Housedify: testing Meta Ads on a brand-new Shopify store
I set up tracking first, then tested creatives and audiences on a small budget to find out what actually turns ad spend into sales.
- Channel
- Meta Ads
- Business
- My own Shopify store
- Category
- Home Organisation & Storage
- My role
- Founder: store setup, tracking, creative and Meta Ads
- Tools
- Meta Ads Manager, Meta Events Manager, Shopify, Canva, ChatGPT
across 2 products
a foldable wardrobe organiser and a multi-functional curtain rod
store revenue (own-store figure)
Overview
Launching a new Shopify store meant starting from zero: no audience, no proven product, and no data on what would sell. I used Meta Ads to test demand quickly, working out which products, creatives and audiences could turn a small daily budget into sales.
Challenge
- A new store with no existing audience
- A limited budget for testing
- No data on which creatives or audiences would work
- A real risk of spending on the wrong creative or audience
Goals
- Generate sales through Meta Ads
- Test demand across products
- Identify the best-performing creative format and audience type
- Build a repeatable, efficient campaign structure
My approach
- 1. Tracking first. Before spending anything, I set up Meta Pixel and Conversion API on Shopify and verified event accuracy in Events Manager.Becaue bad tracking data would have made every later decision unreliable & wrong.
- 2. Creative testing. I tested video against carousel first, because creative usually matters more than targeting in a brand-new account.
- 3. Audience testing. I compared interest-based targeting with broad targeting, since it wasn't clear which would deliver a lower cost per result for a new store.
- 4. Budget discipline. Campaign 1 ran on a ₹800 daily budget, monitored closely so weak creative and audiences could be close quickly.
What I did
- Set up Meta Pixel and Conversion API on Shopify and verified events before launch.
- Created ad creatives in Canva.
- A/B tested video against carousel ads.
- Compared interest targeting with broad targeting.
- Tracked CTR, CPC, CPM, conversions and ROAS daily.
- Paused weak creatives and kept the winners.
Results
All figures are for my own store. ROAS is as reported by Meta Ads Manager. Revenue is store revenue.
| metric | Campaign 1: Foldable Wardrobe Organiser | Campaign 2: Multi-functional Curtain Rod |
|---|---|---|
| Sales | 7 | 3 |
| Ad spend | ₹5,318.98 | ₹1,359.91 |
| Campaign ROAS | 2.15x | 2.17x |
| CTR | 1.69% | 1.57% |
| CPC | ₹23.96 | ₹6.54 |
| CPM | ₹405.26 | ₹102.35 |
| Best ad ROAS | 3.26x (6 purchases) | 4.34x (3 purchases) |
In each campaign, the best ad carried most of the purchases.
Creative test: video vs carousel (Campaign 1)
- Cost per post engagement: video ₹2.13, carousel ₹34.72
- Purchases: the video ad drove 6, and the carousel drove none, on a smaller reach (1,247 against 6,713)
Audience test (Campaign 1)
| Ad set | Spend | Reach | CTR | CPC | Purchases |
|---|---|---|---|---|---|
| 25–50 Women, interest-based (Home Organisation) | ₹3,727.95 | 7,818 | 1.63% | ₹25.02 | 6 |
| 25–50 Women, broad | ₹1,591.03 | 3,620 | 1.84% | ₹21.79 | 1 |
Small numbers, so I treat this as a signal to keep testing, not a conclusion.
What I learned
- Tracking comes before optimisation. Verified events are what made the later reading of ROAS and purchases trustworthy.
- Read each test at the right level. The video's engagement cost and its purchases point the same way, but they're different measures, so I treat them separately.
- Small samples are signals, not proof. Every figure here rests on a handful of purchases. That's why I treat them as reasons to run the next test, not as benchmarks.
- Move budget to what's working, quickly. Paused weak creatives and kept the winners.
This was a test on my own store with a small budget, not a client engagement. Results depend on the product, price, offer and creative, and they aren't a benchmark for other brands.