Provides access to historical data, report generation, charting and statistical analysis of your cafe data. 

To  generate a report, set filtering by selecting radio button and specifying filter parameters in a "Criteria" control group (on above picture you see filtering set to show all records that were added today for Business Objects of type "Computer"). Then press button to retrieve data.

In case you wish to see an entire historical database, select option, it will disable filtering controls. 

Note, that selecting or specifying too broad a filter may result in significant delays depending on the size of your database and hardware performance.

Once report has been generated it can be visually analyzed by pressing button or saved in comma-separated format ( button).

To remove entries that you see in a report  from the database - press button. Those records will be permanently deleted. Deleting historical records and compacting the database will free hard drive space and speed up your system, so you may want to commence this procedure a couple of times a year.

Combining (Merging) Reports

Check box is useful when you need to combine two or more filtering options in a single report. To do so, uncheck this button, specify first filter and generate the report, then check the button, specify second filter and generate second report. You will see that your previous report stayed and new records were added (of course if the filters you specified produced results).  

To quickly asses the size of a newly generated report see "New Entries" field at the bottom of the report list. Other fields from that section also provide quick reference information:

"Total Report Entries" - Entire report size (including merged reports) "Report Amount" Sum of amounts of money for all records in a report (including merged reports) "Report Units" Sum of units (minutes, pages, etc) for all records in a report (including merged reports) "Database Size" Total number of records in the history database

Advanced Filtering

With NetSpy's advanced filtering capabilities you can greatly enhance report generation experience by specifying wild-card characters and filtering patterns. For example following query returned all records within 10 day period for all users with username beginning with GUEST ON. 

Advanced filtering option can be applied to all textual fields in a "Criteria" group. Use following table as a reference to advanced filtering special characters and search options.

Kind of match	Pattern	Match (returns True)	No match (returns False)
Multiple characters	a%a	aa, aBa, aBBBa	aBC
	%ab%	abc, AABB, Xab	aZb, bac
Special character	a[%]a	a%a	aaa
Multiple characters	ab%	abcdefg, abc	cab, aab
Single character	a_a	aaa, a3a, aBa	aBBBa
Range of characters	[a-z]	f, p, j	2, &
Outside a range	[!a-z]	9, &, %	b, a
Not a digit	[!0-9]	A, a, &, ~	0, 1, 9
Combined	a[!b-m]%	Ann, az3, armine	abc, ajia

Charts and Statistics

Internet Cafes generate plenty of data that is often better perceived in graphical rather than numerical form. That data can also be analyzed to produce key statistical values that will bring a little bit of useful science into your Internet Cafe business. Some options of statistical analysis (such as Confidence Interval filtering and Expectations) are being developed and tested as of writing this manual, so expect more in the nearest future. 

Once report has been generated, it can be viewed in graphical form by clicking button. The "Charts and Statistics" dialog that will be invoked and it will take data from current report and allow you to view it in a variety of different ways.

You can resize dialog window to scale the graph up or down.

Charting is organized in simple yet powerful way. In combination with report generator filtering you can take a look at countless of parameters that are usually hidden from you. Peak hours, daily, weekly, monthly incomes, loads on Business Object usages, average user profile, are just a small fraction of what can be done with this tool. It is impossible to describe every aspect of what can be done here, so we will concentrate on some useful examples and leave the rest for you to play with.

Chart Types

Every graph can be represented in three different ways: Bar View, Pie View and Continuous Line. Some data (such as daily amounts) is better viewed as Bar or Line, other (such as comparative income generated from different computers) is more obvious when represented as Pie.

Data values control group

Here you select your data values, or Y axis type. Selecting "Amount" will show you amount per whatever it is you selected as your X axis (hours, Business Objects, etc.). Selecting "Units" will show how many units (minutes, pages, drinks) were sold. Selecting "Event Count" will show the number of events (Logons, Logoffs etc.)

