Sensors Analytics & Personas & Focus

l.im�&� C 1 ..:.. �ii.t.@Uf.Hei ❖ ffil:1-J�ii.t.l 0 B �ffl 1 �• B D �ii.t.]APP1 E III $Jffl 2, Iy! I0 !l!tiA:S!D 124 lll G I+iillnu:S!II H Funnel Analysis1. Tutorial Video Funnel Analysis Basic Functions Funnel Analysis Advanced Functions Before introducing Funnel Analysis, there are some basic concepts that need to be covered. 1. Step: consists of a meta-event/virtual event plus one or more filtering conditions, representing a critical step in a conversion process. 2. Time Range: The time range selected in the interface is the time range in which the first step of the funnel occurs. 3. Window Period: The time limit for users to complete the funnel, i.e., if a user moves from the first step to the last step within the window period, the conversion will be considered as successful. 2.Funnel Analysis Overview The funnel model is used to analyze conversions and churn at each step of a multi-step process. For example, the complete process of a user purchasing a product may consist of the following steps. 1. Browse products 2. Add product(s) to shopping cart 3. Check out the product(s) in the shopping cart 4. Select shipping address and payment methods 5. Click to pay 6. Complete payment The process above can be set up as a funnel to analyze the overall conversion, as well as the conversion rate and median time to conversion for each step. It can also be analyzed in depth with the powerful filtering and grouping features of Sensors Data. Application Sample of the Funnel Analysis Function 3. Funnel Interface Features Introduction3.1. Create a Funnel Click the "Create Funnel" button at the top right of the interface to bring up the "Create Funnel" panel. 3.1.1. Funnel Name Required field. Please name your funnel with a representative and friendly name. Funnels cannot be renamed within the same project. The funnel you create will be visible to other users in the same project.3.1.2. Funnel Window Period Window period: The time limit for users to complete the funnel, i.e., if a user moves from the first step to the last step within the window period, the conversion will be considered as successful. You need to choose a reasonable validity period here depending on the nature of the funnel. The default funnel validity period is 30 days. In versions 1.4 and newer, in addition to the options provided in the drop-down box, the window period can also be customized by the user, from aminimum of 1 minute to a maximum of 3650 days. 3.1.3. Funnel Steps Step: consists of a meta-event/dummy event plus one or more filter conditions, representing a critical step in a conversion process. A funnel contains at least 2 steps, each corresponds to an event (and may be accompanied by one or more filter conditions). For example, a step could be triggered by "Register" and "Invitation code used". Or, " Purchase" and "Category" equals "Women''s Clothing", etc. The order of the steps can be altered by dragging the serial number before the step. In version 1.4 and newer, you can add aliases to the funnel steps to make them easier to read when displaying.3.1.4. Add steps Add more steps to the funnel. 3.1.5. Funnel Associated Attributes Suppose we need to know exactly how a user browsed a certain item and completed the purchase of this item, we would create a funnel with the following steps: Browse Product Detail Page -> Submit Order -> Pay Order, and select the attribute of Product ID in each event, which can be used as the associated ID to ensure that the product browsed in the product detail page and the product in the payment order are identical. If the Product ID is not selected as an associated attribute, the conversion will still be considered as successful even if the product browsed in the product detail page and the product in the payment order are not the same. The attributes associated with different steps of the funnel can be the same attributes or different attributes, but the type of attributes must be the same. For example, if you use the Product ID to identify a product in the Browse Product Detail Page event, but use the item ID to identify the product in the Pay Order event, you can use the Product ID and Item ID to set up the associated properties separately.3.1.6. Save Funnel Click this button to save the new funnel. 3.2. Analyze Funnel 3.2.1. Switch Between Funnels, and View by Group or Filter In this area, you can switch between the funnel data you wish to display, or group and filter the existing funnel queries. The grouping and filtering conditions include three types, including public properties for any step, event properties for each step, and user attributes. Grouping Rules • Public properties for any step: each user will only appear in one group, grouped by the first valid value of the attribute for each user. • YY attribute of step XX: each user will only appear in one group, grouped by YY attribute value of step XX from each user''s first longest conversion status. The user will appear in an unknown group if the user does not convert to that step. • User attributes: Filter by user attributes.Filtering Rules • Common properties for any step: each user will only appear in one group, filtered by the first valid value of the attribute for each user • YY attribute of step XX: each user will only appear in one group, filtered by YY attribute value of step XX from each user’s first longest conversion status. The user will appear in an unknown group if the user does not convert to that step • User attributes: filtered by user attribute value. If the attribute selected here is a numeric type, you can customize the grouping interval. The query engine will dynamically calculate the grouping interval if it’s yet to be set up. This setting only takes effect