Facebook Ads New Ecommerce Store No Data Playbook

Meta ads guide

You have a new Shopify store. Products are listed, the theme looks decent, and the Meta Pixel is installed. You launch your first campaign with PKR 2,000 a day. Three days later: 400 clicks, zero sales, and an ad account that seems to be learning nothing. Welcome to the cold start, the hardest phase of Meta advertising, and the phase where most Pakistani stores burn their first PKR 30,000 learning nothing.

This is a cold-start playbook built for stores with zero data, zero pixel history, and small PKR testing budgets. It covers the cold-pixel learning problem honestly, the exact budget math for testing in Pakistan, and the COD-store realities that foreign guides never mention.

The cold-pixel problem, explained honestly

Meta’s algorithm optimizes by pattern matching. Show it 1,000 buyers and it finds you 1,000 more people like them. Show it zero buyers and it guesses. Your first week is the algorithm guessing, and guessing is expensive.

A brand-new pixel has no purchase history, no add-to-cart history, and no engaged audience. That means your first campaigns cannot optimize for purchases in any meaningful way, because the system has no idea what a buyer looks like for your store. Most cold-start advice skips this and tells you to run purchase campaigns anyway. The result is Learning Limited ad sets from day one, which is exactly the trap our Learning Limited fix for small budgets helps you escape once you have data flowing.

What zero data really costs you

Expect your first 7 to 14 days to be a paid research project. Cost per click will be higher than benchmarks because the algorithm is exploring. Conversion rates will be lower because your store has no reviews, no trust signals, and no retargeting pool. Budget for this. A store that launches expecting profitable ROAS on day three almost always panics on day five and kills campaigns that were actually gathering useful data.

Week 0: Get the plumbing right before spending a rupee

Half of cold-start failures are tracking failures, not advertising failures. If your pixel misfires, every rupee of testing teaches the algorithm the wrong lessons. Do this checklist before your first campaign goes live.

Pixel, Conversions API, and event setup

Install the Meta Pixel on every page of your Shopify store, then add the Conversions API for server-side events. Browser pixels alone miss a meaningful share of events in Pakistan, where shoppers use in-app browsers, slow connections, and ad blockers. The Conversions API recovers a good portion of those lost events and gives the algorithm cleaner data from day one.

Verify with a real test order that these events fire: PageView, ViewContent, AddToCart, InitiateCheckout, and Purchase. One broken event, especially Purchase, poisons everything downstream. If your purchases are not tracking, stop and read our guide to fixing the Meta Pixel not tracking purchases on Shopify before spending anything.

Domain verification and catalog

Verify your domain in Business Settings and configure your 8 prioritized web events. If you have more than 10 SKUs, connect your product catalog through the Shopify sales channel. Catalog sales campaigns are the best cold-start format for multi-product stores because each product ad accumulates its own engagement signal even before purchases exist.

Store readiness: the unsexy part

Your ads will send traffic to a store nobody trusts yet. Before launch, make sure you have: a working COD checkout with clear delivery timelines, a WhatsApp chat button or click-to-chat link on product pages, at least a few product reviews (seed them from friends and early buyers honestly, never fake them), a returns and exchange policy page, and JazzCash and Easypaisa mentioned as payment options even if COD is your main method. Traffic to an unready store does not just waste ad spend, it teaches the pixel that your visitors do not buy.

How much to budget for testing: the PKR math

Here is the honest testing math for Pakistan. A cold-start test needs enough spend to produce readable signal, and readable signal needs a minimum number of events.

Plan on a daily testing budget of PKR 1,500 to 2,500. Run it for 7 full days without touching it. That is a total test fund of PKR 10,500 to 17,500, the minimum spend that buys you real answers about creative and audience. Spreading the same total thinly over a month never accumulates enough daily signal for the algorithm to learn anything.

Only test with money you can afford to lose completely. If PKR 15,000 disappearing with zero sales would hurt your business, you are not ready for paid ads yet. Build organic traction first through TikTok, Instagram reels, and WhatsApp broadcast lists, then come back to Meta with a small test fund and lower pressure.

