on something you personally argued for. I've had to do that, and it isn't fun, but it's also the entire point. If you found this video helpful, please subscribe to my channel, give this video a like, I'm just trying to make marketing leadership more accessible to owners. If you are looking for marketing leadership, go to my website to learn more, baronbelalov.com. And if I can't help you, then I probably know somebody who can.