for the current query, or if you save the query as a bookmark.3.2.2. Select the Query Time Range to Display the Funnel Chart The time interval set by default is the time range of the first event in the funnel. If "Limit window period to time interval" is selected, the occurrence of each step in the funnel is limited to the selected time interval while meeting the window period. For example, in the case of a limited time sale of an e-commerce company, a limited time sale is set from September 1 to 7, and users can enjoy the special price by completing the transaction within that period, and the original price of the product will be restored after the limited time. In this scenario, it is necessary to set this option to know exactly howmany users complete transactions during the limited subscription period. For example, if the funnel has a 4-day window and a user browses on September 6 and completes a transaction on September 8, a funnel without this option will count that transaction as a conversion in that funnel. When the option is set, it will not be counted.3.2.3. View the Funnel as a Whole, or the Conversion Details for Each Step When selecting "Trend" and clicking on a specific step, you can view the number of churned users, the number of converted users, the median conversion time and the conversion rate between the two steps in details. When clicking on the "Total Conversions" node, you can view the trend of conversion rates as a whole or as individual steps. When clicking on a single step, you can view the number of triggered users, the number of churned users, the conversion rate and the median conversion time for the two events of the current step in the table. When you select "Compare", you can set the two indicators for comparison in the "Display Settings" and compare the conversion status of the two groups in the funnel chart below. The table displays all groups, each event’s conversion rate in every step in the funnel and the overall conversion rate. 3.2.4. Funnel Chart Display When you select "Overview", you can click on each step of the funnel, and the table below provides a detailed view of event users count, conversion rates, lost users, and median conversion time for the current step. Click on the arrow in the cell before the group value to expand the current group value to view the number of event users, conversion rate, churned users, and median conversion time for the current step by day in details. When you select "Compare", you will not be able to click on the funnel chart. When there are groups, you can set two groups up for comparison. In the table below, you can view the number of users and conversion rates for each event in the current funnel by the current comparison group in details. Click the arrow in the cell before the group value to expand the current group value to view the number of users and conversions for each event in the funnel by day in details. 3.2.5. Browse User Details The corresponding numbers in the cells of the table are clickable, and clicking on themwill allow you to browse a list of the user, and even browse the behavior sequence of a particular user. 3.2.6. Display Settings When selecting "Trend", you can click on "Display Settings" to set up the indicators displayed for the current line chart. You can also click on the "Total Conversions" node on the left to setup the total conversion rate and the conversion rate for each step of the funnel. Finally, you can click on the individual steps node on the left to set up the conversion rate for the current step, the number of users lost for both events, and the median conversion time. When "Compare" is selected, clicking on "Display Settings" allows you to set up groups displayed to compare funnel conversions between the two group values. 3.3. Modify and Delete Funnel Click on the "Pencil" icon in the list of funnels to edit the funnel in the "Edit Funnel" pane. Click the "Delete Funnel" button in the bottom left corner to delete the current funnel.3.4. Save as New Funnel Go to the Edit Funnel, change the funnel name and click the "Save As New Funnel" button in the bottom right corner to save it as a new funnel. 4. How Funnels Are Calculated In this document, we will explain the calculation rules for funnel analysis in details, especially in the scenario of having filters and groups, for the users to better interpret the results of funnel analysis. In the meantime, we will provide examples of using funnel analysis for some common analytical scenarios to help users use this feature more easily.4.1. Basic Calculation Rules Suppose a funnel contains five steps A, B, C, D, E, and the selected time range is from January 1, 2015 to January 3, 2015, with a window period of 1 day. If a user triggers step A between January 1, 2015 to January 3, 2015, and triggers B, C, D, E in sequence within 1 day of step A triggered, the user is considered to complete a successful funnel conversion. If you add other steps or behaviors in the process, for example, the A > X > B > X > C > D > X > E, where X represents any event, the user will still be considered to complete a successful funnel conversion if it’s within the time limit. If the user triggers A > B >C > Ewithin this time limit, the user won’t be considered to complete the conversion of this funnel and will be recorded as a churned user for step C. Consider a more complex scenario that a user has multiple events that meet the definition of a certain conversion step in the selected time period. The event closest to the final conversion goal is considered as the conversion event and the conversion calculation will stop when reaching the final conversion goal for the first time. Assuming that the steps of a funnel are as follows: Visit home page, Select payment method, Payment successfully. The different users’ behavior sequence and the actual conversion steps (marked in red) are shown in the following example. 