# Marcus Webb, Product Manager turned Indie Hacker — read of Service Desk (claude-smb-customer-service), 2026-06-17 > 11 years in B2B product (last stop was Atlassian, mid-size tools org), 18 months now full-time solo. Two kids, 8 and 11. I run 6 miles at 5am before the house wakes up. Trying to find the right second product before my runway gets uncomfortable. --- ## How I got here Someone dropped a link in the MicroConf Slack with a comment like "first idea platform that doesn't just validate everything." I opened it skeptically. I've bought three of these strategy packages from different places in the last year. Two were useless. One had a decent ICP write-up but the "MVP scope" was basically a Notion doc with vibes. So I came in already a little tired. ## What I clicked first The hero pulled me in immediately, then confused me. "AI customer service platform built for SMBs fleeing expensive Salesforce + Fin" is clear positioning. I know exactly who that is. I've talked to those people. But then I realized: this page isn't selling me a customer service platform. It's selling me a strategy package to BUILD one. Those are completely different products and this page is trying to be both at once. I had to scroll halfway down before I understood what I was actually being asked to buy. ## Where I paused The Fermi math section. Specifically "$-23,400 Year-1 take-home." I stopped and read that three times. A product page showing a negative projected income in year one is not something I have seen before. And then right below: "1 in 8 Meaningful-success odds." That's a 12.5% hit rate. They just told me there's an 88% chance this doesn't work out. I don't know if this is genuinely honest or if it's a sophisticated inoculation play -- show you the bad numbers so you trust the good ones, then you buy the $99 kit anyway. But it made me pause in a way that no "10x your support throughput" claim ever has. ## What I distrusted Two things. First, the metrics in the feature blocks are completely uncited. "Reduces first-response time 70%." "Deflects 40% of support volume." Based on what? Where? On whose install? These are round numbers that feel like they were generated, not measured. I've seen enough product pages to know that 70% and 40% are the exact numbers you pick when you're approximating something that sounds credible but isn't tracked yet. Second, and this one is more structural: "pain intensity: 4/10" is listed as a concern, but SMB support chaos is one of the most documented pains in B2B SaaS. If Wishdeal is scoring this a 4, either their scoring model is off or they're saying something meaningful that the rest of the page isn't explaining. I have no idea which. Also: the brand brief and product brief sections in the source are both blank. The page feels thin below the fold. ## What would convince me I want one case study that isn't a founder. Not "we built this, here's the code." I want to see one person who bought the $99 kit, launched, and got their first paying customer within 90 days. Name, MRR number (even $200/mo), what they changed from the kit. That's it. One real operator who is not the studio. The Fermi math showing -$23,400 year one could actually be compelling if they showed their math. What assumptions drove it? Customer count? Churn rate? Acquisition cost? Right now it's a number hanging in space. If I could poke at it, I might trust it. ## What I'd ask in an email reply 1. The page scores "pain intensity" a 4/10. That's lower than I'd expect. What specifically about this market is lower-pain than it looks from the outside? Is it that SMBs aren't actually looking to switch, or that they don't convert once they find a tool? 2. The $99 adopt tier includes "working code starter." What stack, what level of completeness, and how long before a solo developer has something they could show a paying customer? I've been burned by "starter" meaning a repo with a README and three API calls. 3. Have any of the buyers from this idea's dossier or similar ideas come back and said what they actually shipped? Even anonymized, directional signal matters more to me than the Fermi. ## Verdict: on-the-fence The honesty is real enough to keep me here, but the page hasn't resolved its core identity problem -- it markets like it's selling a support tool, then sells a build kit, without bridging the two. If I got a direct answer to question 1 above, I'd probably buy the $5 dossier unlock at minimum. --- *Memo by skeptic persona, generated 2026-06-17T00:00:00. Studio breaks own self-grading loop.*