Dear Jean-Brice, I am under the impression that it is somewhat difficult to predict consumers' reactions in experimental conditions. Stated differently, in the real wold, there are numerous intervening variable that we may not be able to control. That being said, please find embedded a link to a Google Scholar search with some of your keywords. Hope this helps. 

https://scholar.google.co.za/scholar?start=0&q=e-advertising+and+experiment+and+%27purchasing+behaviour%22&hl=en&as_sdt=0,5&as_ylo=2012

Time delay between the exposure and the purchase can be conducted as an experimental method.it can be done with a focus group.

https://www2.bc.edu/~woodsiar/publications/CJAS%20MacDonald.pdf

Dear Jean,

The attached paper may help you. You also may use the references of the attached paper. I also suggest that you find papers with similar constructs to your own constructs and relate them to your study.

Best wishes,

Amin Ansary

Leona, Ligo and Amin, thanks for your answers.

I would look into Avinash Kaushik's blog on google analytics. there is some useful information about how e-sites( commerce and non commerce). Use KPIs to predict consumer behavior. There are also environmental and economic factors at play when measuring consumer behavior. Price, possible events (natural disasters, social events, anything that creates the perception that the price of a product or service will increase/decrease) Also economic factors such as recessions and economic downturns/upturns will drive price. 

Thank you Christopher Conyers for your response.