Brief to DataPOEM
Being a well-funded startup, The brand was dependent more on their paid media spending to drive growth for their brand on the e-commerce platforms on both Brand website and marketplace. The major problem for the brand was to measure what is the impact the brand's marketing spends on their marketplace sales.
Below is the Data we have taken as an input.
Media Data:We have taken access to all the advertising platforms used our APIs to fetch the data automatically to avoid any data loss.
Web Analytics Data:We have taken web analytics access i,e google analytics and amazon seller central and used our API’s top fetch the data.
Market Place spends data:We have taken the spend data that the brand is spending on Amazon Ads.
Retail sales data:Weekly Sales data of retail channels.
The platform fetched the data from different platforms with as much granular as possible such as campaign/ad group/stage/city with all the key metrics and ingested the data to identify what is impacting the sales on the e-com platform and marketplace sales.
We have built 20+ models before finalizing the best model that gives 94% accurate attribution of the marketing channels.
The strategy was to predict the latest 1 month's sales from the available data without exposing the latest sales data to the model but by inputting the marketing spends of each channel.
1. The client was able to see the real attribution of their marketing spending on the sales.
2. The client was able to identify what is impacting their marketplace sales and howmuch it is impacting.
3. The client was able to see the plan that they can execute in the following month and estimate the outcomes.
4. The client was able to test the hypothesis they have on the media mix of their marketing activities.
5. The client identified the no impact marketing activities and saved the costs.
1. 25% savings on the media spends by delivering the same results.
2. Scenario planning with 15+ hypothesis.
3. 94% Accurate model results.