Data series control group

Altering data series will allow  you to change the vale type you want to see as your X Axis. Selecting "Time Intervals" will activate interval selection combo box where select the smallest period of time per which to summarize the Data Values as shown on a second picture.  Selecting "Business Objects" will show Data Values for every Business Object in the report. Click on "Event Types" to see how Data Values are distributed per type of the event that happened (Computer or Printer usage, Additional sales, etc. )

Lets take a real example. We want to see for instance daily sales from all computers for the last 10 days. First we need to go to the "History and Reports" page and generate an appropriate report. After specifying filter and clicking button we will receive something like this

We can already see that total amount of sales for last 30 days is $5681.23 (its been rough) and Total number of minutes computers have been used is 30750.29

Now lets press button. Graph window will appear with daily distribution of sales shown in graph format by default.

We can immediately see that the computers were making around $180 daily for the last 30 days with some fluctuations. The reason last entry is the smallest one is probably due to the fact that the day is not over yet. To exclude it we might specify our filter to generate the report for exact days. If Bar view is not the best way to look at this graph, we can select "Line" in the Chart Type group and here is what we will see.

But lets go back to "Bar" view and select "By Week Day" as a time interval. This will show us how much we were making in average depending on a week day.

As we can see, Tuesday seems to be the best weekday with Sunday lagging behind. To better asses that difference at a glance, select a Pie View.

Now we can exactly tell what is the difference between the week days (from Tue at 17.3% to Sun at 10.9%). Lets continue and see how these sales are distributed amongst computers. Select "Business Object" as Data Series and receive this

As you can clearly see COMP_1 is outperforming all other machines considerably. You might think that users for some reason like it more (may be the mouse pad is clean), to check that theory click "Units" as a Data Value.

It now becomes clear that COMP_1 is actually the least used workstation (though by small difference) this is no mismatch, COMP_1 is set to be twice as expensive as other machines (it has 21" flat screen, 2400 dpi scanner and  Adobe Photoshop) 

We hope this small exercise demonstrated some basic features of your charting tool. The number of different types of charts is big and hopefully you will soon find your own favorites.

Statistics

So far we we've been looking at plain data representation without any sort of statistical analysis. Our research shows that some level of statistical processing is helpful and even necessary for determining the optimal pricing, discovering suspicious deviations and such. "Statistical Calculations" control group is developed to provide you with that data. 

This group is undergoing big changes as of writing this documentation, so make sure to update your installation in future.

By default statistics is disabled when you first invoke Charting dialog ("None" is selected). "Min, Max, Averages" will show you the Averages and Confidence Intervals for the report you generated. "Fluctuations" dumps the percentage deviation of every data series from an average of all series.

Lets go back to the first graph we made with Daily View of sales. To do that we will set time intervals to Daily, Data value to Amount and Chart type to chart. But now we will select "Max, Min, Averages" radio button and receive the following graph.

This is an interesting report. It tells us that the best day was 2002 / 03 / 29 with sales of over $130, the worst day was 2002/03/02 at less than $90 (see last two data series called Max - ... and Min - ...). It also shows us that the simple average daily sale was about $110.

You can see other Data Series called Confidence Intervals.

Confidence Interval 95% High and Low on this particular graph shows that 95% of days the sales were between $110-$116. This is truly what you can count on as an estimation of average daily performance, because simple mathematical average does not filter out some unusual low or high fluctuations. 

Similarly, 75% and 50% Confidence Intervals show us 75% and 50% of daily sales were somewhere between 110 - 115 and 111 - 113 respectively. Now this cafe has been running very smoothly (which is good) and that's the reason confidence intervals are so close to each other. If you see big deviation in confidence intervals (especially 75% and 95%) it means your operation is not smooth and is interrupted by some internal or external events.

Confidence intervals are important values for determining the average profile of your user. And determining it may help you optimize the performance of your cafe, and simply understand the Cyber Cafe Business Model better. Applied to the users, it will show you the correct averages on amounts of units (pages, minutes) users purchased, and you will be able to say something like "95% of my customers login between 10:30 AM and 4:00PM, stay online for 15- 20 minutes, do not print any documents and buy no CDs"

Clicking "Fluctuations" will show the percentage difference between the simple average of all and particular record in the report. "Line" view is probably the best way to look at this graph.