1. Example 1: Visit home page -> Select payment method (Alipay) -> Select payment method (WeChat) -> Payment successful. 2. Example 2: Visit home page -> Select payment method (Alipay) -> Visit home page -> Select payment method (WeChat) -> Payment successful. 3. Example 3: Visit home page -> Select payment method (Alipay) -> Visit home page -> Select payment method (WeChat) -> Payment successful -> Select payment method (WeChat) -> Payment successful. 4.2. Meaning of the Numbers Shown in the Funnel The number shown in the funnel analysis represents the number of unique users who converted or churned, instead of the number of triggered events. Within that time frame, even if a user completes the funnel multiple times, it only counts as once. 4.3. Meaning of the Filter Conditions Like other analytical models, funnel analysis also provides filtering features. Specifically, the filtering feature in funnel analysis selects users from those who have completed conversion or identified as churned users. Filtering feature in funnel analysis includes three different types of filtering.1. User attribute filtering: This filtering type is relatively easy to understand. It filters users on the basis of those who have completed conversion or identified as churned users. For example, if we add the filter condition "gender" as "male", only users with "gender" as "male" in their attributes will meet this filter condition and appear in the filtered funnel analysis results. 2. The filtering of the attribute in a specified step: Suppose we select the attribute " Payment Method" of Step 2 as "Alipay", this filtering means that among the users who have completed the conversion or identified as churned users, those who have converted to Step 2 will have their "Payment Method" as "Alipay"; if there aremultiple possible conversions, please refer to the explanation in the Basic Calculation Rules. 4.4. Meaning of Grouping Like other analytical models, funnel analysis also provides grouping feature. Specifically, the grouping feature in funnel analysis groups users from those who have completed conversions or identified as churned users. The grouping of funnel analysis includes three different types of grouping. 1. user attributes grouping: This grouping is relatively easy to understand. It is a collection of users who have completed conversions or identified as churned users, and you can further conduct a grouping based on the attributes of these users. For example, if we add the grouping condition "gender", the funnel analysis will group the users by "male" and "female". 2. Grouping by the attribute in a specified step: Suppose we select a grouping attribute "payment method" from step 2. This means users who have completed conversion or identified as churned users are grouped by "payment method" at step 2 of the entire conversion; if there are multiple possible conversions, please refer to the instructions in the Basic Calculation Rules; if the user did not convert to step 2, they were to be grouped into the Unknown group. 5. FAQ 5.1. The Difference Between Filtering Conditions Inside and Outside the Funnel Example 1: The filtering condition inside the funnel is to acquire the funnel according to the conditions, and the filtering condition outside the funnel is to filter the funnel that meets the filtering condition according to the funnel acquired beforehand. The general business application scenario is to set up the filtering conditions inside the funnel, and we recommendyou to set up the conditions directly inside the funnel to acquire the funnel that satisfies the conditions. For example, the selected funnel step is A->B->C->D, and the sequence of a user A is F->A2->A1->A1->B->C->A1->B. Event A contains an attribute of the operation system, and the operating system of A1 is Android whereas the operating system of A2 is iOS. If you set the filtering condition of the operating system inside the funnel to be iOS, the sequence filtered by system is A2->B->C. The filtering concept is to search along the sequence of the user''s behavior until it finds the event A with the attribute of iOS. If the filtering condition of the operating system added outside the funnel is iOS, because the filtering condition outside the funnel is the second filtering after the successful conversion of the user''s funnel, without adding any filtering condition, the normal funnel conversion of the user is A1->B->C, where A1 is the second A1, so the user is not filtered out when the filtering condition is added outside the funnel. 5.2. The Number of People Will Change if the Funnel Is Refreshed Mandatorily （ 1） If the data changes only at the first refresh, and the number of people no longer changes after multiple refreshes, it is due to Sensors Data’s query caching mechanism, and you the data after the refresh shall prevail. （ 2） If the number of people keeps changing after multiple refreshes, it is due to the unstable sorting caused by the same occurrence time of adjacent events in the funnel. You can consult the personnel responsible for event-tracking of the corresponding events to adjust the timing of the event reporting to ensure that the two events occurred at different timing. If none of the above applies, you can contact personnel on duty from Sensors Data. 5.3. The Number of Users in an Event Within the Funnel and the Number of Users Triggered in the Event Analysis Do Not Match （ 1） If the number of users in the first step of the funnel is not the same as the number of users triggered in the event analysis query: If you add a filtering condition