Meraki Wraps.

The brief.

Meraki Wraps sells gifts, gifting accessories, and the kind of wrapping paper you keep instead of throwing away. India-based, mobile-first, D2C. The product worked. The dashboard didn't — and the dashboard was the only signal the founder had to steer by.

The ask was specific: audit the entire Google Ads + analytics stack, find what's actually broken, and rank fixes by impact-per-hour so I can see where the money is going. No glossy retainer. Receipts.

Your headline ROAS is lying to you.

The first thing the audit surfaced wasn't a campaign problem. It was a measurement problem. The account-level ROAS the founder had been quoting in board meetings was inflated by the plumbing, not the performance.

Three reasons it was lying:

• Six non-purchase actions were marked primary. Map directions, "clicks to call," local website visits, smart-campaign engagements — all enabled, all primary, all counted as conversions. Every dashboard pulled from this account inherited the noise.
• Local Store ₹1 walk-in conversions, 8–17 per day, were rolling into the trailing average. Real foot traffic, real value — but priced at 1 INR each, they made the conversion count look healthy while telling you nothing about whether ads were actually selling product online.
• A single March-31 outlier (₹15,409 spend / 31.84× ROAS day) was skewing every trailing window. One day was carrying the story.

Filter to PMax-only and exclude the noise, and the true account ROAS is closer to 2–3× on PMax alone. That's the honest number. Everything that follows starts from that number, not from the dashboard's number.

The smoking gun: GA4 was linked to a dead domain.

Google Ads > Tools > Linked accounts > Google Analytics — the GA4 property linked to the Ads account was www.merakicards.com. The live store is merakiwraps.com. A legacy domain, never unlinked, was where every "GA4 web purchase" signal had been routing for months.

Downstream consequences, observed live:

• Cart Abandoners 30-day audience: 0 members after 6 days populating. Remarketing was structurally dead.
• Checkout Abandoners 14-day audience: 8 members. Not a strategy problem — a wiring problem.
• 4 intent-tag batches (wedding / gift / party / decor) showed 0 products on audit. The Shopify tagsAdd mutations had silently failed because the app token was missing the write_products scope.
• /collections/gift-wraps was serving ribbons. Confirmed by HTML scrape — the SKUs returned were sheer-checkered, sheer-metallic, grosgrain-edge ribbons. Anyone Googling "gift wrapping paper" landed on a ribbon page and bounced.
• Landing-page meta titles were leaking B2B framing into D2C ads. "Wholesale Price." "Supplier in India." Users clicking a gifting ad landed on what read like a wholesale catalog.

The GA4 relink itself is a 3-minute UI walkthrough. That single fix unlocks valid remarketing audiences, valid attribution, valid conversion import. Every downstream growth lever was bottlenecked behind it.

Ranked by impact-per-hour.

Ranked by impact-per-hour. Do 1, 2, 3 first — they unlock everything downstream.

The audit deliverable wasn't a slide deck. It was a dated, prioritized action list — every item with an effort estimate, an impact prediction, and a dependency map. Top of the list:

1. Checkout UX audit. The biggest revenue lever in the account. ATC→Purchase rate sits at 5.7% against a 15–25% benchmark. Over the trailing 90 days that gap is 364 lost purchasers — buyers who hit add-to-cart and never checked out. No ad strategy beats fixing that funnel.
2. Fix LP meta titles. Strip "Wholesale" / "Supplier" framing from collection SEO. Quality Score up, CPC down, relevance up. ~1 hour of work.
3. Fix /collections/gift-wraps. Either correct the product-type filter or split the taxonomy so the gift-wrap keyword universe lands on actual gift-wrap paper. ~1 hour.
4. GA4 relink. 3 minutes in the UI.
5. Demote the 6 junk conversion actions to secondary. Stop polluting Smart Bidding inputs.
6. Add Shopify write_products scope, re-authorize, re-run intent tagging. Then build the smart collections that were supposed to power 4 ad groups.

The discipline isn't novel. The discipline is doing it in that order, with that honesty, instead of starting with "let's launch more campaigns."

What's actually working.

Ad spend is not the story. Ad discipline is.

• PMax-1 → 2.98× ROAS at ₹300/day over the trailing 14 days, with URLs correctly routed to /collections/ribbons-bows and /collections/decor.
• Ribbon Types and Use Cases search campaigns paused after a clean read on zero conversions. Together they had burned ₹2,929 (₹1,978 + ₹951) before being shut. That number is the entire isolated waste — itemized, dated, no hand-waving.
• Brand Defense paused as a deliberate call, not a default.
• 59 negative keywords on Ribbon Types, 57 on Use Cases — verified live against the API, not a status meeting.
• Geo and budget structure verified end-to-end. Where money flows, why it flows, what each rupee is being asked to do.

The point isn't that ₹2,929 of waste was discovered. The point is that ₹2,929 is the whole number, and it's traceable to two specific campaigns on specific dates, because the audit was structured to make waste visible instead of average it away.

Why this work is the reason Sentinel exists.

Meraki's audit is the canonical pattern. A founder who runs a real business with a real product opens her dashboard and sees a number. The number is wrong — not because the math is wrong, but because the plumbing routes through a dead domain, six junk events are flagged primary, a one-day outlier is dragging the trailing window, and a Shopify scope is silently dropping mutations. None of that is visible from the dashboard surface. All of it is visible to an operator who knows where to look.

Sentinel is what we built so a founder doesn't have to wait 90 days and lose 364 purchasers before someone external runs an audit. It's the always-on version of this audit, pointed at Shopify, surfacing the same class of dashboard-lying-to-you patterns the day they appear instead of the quarter they compound.

Your headline ROAS is lying to you. We figured out by how much, line by line, in rupees.

Where the engagement is now.

Live. The audit (dated 2026-04-19, verified against the client's Google Ads and Shopify accounts on 2026-04-20) is converting into execution. Cart Abandoners remarketing launches once GA4 is relinked and the audience populates. A new-customer value rule (1.5×) goes in alongside it. GMC OAuth + feed audit unblock Shopping reactivation. Gift-occasion keyword expansion and a Home Decor PMax campaign are queued behind the funnel fixes — because spending more on traffic before fixing a 5.7% ATC→Purchase rate is how accounts burn money, not how they grow.

If you're running a Shopify store and your dashboard ROAS feels suspicious — read our field notes, look at Sentinel, or book a 30-minute fit call. Tell us the number you don't trust. We'll tell you whether your dashboard is lying.