Analysis of advertising campaign to reassess where to allocate resources
This report is comprised of two datasets, one is total sales and the other is a report from Amazon on the effectiveness of an advertising campaign. The goal of this project is to summarize the two datasets and uncover any significant efficiencies/inefficiencies in order to optimize advertising spending in the future.
For the order report, I created several visualizations showcasing sales over time, how products are doing compared to each other and a geographical breakdown of where the bulk of sales are coming from.
For the advertising report, I focused more on trying to draw correlations between sales and the different metrics offered by Amazon.
Moving forward, I will follow along as I receive new data to see how advertising efforts are affecting sales numbers.
-
All cleaning was done in order report cleaning and ad report
-
All visualizations were done in Orders_Viz and Advertising Viz
- Looking at the monthly breakdown of sales we can see a steady increase in sales over time with a steep increase in May and June
- IMPORTNAT NOTE, both December and June are only half of a month
- Looking forward I will be updating as time goes on to determine if these rises are seasonal or if they are going to continue to rapidly increase.
- Here we can see that Rainforest Bowls are outperforming Floral Bowls every month.
- The one unique trend is the dip in Rainforest sales in March along with a steep rise in Floral sales. It could be worth looking into the following year to see if a similar trend takes place.
- Initially I wanted to see what states placed the most orders in order to hone in on what regional demographics are worth targeting. However, the first thing I noticed was that all of the states with high order totals all have massive population sizes.
- This led me to add a second visualization to show which states have the most sales in relation to their population size.
- After viewing both graphs I'd conclude that California, Vermont, Washington, Massachusetts and Colorado are the best states to market to.
| Orders By State | Orders By State Per Capita |
|---|---|
 |
 |
- Click this link to view a map that includes a pin for all purchases including international purchases https://www.google.com/maps/d/edit?mid=1LwwXuNbNnIjzqFwPGB1MRaMHgiwj1_GA&usp=sharing
- Here, I am trying to see what factors impact how much we spend.
- As expected, Clicks are almost a perfectively direct correlation.
- What I found interesting was that Sales had a clearer correlation than impressions.
- With this visualization I wanted to see the difference between how Impressions and Sales effect how much we spend.
| Impression and Spending | Sales and Spending |
|---|---|
 |
 |
- ACoS (Advertising Cost of Sale): shows how much you spent on ads to gain a dollar from attributed sales.
- ROAS (Return on Ad Spend): tells you how much money you earn for every dollar you spend on advertising.
- In order to effectively adjust how we spend our advertising money I wanted to look at what key words we're spending money on but yield zero profits.
- We can clearly see that profits are trending upwards and that if we continue to actively identify efficiencies as well as inefficiencies we can easily justify all adversising spending.
- We will look to monitor these trends over the upcomming months to make improvements.