Predict number of users in an upcoming month and change marketing strategy accordingly

Why use forecasting: Businesses around the globe have been forecasting their sales for a long time. The primary reason for it is planning, for instance, amount of inventory to store or when to put the budget on marketing etc.

Requirement: Predict Trend for the month of September

The domain of the Dataset: Products and Retail. However, the application of the algorithm is not limited to only products and Retail. The technique can be applied wherever we want to discover the upcoming trend.

Following data points will be helpful: Organic Data

  1. Date
  2. Users
  • Methodology: Facebook recently released a forecasting library for Python and R, called Prophet. It’s designed for forecasting future values of time series of any kind and is remarkably easy to get started with. Pretty impressive output for so little work required! A prophet is fast too – these results were computed in only seconds.

Using Prophet to generate predictions turned out to be very easy, and there are several ways to adjust the predictions and inspect the results. While there are other libraries that have more functionality and flexibility, Prophet hits the sweet spot of predictive power versus ease of use. No more looking at weird plots of predicted values because you chose the wrong algorithm for your use case. And no more spending hours to fix input data that has gaps or timestamps in the wrong format.

The Prophet forecasting library seems very well designed. It’s now my new favorite tool for ad-hoc trend analysis and forecasting!

 

  • Description & results: Businesses around the globe have been forecasting their sales for a long time. The primary reason for it is planning, for instance, amount of inventory to store or when to allot the budget on marketing etc. Now a day every e-Commerce business, big or small, has their online presence. This gives them an opportunity to accumulate data about their consumer’s behavior, demographics, sources, no of new visits, etc. which can be indirectly used to predict sales. A precursor to sales can also be found by calculating some correlation between the number of visitors and sales. (If this is not the case then one should first rethink about the efficacy of the website in generating revenue) Since many consumers thoroughly research the products and services online before they buy, the web analytics forecasted no of visitors can quickly alert you on any new trend, then what the sales data can.

 

ABC of forecasting:

Forecasting is the process of estimating a future event based on recent and past time series data. It may not reduce the uncertainty of future; however, it gives the decision makers an idea and a basic premise for planning. Short term forecast will always be more accurate than Long term forecast.

 

We will use a Time-Series Model for our forecasting purpose. People may visit a particular website for many different reasons which are next to impossible for us to fathom all the underlying factors. So, we presume to know nothing about the causality that affects the variable we are trying to forecast. Instead, we examine the past behavior of a time series in order to infer something about its future behavior. Time-series models are particularly useful when little is known about the underlying process one is trying to forecast.

 

Graph of Weekends User’s: Without prediction i.e. based on already available data from 1st January 2016 to 31st August 2017

  • From the graph, we can understand that our website is not performing well on weekend’s as compared to other days of the week. As most of us as an assumption website do well during the weekends, the graph clearly shows our assumption is not correct.
  • Also, website performance has gone up every half yearly.
  • Since it is not possible to answer on which day our website is performing well. We go ahead and plot one more graph for all days.

 

  • plot_weekend

 

Graph of All Day User’s: Without prediction i.e. based on already available data from 1st January 2016 to 31st August 2017

  • From the graph, we can see our website is performing well during mid-week days i.e. Wednesday and Thursday.
  • In order to check all days when website performs well, we plot box plot.

plot_allDays

 

Box Plot of All Day User’s: Without prediction i.e. based on already available data from 1st January 2016 to 31st August 2017

  • From the box plot, we can see our website performance order.

Wednesday ==> Thursday ==> Friday ==> Sunday ==> Tuesday ==> Saturday ==> Monday

So, Website has the least user on Saturday and Monday.

boxPlot_AllDays

Forecast Users:

  • From the Forecast graph, we can predict our website will perform well in a coming month.

forecast_plot

 

From forecast Data Output: The data show users are increasing everyday

 

forcasted_data

 

  • days_month wise_plot: With Predicted Data

 

days_monthwise_plot

  • From the trend graph, we can predict our website performance will increase will time.
  • From the weekly graph, we can predict user’s visits will increase gradually from Sunday to Wednesday and then will drops from Wednesday to Saturday.

Wednesday ==> Thursday ==> Tuesday ==> Friday ==> Monday ==> Saturday ==> Sunday

As we can see order is different from the one given by box plot (plotted using existing data)

Wednesday ==> Thursday ==> Friday ==> Sunday ==> Tuesday ==> Saturday ==> Monday

 

Sample Data and Script: //drive.google.com/open?id=0B5Onjfjgz6h_bFhHS1pxMURCRGs 

Please follow and like us:
569

You may also like...

1 Response

  1. sanction says:

    I sаvour, гesult in I found exactly what I
    was taking a look for. You’ve ended my 4 dаy lengthy hunt!
    God Blesѕ you man. Have a nice day. Bүe

  2. WilliamBuM says:
    Your comment is awaiting moderation.

    microsoft outlook
    microsoft solitaire
    microsoft store windows 10
    office online
    xbox store

Leave a Reply

Your email address will not be published. Required fields are marked *

www.000webhost.com