outside the funnel in the process of funnel query, the number of users in the funnel may be different from the number of users corresponding to the same filtering condition within event analysis. You can add the filtering conditions inside the funnel and compare the query data. If the results of the query, which its filter is applied inside the funnel, are consistent with the event analysis, the query results are normal. For the specifics, please refer to the difference between in-funnel and out-funnel filtering conditions in Question 1. （ 2） If the number of users in the events that are not the first step in the funnel does not match the number of users triggered in event analysis: The difference is normal, because the number of users from an event in the funnel is the number of users filtered out after satisfyingthe process of the funnel. For example, if the funnel process is A->B->C, and the number of users of event B in the event analysis is 100, the number of users who trigger event A and event B after in the funnel analysis would be 20. 5.4. The Daily Conversion Rate Is Not as High as the Overall Conversion Rate For example, let''s say there are 3 people in the 3 days fromNov 1st to Nov 3rd, and the funnel process is A->B->C. On Nov 1st, all 3 people trigger event A, but only the first person completes the B->C conversion within the window period; on Nov 2nd, all 3 people trigger event A but only the second person completes the B->C conversion within the window period; on Nov 3rd, all 3 people trigger event A but only the third person completes the B->C conversion within the window period. The conversion rate by day is 33% for Nov 1st, 33% for Nov 2nd, and 33% for Nov 3rd. As a whole, the conversion rate for the 3 days from Nov 1st to Nov 3rd is 100%. 5.5. Overall Conversion Rate Is Not as High as The Daily Conversion Rate For example, for the 2 days from Nov 1st to Nov 2nd, the funnel process is A->B->C. On the first day, three people, a, b, and c, triggered event A, and only two people, b and c, completed the B->C conversion within the specified window period, with a conversion rate of 66.6%. On the second day, four people, b, c, d, and e, triggered event A, and two people, b and c, completed B->C conversion within the specified window period, with a conversion rate of 50%. Altogether, five people, a, b, c, d, and e only visited, b and c placed an order, and the conversion rate is 40%. Heatmap Analysis Explanation Video Explanation Video on Heatmap Analysis Login Page Thermal Analysis Overview Heatmap analysis is mainly used to analyze the user''s clicks and depth of reach on the web page, and present it to the user with intuitive visual effects. For more information about theapplication of click-through analysis, please refer to: How to use click-through analysis to optimize product experience Introduction of Heatmap Analysis Function Original Page The description of this page is as follows, for click maps: 1. Display Content: There are two categories, original pages and page groups. The original page is used to analyze the click situation of a single page, while the page group is used to analyze the browsing and click situation of a series of web pages with similar interface structure as a whole (for example, the product detail page of JD.com can be analyzed as a page group as a whole). 2. Event Filtering Conditions: Like all other analysis functions of Sensors Analytics, Heatmap Analysis can also be filtered according to the attributes of events, and only look at events that meet the filtering conditions. It should be noted that the filtering conditions of Click Analysis are the public attributes of the events Web Element Click and Web Browsing. 3. User compliance: Like all other analytics functions of Sensors Analytics, Heatmap analysis can also be filtered by user attributes. 4. Select the time of the occurrence of the event Web element click and analyze only the user clicks during this period of time.5. This area is sorted by page views by default, and can also be sorted by the number of users who viewed the page and the number of clicks on the element. The page name refers to the title of the page we acquired. 6. Views: The PV of this page, also known as the total number of page views of the event Web. 7. Number of users: The number of clicked UVs for each interactive element on this page, which means the number of users triggered by the event Web element click. 8. Number of clicks: The number of clicks on each interactive element on this page, i.e. the total number of clicks on the event web element. Page group The description of this page is as follows. 1. Define a name for the page group 2. Background page, specify a specific page address to be used as a template page for the click-through image demonstration 3. Select the address of the page to be added, you can use include and regular to match 4. You can select multiple filter conditions 5. Save this page groupInstructions for Using the Click Map Here we introduce the specific use of the click analysis function, using the official website of Sensors Data as a sample page (the data of the click is simulated). 