Three hundred episodes in, and the most useful thing Matt has learned isn't a tactic. It's that the question we all use to filter ideas doesn't filter anything at all. In May, on episode 289, Matt committed on air to a 90-day Instagram experiment, personal brand against a paid-ads benchmark, numbers published either way. This is the honest report. The benchmark never got run, the reels never hit cadence, and the two-to-three hours a week he'd allowed for it were never once spent. So it wasn't time, and he argues it wasn't discipline either. What it was is a cost he never budgeted for, and a coaching question he's asked of a thousand other founders and failed to ask himself. This is also the last episode ever recorded in the home studio, the Edmundsons move house in a week, after 22 years. In this episode 01:58 - The 90-day Instagram experiment, and what actually happened06:17 - Origination cost, the expensive part was never the filming09:27 - Entrepreneurial optimism, and the two coaching questions12:33 - Why a notebook full of good ideas creates overwhelm17:21 - Why 8 out of 10, and why "consistent" is the word that matters22:51 - Does it fail because of you, or because of the task?25:51 - The conversation that lifted mobile conversion over 400%31:58 - When the guest is right and you don't want them to be36:01 - What AI actually changed39:16 - How this podcast started, and what 300 episodes have been worth46:12 - Should you start a podcast? The Promise That Didn't Get Kept (01:58)The setup on episode 289 was a proper argument. Davie Fogarty, the Oodie founder doing around $200m a year, reckons founders under $10m shouldn't bother with personal brand, the opportunity cost is enormous and that time is better spent on ads creative. Alex Hormozi and Daniel Priestley say the opposite. Show up on camera or AI eats your business. Matt said he'd settle it. Ninety days, head to head, numbers published. "The ads benchmark never actually got run. Not that it underperformed, it never actually happened. So I can't give you the comparison that I promised you, which I can only apologise for."Origination Cost (06:17)The filming was never the problem. Matt has three studios and, in his words, more iPhones than he knows what to do with. The expensive part was deciding what to say and then getting in front of a camera cold, from a blank page, every single time. The one reel that did work, building an AI assistant like KITT from Knight Rider rather than Jarvis, hit eight or nine thousand views, the best on his channel. And he didn't script it. Hook written word for word, landing written word for word, prompts in the middle, everything else ad-libbed. The Two Questions (09:27)"Will this move the needle between where I am and where I want to be? And can I be a consistent 8 out of 10 at doing this?"Two questions Matt has put to a hundred, maybe a thousand founders over the years. He broke the second one in public, on his own podcast, and the thing that got him there he calls entrepreneurial optimism. The blind confidence that we can just do the thing. Your Notebook Isn't Full of Bad Ideas (12:33)Matt takes pages of notes on every episode, conference, book and coaching call. It's a genuine gold mine. It's also, he suspects, a library of good intentions. "The problem isn't that the ideas themselves are bad. The problem is that every single idea written in my notebook is good."Which is exactly why question one filters nothing. Everything in the notebook passed it, that's the only reason it got written down. Why 8, and Why Consistent (17:21)Ten out of ten is a fantasy, and chasing it means never starting. Eight is where things produce real results and where Matt and the team can still sustain them on a Thursday afternoon when everyone's tired. "Ten out of ten is perfectionism wearing a business suit."The word doing the real work is consistent. Hitting an 8 once is a good afternoon. Week nine is the test. And a brilliant idea managed at 4 out of 10 sporadically is worth close to nothing, arguably less, because of what it eats. "It could slowly improve your guilt score. But that's a KPI not worth measuring."There have been three or four points across seven years where Matt seriously considered stopping the show. He didn't, and that consistency is the entire argument. Drop, Delegate, or Systematise (22:51)When something fails question two, there's one more question. Does it fail because of you, or because of the task? If it's the task, drop it and feel good about it. If it's you, work out which steps genuinely need you and which you've only assumed do. With the reels, the honest answer was thirty seconds of face on camera, the ideas, the beat sheets, the edit and the scheduling never had to be Matt's. Three exits. Only one of them is a no. What It Looks Like When It Works (25:51)Episode 280, March this year. Adam Pearce of Blend Commerce on the first three thumb scrolls of a mobile product page, you get about three before somebody buys or leaves, and most of us waste them. Across his client base he was seeing 30-50% conversion lifts from fixing them. Question one, obviously yes. Question two, no. Matt isn't a CRO specialist and won't become one. Had the plan been "Matt learns mobile UX," it would have died in week three. So they changed who does the doing. The conversation became research, then a Claude Code tool that audits the mobile experience and tells the developers how to fix what it finds, then a three-phase implementation plan. "We have worked through two phases of that three-phase plan. As things currently stand, our mobile conversion is up by over 400%."Capacity, Not Appetite (31:58)A warning, because the filter can be abused. Back on episode 35 in October 2020, Chloe Thomas told Matt to get moving on Black Friday. He agreed out loud and dismissed it internally, they'd done this before, they knew what they were doing, there was time. There was not time. They've been early every year since. Question two is about whether you can sustain something, not whether you fancy it. "I could do this at 8 out of 10 every week, I just don't want to" isn't a fail. That's a Tuesday afternoon. What AI Actually Changed (36:01)The gap between an interesting idea and a decision you can act on is enormous, because answering question two honestly means holding the whole business in your head at once. SAM closes that gap, those conversations run from fifteen minutes to two or three hours of genuine pushback. "It doesn't do the thing for you. It doesn't get me in front of a camera. But it does close the gap between me doing it and not doing it."Where It All Started (39:16)Matt's best friend Tony told him he had a great face for radio. That was, more or less, the strategy. They already had the kit from a 2012 podcast run for a beauty business, which made hundreds of thousands of pounds in ways nobody predicted. Season one of EP was mostly Matt talking, until he got bored of his own voice and switched to interviews. He has never scripted an episode or his questions since. And the most listened-to episode of all 300 is still episode one, the one he made before he'd learned anything. There was never any sponsorship. Measured that way, the show is a failure. Measured on seven years of weekly hours with people who know things he doesn't, it's worth millions, plus the part nobody plans for. Jared Mitchell (episode 74) and his family have stayed at Matt's house, and Matt and his daughter have stayed at theirs. Chris George (episode 130) led to recording live at SubSummit and doors that wouldn't have opened otherwise. Should You Start One? (46:12)Yes, with one condition, which by then you can guess. Only if you can be consistent at it. Ignore the download numbers and have great conversations with interesting people. "This podcast is definitely not the best idea that I had in 2019. Not even close. It's just that this particular one I have been able to do at a consistent 8 out of 10 for seven years, week after week. And that, I think, is the entire reason it has worked."Your TurnGo and find your list, the notebook, the doc, the Slack channel, wherever the graveyard is. Run it through both questions. Where something fails the second, ask whether it's you or the task before you delete it. Most people end up with two or three things. And that tends to feel like an enormous relief. Download The 8/10 Filter, one page, free, at ecommercepodcast.net under the resources link.
