# Derek Holcomb, Owner/Operator at DH Resale (Solo + 1 VA) — read of Listing AI, May 8, 2026 > "Eight years flipping on Amazon and eBay, currently running about 180 active SKUs, mostly home goods and vintage kitchen. Use Helium 10 for keyword research, InventoryLab for accounting. Coach my daughter's U10 soccer team on Saturday mornings, which means Friday nights I'm usually listing 30 products instead of watching TV." --- ## How I got here Watched a YouTube short from a guy I follow in the FBA space. He mentioned it offhand, not a sponsored segment, just "been using this lately." That detail matters to me: unpaid mentions from people who source for a living carry more weight than any ad. I Googled "ai listing tool amazon ebay" right after and this was the third result. Landed on the homepage directly, no landing page funnel. ## What I clicked first The before/after block. I scrolled past the hero and went straight to it. "Handmade ceramic mug, blue. A nice mug made by hand. Good for coffee or tea." That's painfully accurate. That is exactly what my listings looked like in year one. The after version is genuinely good. "Holds heat longer than thin porcelain, fits under standard coffee makers" is the kind of objection-handling copy that takes experience to know buyers even care about. I stopped on that section for probably two minutes. The hero claim that got my attention: "4 min saved per listing vs. writing from scratch." That's credible. Not "10x your sales." Not "the last listing tool you'll ever need." Four minutes is a number I can actually verify in my own operation. ## Where I paused The A/B listing variants feature. Page says: "Generate two versions with one click, test them in your storefront, and let your sales data decide." And then: "Sellers using A/B variants see an average 28% lift in conversion over single-version testing." That's interesting because Amazon does not let you run a true A/B listing test unless you are in Manage Your Experiments, which requires brand registry. eBay definitely does not have this. Etsy does not either. So what does "test them in your storefront" actually mean in practice? Run version A for two weeks, switch to version B? That is not an A/B test. That is a time-series comparison with zero controls. The claim of 28% lift is doing a lot of work on a methodology that I am not sure holds up. Priya K.'s testimonial gets at this too: "Two versions, post both, check the data after two weeks, keep the winner." Post both where? Two separate listings on the same platform? Duplicate listings on Amazon get suppressed. There is a workflow gap here that the page is skating past. ## What I distrusted Three things. First, the "47% average improvement in search click-through rate." That is a very specific number with zero sourcing. No "based on X sellers over Y weeks." No methodology note. Improvement over what baseline? Their own prior listings? Industry average? I have been around enough SaaS marketing to know that number comes from somewhere convenient. Second, the testimonials. "Sarah M. Etsy seller, handmade ceramics, 3 years." No shop name. No way to look her up. "Marcus T. Amazon FBA operator, 400+ active SKUs." Same. I am not saying these are fake. I am saying I cannot verify them, and any product with 12,400 active sellers should be able to surface one person willing to put their actual Etsy shop URL next to a quote. Third: "Views-to-sales analytics" on the Scale plan. How? Are they pulling data from platform APIs? Do I have to connect my seller accounts? If so, what permissions does that require and what data are they touching? The privacy FAQ says "we do not sell your product data" but does not say anything about what happens when you connect a marketplace account. That section is doing work the page is not explaining. ## What would convince me One real seller I can look up. Not "Sarah M., ceramics." A live Etsy shop link or an Amazon storefront where I can see the listings and verify the copy style. Even one. The A/B methodology explained plainly. If it is "write version A, swap to version B after two weeks, compare" just say that. I will accept that workflow. What I will not accept is calling it an A/B test when it is a before/after comparison, because that is the kind of thing that makes me trust the rest of the claims less. And I want to know whether the keyword data is live-pulled (like Helium 10 pulls from actual search volume estimates) or baked into the model at training time. If it is the latter, those "highest-volume search terms" could be six months stale. That matters a lot in trending categories. ## What I'd ask in an email reply 1. How does the A/B variant workflow actually function on Amazon for sellers without Brand Registry access? Is there a workaround you recommend, or is that feature primarily for Etsy and Shopify? 2. Where does the keyword data come from? Is it a live integration with any keyword research tool, or is it derived from the model's training data? How often does it update? 3. The "47% improvement in click-through rate" stat: can you share the methodology? How many sellers, what time window, and what were they comparing against? ## Verdict: curious-enough-to-reply The before/after copy alone is enough for me to send an email. But I would not pay $49 before getting a real answer on the keyword data freshness, because that is the core of the product promise and the page does not explain it once. --- *Memo by skeptic persona, generated 2026-05-08. Studio breaks own self-grading loop.*