1. Indicates the clicked interactive elements, where 24.70% of the clicks are shown here. When the mouse is hovered over these clicked buttons, the click details information box is displayed. And if the mouse clicks on this interactive element, the corresponding interaction will follow the logic of the original page 2. Current element content, which refers to the text content inside this interactive element 3. Number of clicks, how many times this button has been clicked 4. Click-through rate, the number of clicks on this element / the number of views (PV) of the whole page 5. Click-through ratio, the number of clicks on this element / the total number of clicks on all visible elements in the whole page6. History content, which means that the most frequent value of this button''s history is taken. For example, in the case of news, the headline position is different every day 7. Clicking on the list of users to see which specific users have clicked on this button 8. Click-through rate corresponds to the illustration 9. Switching the display options in different screen sizes 10. After clicking on the interactive element of the current page for page jumping, you can click here to return to the previous page 11. Open in a new window and show the click graph directly in the original page, which is a more natural way to view the data 12. Dropdown toggle between click-through and reach charts 13. Different options for switching the click map 14. Share the current page 15. Refreshing the data of the current page 16. Close toolbar Switching Between Different Versions of the Click Map (Supported by Version 1.12.1 or Above) (A) Differences Between the Two Versions of Click Maps Version 1 Click Map Styling on the click element itself adds style to the element by adding after and before pseudo elements, and will change all a tags to inline-block style. If the after and before elements on the page are already styled, this may cause a style conflict, resulting in some elements on the page not having clickable images, or a change in page structure.Version 2 Click Map A click map is generated at the top of the page based on the relative position of the clicked element on the screen, and is re-rendered each time the scroll bar is scrolled, or the page size is changed. The second version of the click map will not modify the original style of the click element, which can effectively avoid the problem of style conflicts. The second version of the click map does not render hidden elements, but it will render some elements that are hidden by parent element overflow:hidden, etc., which cannot be determined whether they are hidden or displayed. (B) Method of Switching Click Maps 1. Switching via the first version/second version button in the upper left corner 2. Switching via keyboard shortcuts ''z'' key: the first version of the click map ''x'' key: the second version of the click map Method of Refreshing the Click Map The second version of the click map renders the clicked elements based on the relative position of the screen, thus the click map is re-rendered each time the scrollbar is scrolled and the window size is changed. Modifying the rendering delay can be done by referring to the JS SDK documentation for the relevant parameter renderRefreshTime.Version 2 click maps will only render elements that are not hidden. Some elements that are manually controlled to be hidden/shown (e.g. drop-down menus, etc.) need to be manually refreshed when the element is displayed. 1.Refresh the page via the refresh button in the upper right corner to re-render the click map data. 2. Refresh the click map data by using the ''r'' key on the keyboard. Notes The second version of the click map can solve the problem of style conflicts in the first version of the click map (such as after, before pseudo-class conflicts, inline-block caused by style exceptions, etc.). The second version of the click map cannot determine whether certain elements are shown/hidden. For example, by changing the position of the child element to produce a show/hide effect through the parent element overflow:hidden, and by changing the height of the parent element to control the show/hide of the child element, etc. The second version of the click map will render all elements judged to be displayed on the page, so if the second version of the click map affects the display of your page, please switch to the first version of the click map. Translated with www.DeepL.com/Translator (free version) Share Button The current pre-configured $WebClick click events are data collected under various screen sizes of various devices of users, but the display of click analysis is shown under one screen (usually PC side). Option 1: Click the share button and scan the QR code with your phone to view the click analysis on your mobile device. Later we will develop a display solution based on multiple screen sizes, so please stay tuned.Instructions for Using the Reach Chart Effective Dwell: Focus on the web area without scrolling, during which the mouse can move, click and perform other operations. Valid Dwell Time: the dwell time is longer than the specified time, and the default is 4 seconds in javascript sdk. This parameter can be set. If a page scroll occurs and the previous page dwell is a valid dwell, which means that it exceeds the default 4 seconds or a custom time, the javascript sdk will send a page dwell event. Reach rate is the percentage of users who eventually reach a location on a web page given the current filtering conditions. The reach graph is generated by calculating the depth of reach of a page by counting where the user ended up before exiting the page. The reach chart can be used to analyze the depth of users'' browsing on detail pages, landing pages, and other types of pages to help optimize the design of the content and structure of the page.Using the Sensors Data blog as a template page (as simulated data), here we introduce the reach analysis. 1. Indicates the proportion of users (expressed as a percentage) who reached the current location given the current filtering conditions 2. Baseline for the current page reach 3. Refresh the data on the current page 4. Collapse (drop down) the action toolbar The data in the reach graph, calculated by uv, is based on the maximum height of a person''s reach for this page. FAQ To be added">