Web Science/Part2: Emerging Web Properties/Web Search Ecosystem/Factors that have impact on advertisement campaigns/script

After we understand the basic metrics that are being used in online marketing we are able to measure how successful or unsuccessful our campaign is running. But what happens if we are not happy with the results. Let us say we want to decrease the acquisition costs per customer which (we recall from last lecture) is defined as $$\frac{\mathrm{CPC}}{\mathrm{CR}}$$. We have already seen there are two factors which have an impact on the acquisition cost. Of course decreasing the CPC will decrease our total cost. But the CPC is usually something we cannot really manipulate since it is - as we will see in a future lecture - a collaborative auctioning process in most cases that helps to determine the CPC.

This means that another way is to increase the conversion rate of our advertisements in order to decrease the acquisition cost per customer. Since the conversion rate is something we can work on we will now investigate techniques that might help to increase the conversion rate.

Obviously if half of your Clicks that you bought bounce away your conversion rate will suffer. So decreasing the bounce rate is one solution to the problem. Since the bounce rate shows how relevant users find your site in comparison to the advertisement you might want to over think the advertisement text or banner. Maybe your ad is raising to many expectations that your site and offer cannot hold. in this sense a truthful ad will be most useful for you.

In general you should also try to target your advertisement only to people that will be interested to your product. The problem here is that you don't want to run the advertisement in moments and situations where people see it that are your customers anyway. Still carefully selecting relevant keywords and the target audience will help reducing the bounce rate and help increase your conversion rate. one technique that is used nowadays is that of a negative keyword. This techniques prevents displaying your ad if a certain keyword is in the query. A typical example might be jobs in the query hotel tokyo. If you are a booking website you don't want to display your ad to people that are looking for hotel jobs tokyo.

Another reason for a high bounce rate could actually be the layout and design of your site and in particular your landing page. It could just lack usability and user experience. Depending on the entire budget of your campaign you could pretest your landing page with a smaller budget and do A/B testing of various design options and layouts. Here one could even use techniques from statistical machine learning in order to do a real time decision marketing and adapting the site directly to the potential customer.

Last but not least your product should be the best that is actually possible. Increasing the quality of your product and your service and decreasing the burden of interacting with your site will make customers more happy and increase conversion rates.