Electronic Data Gathering, Analysis, and Retrieval—is the electronic filing system created by the Securities and Exchange Commission to increase the efficiency and accessibility of corporate filings. The system is used by all publicly traded companies when submitting required documents to the SEC. Corporate documents are time-sensitive, and the creation of EDGAR has greatly decreased the time it takes for corporate documents to become publicly available.  

Most common filings:

10K-annual statement of financials - generally in the Notes to Consolidation of Statements, or  legal factors, management discussion. Audited statements. 
10Q-Quarterly statement of income and balance sheet information.  It is a snapshot of the company's financial performance, but it is unaudited. 
8K- Current News or events that may affect the financial health of the company
20K - annual report of financials for non-US companies which file on the various US exchanges.  However, their filing requirements are less robust than for US companies. 

Example of a simple company search, looking for 10Ks

After searching for a specific company, you can find the company profile in EDGAR along with its filings. The default view shows the latest filings and popular selected filing types (right box) Also, one can view latest documents by clicking the green button "View Filings.".

Results 

Click on 10K-and 10Q filings in the right box.  

Then scroll down the list and look for 10K. 

Certain marketing data items are in the operations expenses of the income statement.  Those include either "marketing" but more common, Cost Sales or Cost of Revenue.  If not broken out by those two items, it may be under "Sales & General Administrative Expenses".  

If one is searching for Research & Development expenses, those are generally also in the Operating/Operations Expense section.  Not all companies will have research and development expenses, especially if it is a service company as opposed to a goods-oriented company. 

Search by company name or ticker.  Look for historical data (highlighted in screenshot) .  

To export data from Yahoo! Finance using R, here are some instructions.   Another site, Data Science Central, has instructions for importing historical data to R,