The July Report

The month in numbers

Both channels, one ledger. Every figure carries three comparisons — last month, last July, and the January 2025 baseline this account started from.

The tills had their second-best month of the year. The ad account had its most expensive. Those two facts are the whole report.

The lift panel

One table, three horizons. MoM is momentum, YoY is the real trend once seasonality is stripped, and vs Jan ’25 is the whole transformation. Green is good — including when the number falls, for cost metrics.

Bar length scales with the size of the move (capped at 150%). Revenue-per-rupee and cost-per-order have no comparable Jan ’25 basis — Meta spend wasn’t tracked then.

MQL → SQL, month by month

Marketing (MQL) — what Google Ads bought — against sales (SQL) — what the tills recorded. June’s bill count has now landed and is filled in; July is restated to the exact campaign-report totals.

MQLs = Google Ads conversions · SQLs = billed orders. Cost/order and Rev/₹ use Google spend only so the full series stays comparable — July’s all-in figures including Meta are ₹54 and 49×.

The nineteen-month arc

The same account, transformed. Clicks cost ₹14.23 when this run began; they cost ₹4.70 now — but that number went the wrong way this month for the first time since October. The gold dot is July.

The efficiency curve

The question every extra rupee has to answer: did it buy anything? Every 2026 month plotted against the account’s own response curve, and what each additional MQL actually cost.

Since March, spend is up 45% and MQLs are up 4%. Those 821 extra leads cost ₹164 each — roughly 7 times the ₹21.91 the account reports as its average.

What the spend is actually worth

Blended ROAS flatters everyone: it credits paid media with revenue the salons would have taken anyway. Regressing 19 months of orders against MQLs separates the two.

Orders = 6,984 + 0.1385 × MQLs, fitted on Jan ’25 – Jul ’26. Incremental revenue = 2,745 orders × ₹2,655 average bill = ₹72.89L against ₹5.47L invested. A blunt model — it attributes nothing to Meta and treats the baseline as fixed — but it is directionally honest in a way blended ROAS is not.

Meta: the league table

₹1.13L bought 2,249 leads at ₹50 — 49% more leads than June at a flat blended CPL. But the average hides a network split almost exactly in half.

Seven salons take 47% of the Meta budget and return 71% of the leads. Seven others take 38% and return 20%.

Quadrants split at the network medians — ₹69 CPL and 102 leads. Lavelle Road ran no Meta campaign in July, as in June. Frequency above 3 (HRBR Kalyan Nagar, 3.7) means the same people are seeing the same creative repeatedly.

August, three ways

Built from the account’s own response curve, the orders model, and 2025’s August seasonality (orders +10.2% on July, average bill +5.7%). August is structurally a good month — the question is only how much it costs to get there.

Assumptions: MQLs from the fitted 2026 response curve, discounted 9% to match July’s actual underperformance. Orders = 6,984 + 0.1385 × MQLs, × 1.102 August seasonality. Average bill ₹2,805 (July’s ₹2,655 × the 2025 Jul→Aug ratio). Meta leads are not in the orders model, so all three revenue lines understate the reallocation case.