Ridgeline Outlets

00 · Verdict

The account is profitable, and two campaigns are quietly taking from it.

Across both ad platforms you spent $7,600 and earned $23,950 in revenue over 30 days, a blended 3.15x return. After cost of goods that leaves +$5,574 in contribution margin. FACT

The headline is healthy. Underneath it, Search | Competitor returned $300 on $560 of spend, and Search | NonBrand | Gear sits at 1.65x, just under the 1.82x you need to break even at a 55 percent margin. The second one is the reason this report exists: at 1.65x it looks fine on any return-on-ad-spend dashboard while giving back $81 a month. FACT

01 · Period snapshot

The numbers, both platforms.

Microsoft Ads returned 4.72x on $1,700 of spend against Google Ads at 2.70x on $5,900. The gap is real and it has held for three audits, so the budget question in section 04 is about weight rather than about a single good month. FACT

02 · Profitability

Contribution margin after COGS, by campaign.

Contribution margin = revenue × 0.55 − spend, where 0.55 is your blended gross margin. Break-even is 1.82x return on ad spend.

Search | Brand | TM carries 31 percent of revenue on 14 percent of spend across both platforms. Brand demand is finite, so this is bounded by how many people search for you rather than by budget. FACT

Two campaigns are below break-even. Search | Competitor at 0.54x gives back $395 and has done so for two consecutive periods. Search | NonBrand | Gear at 1.65x gives back $81, which is small enough to survive a dashboard review and steady enough to matter over a year. FACT

03 · Attribution validation

Whether the platform numbers match your store.

Your ad platforms reported 513 conversions this period. Your store recorded 479 orders traceable to paid campaigns. The 7 percent gap is the part worth reading. FACT

Most of it sits in Shopping, where the platform counts a conversion that the order record does not confirm. Every profitability figure above is computed from the store side, so the grades in section 02 already exclude the unmatched conversions rather than inheriting them.

Click identifiers are stripped at checkout on this storefront plan, so the match runs on campaign tags. That is the normal case for stores like yours and it is why the tracking templates matter at setup. INFERRED, storefront plan behavior

04 · Recommendations

Four actions, in the order worth doing them.

Each one carries the size of the move, how long to wait before reading the result, how much of the account it touches, the contribution margin it is expected to move, and the condition that would make it wrong.

Three further findings were held this period because the evidence behind them was not enough to support a call. They stay on file and are re-read at the next audit. INSUFFICIENT DATA

05 · Methodology and confidence

How much to trust this report.

Statistical models and rule-based checks run first and produce the evidence. The plain-English write-up happens last, inside the constraints that evidence has already set, so the wording can change how a finding reads but not what it says.

Every figure above comes from your ad platform data joined to your store's order record, with your gross margin applied. Where a cost cannot be resolved for a product, that line is excluded rather than estimated, because a placeholder margin produces a confident looking number resting on a figure you never supplied.

This is your sixth audit, so the full picture is available: what is normal for this account, what is genuinely unusual, and projected effects with ranges. At the first audit only the current period can be graded.

06 · Data dictionary

What every term here means.

Contribution margin

Revenue times your gross margin, minus ad spend. What the campaign added after the cost of the goods it sold and the cost of the clicks.

Break-even return

The return on ad spend at which a campaign covers its own costs. One divided by your gross margin, so 1.82x at 55 percent.

Grade

Outperform, Profitable, Marginal or Loss, set by contribution margin against break-even rather than by return alone.

Impact

The contribution margin the action is expected to move, computed before the write-up rather than estimated in it. Reads qualitative only, with the reason, when a figure would have to be invented rather than computed. The slot is filled with that reason rather than left blank.

Impr. share

The share of available impressions the campaign won over the period, as the ad platform reported it. Google withholds the exact figure outside the 10 to 90 percent band, so those campaigns read Below 10% or Above 90% rather than a number we would have had to invent. A campaign the platform reported no search impression share for at all reads Not reported, which is not the same as zero and is never shown as zero. Unwon share splits into lost-to-budget and lost-to-rank; a budget increase only buys the first of those.

Confidence

HIGH, MEDIUM, LOW or INSUFFICIENT, set by how much evidence sits behind the finding rather than by how strongly it is worded.

Held

A finding the evidence could not yet support. Shown rather than dropped, and re-read at the next audit.

FACT

Directly measured from your data this period.

INFERRED

Derived from a named basis rather than measured directly.

PROJECTED

An extrapolation, with its assumption stated.

INSUFFICIENT DATA

Below the evidence needed to make the call.