Open Ads Manager and you get forty columns of numbers: ROAS, CTR, CPC, CPM, frequency, relevance diagnostics, video watches, outbound clicks. Most store owners stare at this wall of data, pick the one number that looks good, and make decisions on it. That is how profitable campaigns get killed and losing campaigns get scaled.
You need eight metrics. This post defines each one, shows you the PKR math behind it, adjusts everything for COD and RTO realities, covers WhatsApp attribution that most guides ignore, and tells you which popular metrics to stop looking at entirely.
The 8 metrics that actually matter
1. ROAS: return on ad spend
ROAS is revenue divided by ad spend. Spend PKR 10,000, generate PKR 40,000 in attributed revenue, and your ROAS is 4.0. It is the headline efficiency number, and it is also the most misunderstood, because Ads Manager’s ROAS uses raw attributed revenue with no adjustment for returns, delivery failures, or WhatsApp orders it never saw.
Always read ROAS alongside your real numbers. We will fix its blind spots below with COD-adjusted ROAS and blended MER.
2. CPA: cost per acquisition
CPA, shown as cost per purchase in Ads Manager, is spend divided by purchases. Spend PKR 10,000 for 14 purchases and your CPA is PKR 714. For Pakistani COD stores, CPA is often more actionable than ROAS, because your costs are per-order: courier charges, packaging, and RTO losses all scale with orders, not revenue. Know your target CPA before you launch: for most stores it sits at 25 to 35 percent of average order value.
3. CTR (link): click-through rate
Link CTR is the percentage of impressions that produced a click to your site. This is your creative report card. Most Pakistani ecommerce prospecting campaigns land between 1 and 2 percent. Below 0.7 percent after a few thousand impressions, the creative or hook is failing. Above 2.5 percent, you have a winner worth scaling. CTR tells you about the ad. Everything after the click tells you about the store.
4. CPC: cost per click
CPC is spend divided by link clicks. Healthy Pakistani prospecting CPCs typically fall between PKR 30 and 80. CPC is mostly a function of your CTR and the auction: better creative earns cheaper clicks. Watch CPC as an early warning, but never optimize for cheap clicks alone. The cheapest clicks in Pakistan often come from curiosity traffic that never buys.
5. CVR: conversion rate
Conversion rate is purchases divided by link clicks, and it measures your store, not your ads. Typical Pakistani Shopify stores convert 1 to 3 percent of ad clicks into orders. Below 1 percent with healthy CTR, your product page, pricing, or trust signals are the bottleneck. Fix the store before you blame the ads. Above 3 percent, your funnel is strong and the constraint is traffic volume.
6. CPM: cost per thousand impressions
CPM measures auction competition for your audience. Normal Pakistani ecommerce CPMs run PKR 400 to 900, spiking 30 to 60 percent higher during Ramadan and the November sale season. CPM is context, not a goal. A rising CPM with steady CPA means the auction got hotter but your funnel absorbed it. A rising CPM with rising CPA means you are being priced out.
7. AOV: average order value
AOV is revenue divided by orders, and it is the lever most stores ignore. Raising AOV from PKR 2,500 to PKR 3,200 with bundles and free-shipping thresholds transforms your allowable CPA overnight. Every metric above gets easier when AOV rises. Track it weekly; a falling AOV silently erodes ROAS even when everything else looks stable.
8. MER: marketing efficiency ratio
MER is total store revenue divided by total ad spend, across all channels. If your store did PKR 500,000 in revenue last month and you spent PKR 100,000 on all ads, your MER is 5.0. MER catches what ROAS misses: WhatsApp orders, organic spillover, and returning customers that ads assisted. When Ads Manager ROAS says 2.5 but your bank account disagrees, MER tells the truth.
