The following data is related to a local online pizza store. Data are coming from a third-party ad provider that runs display ads for our pizza brand. The goal is to optimize the campaign data and see which deal is performing better and why.
| KPIs | $8 Pepp | $8 Hawaiian | $7 2-Top | 40% Off | Grand Total |
| Average Check | $21.49 | $20.35 | $23.48 | $20.57 | $20.83 |
| Conversion Rate | 2.50% | 3.00% | 1.00% | 1.50% | 1.55% |
| Cost / Transaction | $4.09 | $5.67 | $4.39 | $4.21 | $4.27 |
| CPM | $2.38 | $1.87 | $14.30 | $9.42 | $6.31 |
| Transactions | 5,830 | 2,268 | 3,353 | 44,362 | 55,813 |
| Engagements | 12,430 | 6,784 | 5,873 | 134,555 | 159,642 |
| Spend | $23,844 | $12,868 | $14,734 | $186,900 | $238,346 |
| Impressions | 10,008,900 | 6,890,004 | 1,030,000 | 19,834,930 | 37,763,834 |
| Net Sales | $125,298 | $46,158 | $78,728 | $912,383 | $1,162,567 |
| All Clicks | 233,200 | 75,600 | 335,300 | 2,957,467 | 3,601,567 |
* (Data and numbers are fake and only used as example rates.)
1. I have started with sorting the KPIs based on the sales-funnel structure to have better visualization and easier traceback.
2. Then transposed data to be able to sort data based on KPIs and better visualization.
To better understand the data, I have added a few other KPIs that makes the others more meaningful. The metrics are:
- CPE (Cost per Engagement)
- CTR (Click Through Rate)
- CPC (Cost per Click)
- Average Transaction Value
- Profit
- ROAS (Return On Ads Spend)
ROAS is the main KPI to optimize for the maximum profitability of the campaign.
1- The best deals in terms of profitability are “7$ 2-top” and “8$ Pepp” – Using only 11% of the budget, they bring 17.5% of the profit
2- All the offers are profitable and each is best for different KPIs. So as long as budgeting is not a concern, we might keep them all, but with some adjustments in the budgeting for optimal results. If we have to stop any of the offers, I’d suggest stopping Offer A based on the lower ROAS, and higher cost per transaction. But I also suggest taking a deeper look into the deal as I described in the following.
3- Surprisingly, the “8$ Hawaiian deal” has the best Conversion rate, yet the lowest return. This is because of a very low CTR compared to the other deals. This might be to the poor design of the deal on FB: either it is photos, titles, call to action, positioning, targeting, etc. We might consider optimizing the Call to Action and improving the CTR by performing A/B testing for the ad. If we could increase the CTR only by 1% for this deal, with the same conversion rate, we will double the transactions, and net sales, and will make it the top-selling deal and actually the most efficient deal.
4- The “40% OFF deal” is the best for customer acquisition, Social engagement, traffic on the website, brand visibility, and lowest CPC – We might consider running this deal for better customer acquisition and hit non-monetary goals. This is also a really good deal to increase the customers’ lifetime value (LTV) along with more customer data.
1- Measure and add more website KPIs (Time on Site, Page visit/Session, Bounce Rate, Exit pages, Behavior flow, etc). If possible combine with other CRM data to optimize the campaign data for different purposes.
2- Consider using automation techniques (Retargeting, Abandoned cart, etc.) to engage dropped traffic and improve the conversion rate of the campaign. We can automate sending deals to customers’ emails based on their segmentation.
3- Calculate Lifetime Value to see if any of these offers do more better across multiple campaigns (i.e. “40% off offer” might bring us more LTV than the other deals)
4- Gather customer location data (States, Cities, Branches) to understand local behaviors, compare data, and GEO optimization of the campaigns
5- Since the actual net profit relies on the “profit margin” of each different product, having access to those data enables us to calculate the exact profitability of campaigns and optimize more accurately. The above suggestions assume that the profit margin of the deals is the same.
