Knowing how to use benchmarking for performance improvement does not guarantee that businesses will use it...
Benchmarking Survey Tool Development 2019
November 2018, Combo Benchmark 2.0 was created and received a rename: Benchmarking Survey Tool, version 2.0. Now, we are a year further and have issued several releases to execute our mission, which is to make sure that you are successful at benchmarking. To help with that, we offer you a standardized survey benchmarking system, that can be completely customised to your wishes.
What has changed between the introduction of 2.0 and now (2.0.10)? You can read that in this years overview.
Multilingual front-end, system in 2 languages (Dutch and English)
It is possible to use many different languages in the front-end, while the back-end is limited to Dutch and English. In the account settings in the back-end, you are able to select which languages you want to use. These languages are then available when you implement all surveys.
Afterwards, you can select per front-end (so per participant environment) which language you want to use. This way, you can create a front-end per language or one specific front-end with a language selection.
Roles and memberships expanded
Roles and memberships are now available for use and selection in all functions. Now there can be more variety in authorisation per user and participant respectively.
Logging of mutations expanded
A lot of changes in the master data were already logged. This now has been further expanded upon so that nearly all types of mutations are now logged. All sent emails are now also logged. All data entry by users are now logged as well.
With surveys at a greater scale, it would sometimes take a considerable time to retrieve the result of one participant as a benchmarking survey manager. This is because results were calclulated in real-time. Because of this, data that was necessary multiple times, was recalculated per retrieve which could lead to poor performance in regards to time results.
We have modified this and determine results when handing in or validating questionnaires, modifiying referral groups, variables and recalculating results by caching the results. This way, the tool makes sure all results of standard retrieves have been calculated before the retrieve is made.
It has been made easier to connect users and participants together by the use of an import. A new option has been added in the account menu in which you can export a list of all participants with the linked users. The exported list can then be edited, and imported again.
Deduplication of participants by import improved
While importing participants and users (through import definitions), free fields can now also be used for deduplication.
It is now possible to anonymise participants of a survey. A new, separate symbol ? has been added to Account - Participants. Click on the ? to anonymiseds a participant.
An anonymised participant receives a hash-code and remains linked to surveys. All users to the participant will be unlinked.
Deleting masterdata and surveys
It is now possible to delete entire surveys. Doing so will delete all survey data and all master data can also be deleted, including user logs. With this, the Benchmarking Survey Tool is fully GDPR-proof.
Configuring password strength policy and necessity emailaddres activation
It is now possible to determine the strength of a password, a user is required to have. You are able to pick from the three following options:
- Weak – minimum length is 6 characters
- Normal – minimum length is 6 characters and has to contain at least one number or special character.
- Strong – minimum length is 8 characters and has to contain at least a number, a letter, a capital letter and a special character.
It is now also possible to make activation non-mandatory when changing an email address.
2 factor authentication benchmarking survey managers
Benchmarking survey managers are now able to configure two-factor authentication or two-steps authentication. With this you are now able to protect your account better. This setting can be found in “my account”.
You have the choice between a time-based one time password (TOP) or a universal 2 factor authentication (U2F) via a dongle.
The first option works with your mobile phone through an authentication app, i.e. Google Authenticator.
The second option works through a key, i.e. the Ubikey. This doesn’t work yet on all browsers as this option is currently still in BETA. While it can be used, you might encounter issues.
General surveys and performance benchmarks integrated, business scans added
We are currently working on integrating all surveys into one survey platform, where each survey makes use of the same technique and by the use of settings, the functionality per survey is determined.
Within each type of survey, there are different settings possible and available. You determine which function you need per survey. Whilst creating a survey you still have multiple choice from three option types.
While configuring a survey, you select which functionality you want to use in the survey, wherein the choice is different per each type of survey.
- General survey – contains one questionnaire and one period (simple query), sign-up (before or after), direct accessible survey (through an individual or anonymous link), approval by participant.
- Business scan – contains one questionnaire and one period (simple query), sign-up (before or after), direct accessible survey (through an individual or anonymous link), approval by participant, validation, groupdata, constants, reports. The report can be sent to the participant directly after completing a survey.
- Performance benchmark survey – contains one questionnaire and one period (simple query), sign-up (before or after), direct accessible survey (through an individual or anonymous link), approval by participant, validation, group data, constants, reports, dashboard, referralgroups, data analysis, data export by the participants, fixed value input and much more. The report can be sent to the participant directly after completing a survey but can also be generated manually after logging in.
At the settings of the pages in the front-end you can also select the pages that belong to the chosen function.
Copying a survey with all underlying elements
It is now possible to copy an entire survey with all underlying elements, this however will be without any data. While using the copy function you can select which elements you wish to copy.
Reports and analyses are also copied and the variables in the elements are immediately converted to the new variables. With this you have a fully functional copy so that you can easily test modifications.