The 8 metrics that matter: what each tells you and the healthy range for Pakistani stores
| Metric | What it tells you | Healthy range |
|---|---|---|
| ROAS | Revenue per rupee of ad spend | 3 – 5 established; 1.5 – 2.5 new stores in month one |
| CPA | What each order costs you | 25 – 35% of AOV, judged on delivered orders |
| Link CTR | Whether your creative works | 1 – 2%; below 0.7% is a creative problem; above 2.5% is a winner |
| CPC | Cost of each click | PKR 30 – 80; above PKR 120 needs investigation |
| CVR | Whether your store converts | 1 – 3% of clicks; below 1% is a store problem |
| CPM | Auction competition for your audience | PKR 400 – 900; 30 – 60% higher in Ramadan and November |
| AOV | Order size, your biggest lever | Track weekly; bundles and free-shipping thresholds raise it |
| MER | True efficiency, including WhatsApp and offline | Total store revenue / total ad spend; catches what ROAS misses |
COD-adjusted ROAS: the number that tells the truth
Here is where Pakistani stores must diverge from every foreign metrics guide. Ads Manager counts an order when the purchase event fires. It does not know that 25 percent of your COD orders will be refused at the door.
The RTO adjustment
Take a real example. Your ads generate 100 orders at PKR 3,000 AOV. Ads Manager shows PKR 300,000 in revenue against PKR 75,000 spend: a 4.0 ROAS. But your RTO rate is 25 percent, so only 75 orders deliver. True delivered revenue is PKR 225,000, and your COD-adjusted ROAS is 3.0. That is a full point of ROAS that never existed.
It gets worse. Those 25 returned orders still cost you courier charges both ways, typically PKR 250 to 400 per round trip. At PKR 300 per return, that is PKR 7,500 in dead courier cost. Your real economics: PKR 225,000 delivered revenue against PKR 82,500 in total cost, which is a 2.7 true ROAS.
Build your own adjustment
Every month, compute: delivered orders divided by total orders gives your delivery rate. Multiply Ads Manager revenue by that rate for adjusted revenue. Add return courier costs to your ad spend for true cost. Divide. That is the number to judge campaigns against. A campaign at 2.5 reported ROAS with 20 percent RTO is really at 2.0, which may be below your break-even. Our break-even ROAS guide for COD returns walks through the full calculation with your margins.
WhatsApp conversation attribution
A large share of Pakistani ecommerce sales starts as a WhatsApp chat, not a website checkout. Click-to-WhatsApp ads, product-page chat buttons, and COD confirmation flows all generate orders that Ads Manager undervalues or misses completely. If you ignore this channel in your reporting, you will kill ads that are actually working.
The two WhatsApp metrics
First, cost per conversation started: ad spend divided by WhatsApp conversations initiated. Most Pakistani stores see PKR 80 to 200 per conversation for click-to-WhatsApp campaigns. Second, conversation-to-order rate: what share of conversations becomes a paid or COD-confirmed order. Well-run flows with quick replies and COD confirmation scripts convert 15 to 30 percent of conversations.
Combine them for true cost per WhatsApp order. At PKR 120 per conversation and a 20 percent close rate, your cost per order is PKR 600. Compare that directly against your website CPA. Many stores discover their WhatsApp path is their cheapest acquisition channel, and it was invisible in their standard reports.
Making WhatsApp visible in your reports
UTM-tag every WhatsApp link in your ads so you can trace conversations to campaigns. Ask every customer where they found you and log it. Once a week, reconcile: Ads Manager purchases plus WhatsApp-confirmed orders minus overlap. That reconciled number is what your MER should be computed against. Stores that do this reconciliation routinely find 15 to 30 percent of their revenue was unattributed, which changes every scaling decision.
Metrics that lie: what to stop watching
Impressions and reach are volume, not performance. A million impressions with no sales is a million wasted. Likes, comments, and shares feel good and mean almost nothing for revenue; engagement campaigns manufacture them cheaply, which is why engagement-optimized campaigns look amazing and sell nothing. Video view counts without watch-time context hide whether anyone actually watched. Relevance diagnostics in isolation cause more panic than insight; a “below average” ranking with a profitable CPA is a profitable ad, full stop.
