Systems and methods for predicting traffic on internet sites
View Patent ↗Systems and methods are provided for predicting visitor traffic to a network of web site pages. The systems and methods are used, as an example, to predict the inventory of total available online advertisements available within the network for a forthcoming period. The visitor traffic includes page viewing, listening or transacting on web pages within a web site, wherein the web pages are categorized by subject, interest areas or specific user queries such as word or phrase searches. For each page whose traffic is being predicted, the system takes into account annual seasonality, day-of-week, holidays, special events, short histories, user demographics, user web behavior (viewing, listening and transacting) and parent and child web page characteristics.
1. A computer implemented method of predicting traffic for a web page in a network of web pages, the method comprising:
automatically generating, using a processor, historical traffic data for the web page based on records of daily traffic for an ancestor web page of the web page if the web page has been available for less than a predetermined period of time or based on records of daily traffic for the web page and descendent web pages of the web page;
automatically back testing, using the processor, the historical traffic data using a growth and seasonality separation (GSS) process;
automatically back testing, using the processor, the historical traffic data using a trend copy (COPY) process; and
automatically predicting, using the processor, future traffic to the web page based on the historical traffic data, wherein predicting includes applying to the historical traffic data either the GSS process or the COPY process that yielded a better back testing result.
2. The computer implemented method of claim 1 , wherein generating historical traffic data includes generating a pageview for the web page.
3. The computer implemented method of claim 1 , wherein generating historical traffic data includes generating a runview for the web page, wherein the runview is a sum of pageviews of the web page and pageviews of the web page's descendents.
4. The computer implemented method of claim 1 , wherein predicting future traffic includes:
identifying a holiday; and
removing from the historical traffic data fluctuations in traffic due to the holiday.
5. The computer implemented method of claim 4 , wherein removing from the historical traffic data includes replacing the data for the holiday in the historical traffic data with the data from a prior day.
6. The computer implemented method of claim 5 , wherein the prior day is seven days prior to the holiday.
7. The computer implemented method of claim 1 , wherein predicting future traffic includes:
for each day, replacing in the historical traffic data the traffic for that day with the average traffic during the week of that day.
8. The computer implemented method of claim 1 , wherein predicting future traffic includes:
determining a long term growth rate based on a traffic increase during a period of time; and
thereafter determining a seasonal change rate based on the determined long term growth rate and the traffic increase.
9. The computer implemented method of claim 1 , wherein the predictions made for the web page are based on the predictions made for one or more ancestor web pages.
10. A system for predicting traffic for a web page in a network of web pages, the system comprising:
at least a server including a historical data generating module and a traffic predicting module, the historical data generating module operable to produce historical traffic data for the web page based on records of daily traffic for an ancestor web page of the web page if the web page has been available for less than a predetermined period of time or based on records of daily traffic for the web page and for descendent web pages of the web page in the network; and
the traffic predicting module operable to process the historical traffic data for the web page and generate a future traffic prediction for the web page, wherein the traffic predicting module is operable to back test the historical traffic data applying each of a growth and seasonality separation (GSS) process and a trend copy (COPY) process, and to apply the historical traffic data to either the GSS process or the COPY process that yielded a better back testing result to predict the future traffic for the web page.
11. The system of claim 10 , wherein the historical traffic data for the web page includes a pageview of the web page and a runview of the web page, wherein the runview is a sum of the web page's pageview and the pageviews of the first web page's descendents.
12. The system of claim 10 , wherein the historical data generation and the traffic prediction modules are implemented on the same computer system.
13. A computer implemented method of predicting web page traffic, comprising:
generating historical traffic data for at least one web page based on records of daily traffic for an ancestor web page of the web page;
automatically predicting preliminary future traffic for the web page in a computer system based on the historical traffic data generated for the web page, wherein predicting includes applying to the historical data a growth and seasonality separation (GSS) process, and wherein the GSS process includes removing holiday effects from the historical traffic data; and
automatically predicting future traffic to the web page in the computer system based on the historical traffic data generated for the web page, wherein predicting includes applying to the historical traffic data the GSS process, and wherein the GSS process includes adding the holiday effects to the predicted preliminary future traffic to predict the future traffic.
14. The computer implemented method of claim 13 , wherein generating historical traffic data includes generating a pageview for the web page.
15. The computer implemented method of claim 13 , wherein generating historical traffic data includes generating a runview for the web page, wherein the runview is a sum of pageviews of the web page and pageviews of the web page's descendents.