A full diagnostic of both Google Ads accounts, measured against Triple Whale as the source of truth, benchmarked against the competitive field — with a prioritized plan to protect today's revenue and unlock the next stage of growth.
Ledisa is a ~$4.9M/month business (90-day run rate, Triple Whale) built on one dominant acquisition channel. Google Ads is the second-largest channel and — per Triple Whale — the most efficient paid dollar in the mix. But it is being throttled by its own measurement, it carries real ad-policy exposure, and the whole category is under regulatory pressure. The plan: fix tracking, defuse policy risk, then scale.
21 active ads limited for Clickbait, 2 disapproved with Prescription Drug / Restricted Drug Terms flags, and 92 of 98 active Search ads land on an advertorial domain — in a category where the FDA has issued 85 warning letters about GLP-1 marketing. A policy sweep could turn the account off overnight. Fixing this comes first. Details on the Search page →
Google's tag captures only ~42% of the orders Triple Whale credits to Google (13% on brand search). Smart Bidding sees half the truth, bids timidly, and gives up 64% of impression share on rank — while Triple Whale measures Google's non-brand CPA at $31.49 against a $60 target. Fixing tracking is the single biggest growth lever. How we measured this →
AppLovin took $6.62M of the $8.56M 90-day budget and demonstrably creates brand demand (its daily spend moves together with brand searches, r = 0.68). It is also the least efficient channel on last-click (0.77). Part of that is by design — high order volume that may include subscription starts — but the concentration itself is a business risk. Demand Gen page →
The only attempt spent $1,297 over 14 days (Nov 2025) — 7 of 8 campaigns never served an impression. There is no evidence for or against the channel. We propose a real, measurable test plus two full campaign builds. Proposals →
Kind, Gentle and Patched have all repositioned to berberine-first naming. The FDA states there are no approved GLP-1 patches, and a consumer law firm is running ads to recruit purchasers of dopamine patches — a product Ledisa sells. Compliance repositioning is brand insurance for both ad accounts. Full competitive analysis →
Disciplined exact-match build (75.6% of spend), 84% of keywords at Quality Score 7+, clean single-purchase conversion setup, zero product-feed waste, and brand campaigns returning $11.28 per dollar on Triple Whale last-click. This program is throttled, not broken.
Every business metric below comes from Triple Whale. Google Ads platform numbers are used only for auction mechanics — impression share, Quality Score, bids and budgets.
Last click gives the whole order to the final ad the customer clicked before buying — it rewards channels that close sales. Linear splits the order value evenly across every ad touch in the customer's journey — it rewards channels that participate early. When a channel's linear number is much higher than its last-click number (AppLovin: $1.24M vs $0.80M), that channel is starting journeys other channels finish. Google scores well on both — it both creates and closes.
On deduplicated Triple Whale attribution, Google returns 2.05 (last click) to 2.46 (linear) per dollar — versus AppLovin at 0.77–1.20, Meta at 1.20–1.64 and Taboola at 0.71–1.93. The mix isn't "wrong" — a demand-creation channel shouldn't be judged on last click alone — but the efficiency gap plus the 77% concentration argues for shifting marginal dollars toward Google capture and building a second demand-creation channel (Demand Gen page).
Ledisa's economics look like a subscription/replenishment business: a higher first-order CPA is acceptable when the same customer keeps paying month after month. That changes how each channel should be judged.
We checked. Triple Whale's subscription metrics come from its Recharge integration, which shows only 2 active subscriptions — clearly not the real recurring base for a store with 52% returning-customer orders. Recurring billing almost certainly runs through the external checkout (CheckoutChamp), outside Shopify's subscription apps, so Triple Whale can't see it as "subscriptions" — only as repeat orders. Recommendation: export subscription status from CheckoutChamp (or tag orders in Shopify) so new-vs-recurring revenue can be split per channel. Until then, the honest proxies are the returning-customer share and LTV/CPA above.
AppLovin's low last-click return (0.77) is partially defensible if its orders are disproportionately new-customer subscription starts that compound over the following months. The same logic applies to scaling Google: a $60 target CPA against a $115 new-customer CPA and a 2.57 LTV/CPA ratio suggests there is headroom to bid for growth once tracking is repaired — the business earns back acquisition cost over time. Verifying the subscription mix per channel (above) turns this from a hypothesis into a budget rule.
| Setting | Search account | PMax & Shopping account | Verdict |
|---|---|---|---|
| Auto-apply recommendations | 0 enabled | 0 enabled | Clean — Google is not auto-changing anything |
| Auto-tagging (GCLID) | On | On | Correct — the click ID exists; it gets lost later in the landing-page hop (see Search page) |
| Conversion actions | 1 primary: Purchase (server-side GTM), data-driven attribution | 1 primary: Purchase (server-side GTM) | Clean single-primary setup; no duplicate counting |
| Optimization score | 79.7% | 70.4% | Informational only — do not chase this metric |
| Enhanced Conversions | Not detected on the purchase tag | Gap — hashed-email matching would recover part of the lost conversions | |