We started working on this client in June 2025, where we started the campaigns with a very low budget to test out a few things first:
1. Audience Targeting,
2. Creatives & Ad Copies,
3. Best Selling Categories/Products,
4. CBO Vs ABO
- Audience Targeting: In Audience Targeting, we tested out mostly the Broad Audiences & kept the age group as 18-65+ with Automatic Placements (Facebook’s algorithm works very well when we give FB the authority to serve to the best audience by just giving the relevant interests).
We segmented the audiences in 5 different ways:
– Broad (Direct Interests like Jewelry, Silver Jewelry etc.)
– Brands (Like Caratlane, etc.)
– Magazine (Fashion Related)
– Influencers (Bollywood Celebs and Designers)
– Indirect Targeting (People who like Luxury, Style, Traveling etc.)
2. Creatives & Ad Copies: We started the campaigns on All Products using Catalog and used different ad copies in different ad sets.
The reason for using Catalog Ads was already having a good photoshoot on the website, so we thought of using the same images.
We kept ad copies very simple and short and to the point.
Just followed the KISS Formula in ad copies.
(KISS – Keep It Simple Silly).
1. Best Selling Categories/Products: After running the campaigns for 10 days, we studied the backend data and analytics to figure out our best-selling categories and products.
After having that data, we created sets in the catalog of best-selling categories and started marketing sets in different ad sets with new interests.
1. CBO vs ABO: We usually test the audiences in ABO campaigns but this time we thought of testing out both ABO and CBO and as always ABO only worked for us.
CBO campaigns have the limitation that it spends more money on the adset on which it gets the most engagement and due to which other potential adsets didn’t spend enough to get an idea of its capability.
In ABO campaigns, we can judge the performance easily as it spends evenly.
CAMPAIGNS SETUP:
We created 2 different campaigns with 4 ad sets each.
In the First Campaign, we used ABO with the budget of INR250/adset.
In the Second Campaign, we used CBO with the budget of INR750/campaign.
In ABO, with 4 ad sets we targeted all BROAD with Single Interest and used All Catalog with Single Ad Copy.
In CBO, we targeted all BROAD with Stack of interests and used All Catalog with Single Ad Copy.
BUDGET ALLOCATION: –
We started the ABO campaigns with a budget of INR 250/adset, whereas in CBO we started with the budget of INR 750/campaign.
We analyzed the performance of all ads for 3 days and then started optimizing it.
The ad set which gave us even a single sale, we kept that on and turned off all other ad sets with 0 sales.
After that we tested out many new interests and followed the same formula for 15-20 days just to understand our audiences.
The ad sets which gave us the sales, we increased their budget by 20% every 6-8 hours.
SCALING:
We scaled up the campaigns using: Manual Bidding, Lifetime Budgets & LLAs
MANUAL BIDDING
We were getting the average CPP (Cost Per Purchase) of INR 150, so we created a campaign with manual bidding and started it on the budget of INR 5000 and gave multiple cost caps from INR 120 – INR 200 in gaps of INR 10.
Waited for 3 days and analyzed which one is giving us the sales with a positive ROAS.
Stopped the ad sets with negative ROAS or 0 sales and started scaling up the one with positive ROAS by doubling the budget in every 24 hours
LIFETIME BUDGET
We created a new campaign with a Lifetime Budget of INR 50,000 for 10 days, and let it spend 20% of its total budget to analyze the results.
It started giving us the positive ROAS from day 1 so we scaled up the budget by 20% every 12 hours.
LOOKALIKE AUDIENCES
Using: Lookalikes (View Content, Add to Cart, Purchase, Engagement)
We created the lookalikes of each event type:
1. View Content – 0-1%, 1% – 2% & 2% – 3%
2. Add to Cart – 0-1%, 1% – 2% & 2% – 3%
3. Purchase – 0-1%, 1% – 2% & 2% – 3%
4. Engagement (We created the custom audience of people who engaged with our ads and then used it to create LLA) – 0-1%, 1% – 2% & 2% – 3%
After that we created different campaigns for each Event Type:
Campaign 1: View Content
3 Ad sets with audiences 0-1%, 1% – 2% & 2% – 3%
Campaign 2: Add to Cart
3 Ad sets with audiences 0-1%, 1% – 2% & 2% – 3%
Campaign 3: Purchase
3 Ad sets with audiences 0-1%, 1% – 2% & 2% – 3%
Campaign 4: Engagement
3 Ad sets with audiences 0-1%, 1% – 2% & 2% – 3%
And followed the same testing process of INR 250 per ad set with 3 days analyzing rule.
In the end, we found our best performing LLAs and started scaling them up by increasing the 20% budget every 6 hours.
REMARKETING:
Now here comes the profit-making formula which we implemented to scale up the overall ROAS and increase the number of sales.
So, we created different remarketing campaigns for each funnel level:
1. MOF (Middle Of the Funnel): The warm audiences who either engaged with the ads or visited the website and took any action but didn’t purchase.
1. BOF (Bottom Of the Funnel): So this is segmented into different ways:
People who Viewed Content in the last 7, 14, 21, 30 Days.
People who added to cart in the last 7, 14, 21, 30 Days.
People who Initiated Checkout in the last 7, 14, 21, 30 Days.
People who Added Payment Info in the last 7, 14, 21, 30 Days.
1. Retention Campaigns: The people who have already made a purchase from us.
In MOF, we created all events (FB Engagers, Insta Engagers, View Content, AddToCart, Initiate Checkout, Add Payment Info) of the last 180 days and excluded all events of the last 30 days & Last 30 days purchasers.
We created a dynamic ad for this audience and it gave us the best ROAS and highest number of sales.
In BOF, we created different campaigns for all funnel level:
1. Viewed Content with 4 ad sets
2. View Content (Last 7 Days) – Excluded ATC & Purchasers (30 Days)
3. View Content (Last 14 Days) – Excluded VC (7 days) ATC & Purchasers (30 Days)
4. View Content (Last 21 Days) – Excluded VC (14 days) ATC & Purchasers (30 Days)
5. View Content (Last 30 Days) – Excluded VC (21 days) ATC & Purchasers (30 Days)
We followed the same for all the event types and created different ad copies & offerings at each funnel level.
It helped us in increasing the overall ROAS by retargeting to the people who already know about the brand which increased the relevancy of the ads.