Niche
B2B industrial / metal supply (edge case)
Sherman TX. Metal fabrication, buildings, architectural metal. B2B + DTC industrial supply since 1996.
Recommendation: GO (Donal override 30 Apr - run ads in parallel to quote-followup work)PARTIAL. B2B industrial supply / metal fabrication is not explicitly listed in OpenAI's allowed verticals. Closest match is "professional services" (allowed) or "ecom non-regulated" (allowed). Most likely allowed in practice but worth verifying via OpenAI advertiser support before live spend - the product mix (metal buildings, architectural panels, sheet metal) is dual-use B2B + DTC which adds ambiguity.
2/4 ICP fit. The big question: do MW's customers (contractors, handymen, farmers, homeowners building metal buildings) use ChatGPT to research metal panel suppliers? Likely some, but the buying behaviour is more often phone-driven (per profile: "Most leads come via phone"). ChatGPT free-tier user demographic skews more consumer than tradesperson, so the ICP overlap is thinner than for movers/roofers.
| Check | Result | Note |
|---|---|---|
| OAIQ SDK installed | FAIL | Not installed. |
| Page loads under 2.5s LCP | VERIFY | WordPress on GoDaddy varies. |
| Form / phone CTA above the fold | PASS | Above-fold form (Name/Email/Phone/Zip/Interest/Message), "Get Free Estimate" + "Click to Start Designing your Metal Building" CTAs, phone (903) 465-6699 clickable. |
| Trust signals | PARTIAL | "Since 1996" + "Best of Texoma" badge. No BBB rating visible. Reviews section header exists but content missing - render bug to fix. |
2-3/4 landing page score.
legalName: "anuj" bug as Lankford. Both sites share this. Must be fixed via Tanatsa + IT team conversation.Inputs: $5,000 monthly budget, 5% form-fill rate (B2B sites convert lower than consumer), 20% close rate (longer B2B sales cycle, lower conversion), $5,000 average ticket (small order; metal buildings push up to $20-50k+ but represent a slice of mix), 25% gross margin.
| Scenario | CPC | Clicks | Leads | CPL | Customers | CAC | Gross profit | Net ROI |
|---|---|---|---|---|---|---|---|---|
| Best case | $3.00 | 1,667 | 83 | $60.00 | 17 | $300 | $21,250 | 3.25x |
| Mid case | $4.00 | 1,250 | 63 | $80.00 | 13 | $400 | $16,250 | 2.25x |
| Worst case | $5.00 | 1,000 | 50 | $100.00 | 10 | $500 | $12,500 | 1.50x |
Break-even CPL: $250 (need ~4 customers/month at $5k ticket and 25% margin). Worst-case CPL of $100 still under.
MW's stated near-term priority is fixing the quote-follow-up gap via Paradigm ERP DataHub integration. Tanatsa flagged this as a known revenue leak: "definitely some quotes that we get out there that we don't end up closing because we're not following up enough."
OpenAI Ads adds inbound lead volume. But MW already converts inbound at a lower rate than they should because of the follow-up gap. Adding more leads on top of a leaky bucket is the wrong order of operations. Fix the bucket first.
Donal's call: "Fine to pump him with leads too." Lead volume and quote-follow-up are not strictly sequential - more inbound while OO ships the Paradigm DataHub automation just means more opportunity for the automation to prove ROI on day one of going live. Worst-case scenario is the same as MW already runs (some quotes leak), best case is the automation lands during the ad ramp and converts the new volume cleanly.
Reasoning: 2/4 ICP fit is the only real concern (B2B audience overlap with ChatGPT free-tier is unproven for this niche). ROI math is positive in all 3 scenarios. Strong above-fold form. Strong vertical positioning ("Metal Buildings. Built Right. Priced Right." - the headline reads like a direct response to a ChatGPT-research-stage prompt).
Conditions: fix the legalName: "anuj" schema bug FIRST (shared with Lankford); install OAIQ SDK; verify ticket mix with Tanatsa + Chase to refine simulator; track close-rate carefully in first 30 days to validate the B2B ICP holds; ship quote-follow-up automation in parallel so it lands before ad volume scales above $5k/mo.