The issue with the before and after test is that there are a whole host of ways in which someone might be encouraged to purchase a product from your site that have nothing to do with a paid advert, but coincide with it. For example, someone might read an article about the kind of products you sell and go on to the web to buy one of them. Although you are running adverts against those searches/products, the article was the origin of the sale.

Likewise, it’s very difficult to accurately assess and predict sales on a week-to-week or month-to-month basis, so whilst it’s a good idea in practice, it’s harder in reality to make totally accurate calculations about what would have happened had you not run the adverts.

The technology is improving though and we’re definitely moving towards a better understanding of the direct connection between paid search and sales.