The most dangerous vanity metric is ROAS on a tiny sample. A 9.0 ROAS on PKR 800 of spend is one lucky sale, not a scaling signal. Never make budget decisions on fewer than 10 to 15 purchases per ad set. For the full kill-or-keep framework, see our rules for when to kill a losing ad set.
Setting up your Ads Manager columns
Customize your columns once and stop drowning. Your default view should be: Amount Spent, Purchases, Cost per Purchase, Purchase ROAS, Link CTR, CPC (link), CPM, and Add to Carts. That is eight columns that answer every important question: how much did I spend, what did it buy, how efficient was it, is the creative working, and is the funnel leaking.
Check this view daily but judge weekly. Daily numbers in Pakistan swing hard: payday cycles, weekends, load-shedding evenings, even cricket matches move conversion patterns. Make budget decisions on 7-day rolling numbers. Use daily checks only to catch emergencies like a disapproved ad or a broken checkout.
Set your attribution window deliberately. For COD stores with considered purchases, 7-day click captures the WhatsApp-delayed buyers that 1-day click misses. Just be consistent: changing attribution windows mid-analysis makes trends unreadable.
Typical ranges for Pakistani stores
Benchmarks are ranges, not promises. Most stores we see land in these bands, and your category, AOV, and creative quality move you within or outside them.
- Link CTR (prospecting): 1 to 2 percent. Below 0.7 percent is a creative problem.
- CPC (prospecting): PKR 30 to 80. Above PKR 120 needs investigation.
- Website conversion rate: 1 to 3 percent of ad clicks. Below 1 percent is a store problem.
- CPM: PKR 400 to 900 normally, spiking in Ramadan and November.
- Cost per purchase: varies by AOV, but healthy is 25 to 35 percent of AOV for COD stores after RTO adjustment.
- Reported ROAS: 3 to 5 for established stores. New stores often see 1.5 to 2.5 in month one.
- RTO rate: 20 to 35 percent is typical for COD. Above 40 percent is a confirmation-process emergency.
- WhatsApp conversation-to-order rate: 15 to 30 percent with a proper confirmation flow.
Compare yourself against your own history first and these ranges second. A store improving from 2.0 to 3.0 ROAS month over month is winning, even if a benchmark post says 4.0 is normal. For deeper context on what good ROAS looks like at different stages, read our ROAS benchmarks for Pakistani ecommerce.
If you want these numbers interpreted for your store by people who live in them daily, our Meta ads management service includes weekly reporting in plain language: what happened, why, and what we are doing next. No forty-column spreadsheets.
FAQs
Which single metric matters most?
Cost per delivered order. It captures ad efficiency, funnel performance, and RTO reality in one number. If you only track one thing, track what each delivered order truly costs you, all-in, and keep it below your target.
Why does my Ads Manager ROAS not match my actual profit?
Three gaps: returns and RTO that Ads Manager never sees, WhatsApp and offline orders it under-attributes, and costs like courier returns and packaging that live outside the ad account. Reconcile monthly with the COD-adjusted calculation in this post and judge campaigns on the adjusted number.
How often should I check my metrics?
Glance daily for emergencies, judge weekly for decisions. Daily COD data is too noisy for budget calls. A 7-day rolling view smooths out payday effects, weekends, and the WhatsApp delay between click and order.
Should I use 1-day or 7-day click attribution?
For most Pakistani COD stores, 7-day click. Your buyers routinely click today and order tomorrow after a WhatsApp conversation or family discussion. 1-day click systematically undervalues prospecting campaigns and overvalues retargeting. Pick one window and keep it consistent across all reporting.
Want help with this? Contact us for a free 10-minute audit. We will look at your store and tell you honestly what is possible. No pitch, no obligation.