For a deeper breakdown of what different budget levels actually buy you in this market, read our guide on how much budget Meta ads need in Pakistan.

The cold-start campaign architecture

Simple beats clever when you have no data. Complexity fragments your tiny budget into ad sets that each learn nothing.

One campaign, one ad set, four to six ads

Run a single sales campaign with one broad ad set targeting Pakistan, or your main cities plus nationwide, with 4 to 6 ad variations inside it. That is your entire cold-start structure. Do not build five ad sets for five interests. Do not split men and women. Do not create separate campaigns per product yet.

Why one ad set? Because your PKR 2,000 daily budget needs to concentrate into a single learning pool. Five ad sets at PKR 400 each produce five ad sets that never exit learning. One ad set at PKR 2,000 has a fighting chance, especially if you optimize for an event with enough volume.

Which optimization event to start with

With zero purchase history, start by optimizing for Add to Cart or Initiate Checkout, not purchases. You will get 4 to 6 times more events, the ad set can actually learn, and you will quickly discover which creatives drive buying intent. This feels wrong to most store owners, who want the algorithm hunting buyers from hour one, but an algorithm with no buyer data cannot hunt buyers. It can only guess expensively.

Move up to purchase optimization once you are seeing a steady 5 to 7 purchases a day, or once your pixel has roughly 100 purchase events in its history. That transition usually happens in week three or four for stores spending PKR 2,500 to 4,000 a day with a working funnel.

Catalog sales for multi-product stores

If you sell 20 or more SKUs, run a catalog sales campaign alongside or instead of your manual creative test. Dynamic ads show each shopper the products they viewed, and the catalog format generates engagement signal across your whole range. It is the closest thing to a shortcut through the cold start for variety stores.

Creative that works with zero data

With no pixel data, your creative does all the targeting. The ad itself must qualify the buyer, because the algorithm cannot.

Lead with UGC-style video

Phone-shot, unpolished video of the product being used outperforms studio renders for cold Pakistani audiences. A 20-second clip of someone unboxing your product, showing the actual size in their hands, and naming the price beats a glossy graphic every time. New stores cannot afford brand polish anyway, and authenticity converts better in this market.

Write hooks in Roman Urdu

Your first line decides everything. Roman Urdu hooks consistently outperform English for mass-market products: “Mehngay products se tang aa gaye ho?” or “Ye 2,500 wali deal dobara nahi milegi.” Name the price early, say “Cash on Delivery available” in the first three seconds of video or the first line of copy, and show the product solving one clear problem. One product, one problem, one price, one call to action.

Test angles, not just products

Your 4 to 6 launch ads should test different angles on the same hero product: a price angle, a problem-solution angle, a social-proof angle (even early reviews count), and a comparison angle against the expensive alternative. The angle that wins tells you what your market actually cares about, which is worth more than the sales it generates.

COD-store realities most guides ignore

Pakistani ecommerce runs on cash on delivery, and COD changes the cold-start math in ways foreign playbooks never address.

Pad your numbers for RTO

Most new COD stores see return-to-origin rates of 20 to 35 percent. That means your cost per delivered order is 25 to 50 percent higher than your cost per Shopify order. When you judge your cold-start test, judge it on estimated delivered orders, not raw orders. A PKR 600 cost per order with 30 percent RTO is really a PKR 857 cost per delivered order. Price your test budget accordingly.

Confirm every order on WhatsApp

From day one, confirm every COD order with a WhatsApp message before dispatching: order summary, total with delivery charges, and a request to reply YES to confirm. This single habit cuts RTO dramatically and costs nothing. It also builds you a WhatsApp list of buyers, which becomes a free retargeting channel for Eid, 11.11, and new launches. Cold-start ad spend that also builds your WhatsApp list is doing double duty.

