6 weeks, $800+ spent, fixed a real bug, still zero purchases — trying to decide between two paths before I burn more budget
Full timeline, no sugarcoating: Weeks 1-5: Ran ATC-optimized campaigns across multiple creative angles, ~$650 spent. Decent engagement throughout (CTR 10-17%, cost-per-click $0.50-1). Got 26 Add-to-Carts. But the site had an actual cart bug during this period (double add-to-carts, broken checkout flow in some sessions), so I don't fully trust that data. Then I fixed the bug, rebuilt the landing page from scratch, and tried Purchase-optimization for 3 days. Day 1 looked great (CTR 3.3%, $0.69 CPC). Days 2-3 CPC spiked to $5-10, so I killed it — reasoning that with zero purchase history ever, I'd never hit Meta's ~50 weekly events to exit learning anyway. Went back to ATC with my two historical best creatives on July 19th. Here's the honest breakdown of that window: - July 19-20 (2 days): still running against the OLD cart logic (a cross-sell item was quietly inflating perceived price ~25%, and the checkout wasn't showing "free shipping" clearly until after address entry) - July 21 (Monday): fixed both issues — removed the confusing cross-sell, made shipping cost/timing explicit upfront - July 21-23 (since the fix, so really only 2-3 clean days): this is the actual test of the "fixed" funnel Cumulative since July 19th: $154 spent, 26 Add-to-Carts, 11 Initiate-Checkouts (33-42% cart-to-checkout, better than I expected), still zero purchases. But only about half of that window is genuinely "post-fix" — so my real clean sample is smaller than the headline numbers suggest. What I've ruled out: checkout technically works (real test purchase completed, Purchase event fires correctly), all major payment methods present, shipping is genuinely free and now clearly shown upfront with delivery time (7-12 business days). What I'm weighing right now, two real paths: Path A: Stay on ATC, let the truly clean-funnel data accumulate a few more days before judging anything (told the real statistical threshold before "structural problem" is 35-40 clean Add-to-Carts — I'm not there yet, especially counting only post-fix days). Path B: Switch to Purchase-optimization now, on the genuinely clean funnel — something I haven't actually tested yet, since my 3-day Purchase attempt happened BEFORE the cart fix. Run it a full 7 days without touching it, accept lower volume for higher intent. Site is tryflowra.com if it helps to see the actual funnel instead of just numbers. If you've run into this exact fork — clean funnel, decent ATC signal, zero purchases — which path would you take and why?