Offer JazzCash and Easypaisa, even if barely anyone uses them

Listing JazzCash and Easypaisa as prepaid options does two things: it captures the small but growing segment of digital payers, and it signals legitimacy to COD buyers who equate payment options with real businesses. Prepaid orders also have near-zero RTO, so even a 10 percent prepaid mix meaningfully improves your delivered-order economics.

Reading the first 7 days

Do not judge your cold start on sales alone. With a new pixel, the first week is about leading indicators.

  • CTR (link): above 1 percent means your creative is working. Below 0.7 percent after 3,000 impressions means the creative or hook is the problem, not the budget.
  • CPC: most Pakistani ecommerce cold campaigns land between PKR 30 and 80. Far above that usually means a narrow audience or weak creative.
  • Add to Cart rate: if clicks are cheap but nobody adds to cart, your product page or price is the bottleneck, not your ads.
  • Cost per Initiate Checkout: this is your early efficiency metric. Track it daily and watch the trend, not the absolute number.

One sale in week one does not mean the campaign works, and zero sales does not mean it failed. You are buying information. The question at day seven is: do I have a creative with a healthy CTR and a checkout flow that converts clicks into carts? If yes, you have something to scale. If no, you know exactly which part to fix.

Your 30-day cold-start plan, week by week

Week Daily budget Structure and optimization event Goal
Week 0 (pre-launch) PKR 0 Pixel + Conversions API + domain verified; COD checkout, WhatsApp chat, reviews, returns policy ready Fix tracking before spending a rupee
Week 1 (days 1-7) PKR 1,500 – 2,500 1 campaign, 1 ad set, 4-6 ads; optimize for Add to Cart or Initiate Checkout Test fund of PKR 10,500 – 17,500; touch nothing for 7 days
Week 2 (days 8-14) PKR 1,500 – 2,500 Keep winners; read CTR, CPC, Add to Cart rate Day 14 decision: scale, iterate, or pivot
Weeks 3-4 PKR 2,500 – 4,000 Step up to purchase optimization at a steady 5-7 purchases/day; add a retargeting ad set Scale budget 20% every few days on winners

The 14-day decision: scale, iterate, or pivot

At day 14, make one of three calls. Scale if you have consistent checkouts and a cost per delivered order near your target: raise budget 20 percent every few days and start a retargeting ad set for site visitors, following our retargeting setup for Shopify. Iterate if CTR is healthy but checkouts are thin: fix the product page, pricing, or offer, keep the winning creative, and test again. Pivot if nothing got traction: CTR under 0.7 percent, CPC over PKR 120, near-zero carts. That usually means the product or angle is wrong, not the ad account. No budget fixes a product nobody wants.

If the cold start feels like too much to run alone while also fulfilling orders, our Meta ads management service runs the full testing phase for Pakistani stores: tracking setup, creative direction, and the budget discipline to get you through the learning period without burning cash.

FAQs

How long does the cold start phase last?

Typically 2 to 4 weeks. You exit it when your pixel has around 100 purchase events and at least one ad set has completed learning. Stores spending PKR 2,500 to 4,000 a day with a working funnel usually get there in the third or fourth week.

Should a new store run traffic or engagement campaigns first?

No. Traffic and engagement campaigns optimize for cheap clicks and likes, which teaches the pixel to find clickers, not buyers. Start with sales campaigns optimized for Add to Cart or Initiate Checkout. The events are closer to revenue and the learning transfers when you step up to purchases.

Can I start Meta ads with PKR 1,000 a day?

You can, but keep expectations honest. At PKR 1,000 a day you are buying roughly 15 to 30 clicks daily, which is too thin for fast learning. Treat it as a slow creative test over 3 to 4 weeks rather than a real launch, optimize for Add to Cart, and do not judge results before day 21.

Do I need reviews before running ads?

You do not need hundreds, but zero reviews on a new store kills conversion. Get at least 5 to 10 honest reviews on your hero products before spending: friends, family, and early organic buyers. Real, specific reviews mentioning delivery speed and product quality outperform generic five-star ratings.

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.

Leave a Comment

Your email address will not be published. Required fields are marked *