Deleting a period has now been made more difficult
Deleting a period has now been made more difficult by adding a confirmation prompt. With this it is now harder to accidentally delete a period with loss of all the underlying data of that period. You now also have a direct view of which data is to be deleted. You will also view what each item is linked to and what linked data will be deleted.
When manually creating a period it was necessary to log out and to log back in to enable the period to be used. That flaw has now been fixed. A created period is now immediately visible and useable and can be enabled for use by participants.
Period ID’s are now continuously determined instead of determined per survey. Because of this, it is no longer convenient to refer to period ID’s. Instead, a period can now be looked up by using the start date.
A period can now also be blocked. Blocked periods are not visible to participants.
It is now also posible to configure the name format of a period. By default, a quarter will be given the name “Quarter N JJJJ”.
By filling in the name and the short name with codes such as “Year” and “Quarter” the name format of a period can be configured differently.
Stricter handling of type variables
The system now handles type variables more strictly. This has consequences for references to variables in formulas. A variable of the type ‘numeric’ can only have a number as reference, without quotes, for example #revenue>30000. A variable of the type ‘text’ can now only have text as a reference, with quotes, for example #country==’United Kingdom’. A variable of the type ‘boolean’ can only have true or false as a reference, for example #include==true or #include==false. Ones and zeroes (#include==0) as the result of a boolean is NO longer allowed.
List questions with sliders and images
We have added a number of new question types to make surveys more flexible when it comes to look and feel. For example, it is now possible to work with sliders.
It is now possible to define a survey question with clickable images, yes or no with radiobuttons and explanations per choice.
For questions in a survey, it is now possible to configure conditional colour formatting. This can be applied to a checkbox, as well as the cell that the checkbox is positioned in. This way, conditional colour formatting can also be applied to the background.
Mandatory questions no longer receive a red background but instead a red border. Because of this, it is now easier to configure a custom background, if not conditionally.
Elements of a question in a survey have been simplified and configureable for export
Elements of a question in a survey have received a continueous ID, next to the value and the name. Because of that it is not possible to only fill in the value and to leave the name empty. The value will then be shown to the participant in the questionnaire.
It is now also possible to configure a question in a survey, if underlying elements also have to be included in an export. With certain lists, (e.g. countries) this resulted into a very large export, without any added benefit.
In conclusion, the elements of a survey question can be configured and ordered directly while creating, or modifying a survey question. This results in it no longer being necessary to save the survey question first and then entering the elements.
Conditional elements of survey question
Elements of survey questions have received a visibility check condition. Hereby it is possible to determine the online elements in real-time based on an answer given to a prior question.
Imagine you ask a participant in which country he lives and then next in which province. You can make a list question with countries followed by a list question with all the provinces of all the countries in the prior question. Per province you can then select by the use of a condition, when it is allowed to be shown to the participant.
If the participant selects ‘the Netherlands’ in the country question, the participant will then only be shown Dutch provinces in the follow up question.
Constants now included in questionnaire
It is now possible with the help of a function to include the value of a constant in a survey, inside of a calculation. You can for example use constants to work with exchange rates. The constants are filled in by a benchmarking survey manager in the back-end. Based on the conditional calculation of the getConstant-function, the system selects the correct constant.
Printout submitted survey by benchmarking survey manager
It is no longer required of a benchmarking survey manager to emulate a participant, in order to make a printout of the survey questionnaire. Through the following options: results per survey, results per period, results per participants, the benchmarking survey manager can now directly see the survey results of each participant and export it.
Acceleration of export data and validation
We have sped up the process of exporting data by no longer performing calculations and exporting the data as it was calculated before the export.
Validations have been accelerated as well. Loading of validation rules used to take a long time. Large surveys with a lot of validation rules could lead to an increased loading time for the list “to be validated” participants.
By optimising the rules we have managed to accelerate the loading time of lists and validation of surveys.
List yes/no questions optimised and included in export
All elements of multiple choice questions are now included per element in exports of data.
It is also now possible to refer to elements with expressions in calculate questions or conditions with the help from the “isValueElementSelected” function. E.g.: (isValueElementSelected(getQuestion('ActImprCS'), '1')
Because of this, it’s no longer necessary to create separate yes/no questions anymore for each option.
Period validation expanded with further questions and sorting
It is now possible to select up to 4 (instead of 2) periods in the validation – period comparison. Additionally, all columns can be used to sort. This simplifies the intertemporal analysis for benchmarking survey managers, and makes it easier for benchmarking survey managers to see differences of a participant between periods.
Determining reference group in reports per period
We have made it possible to determine a reference group based on the period of a report. In a report, the reference group can be automatically determined based on the settings of the report groups. Because a participant can be a member of different reference groups of the same class in different periods, this sometimes caused issues when determining the correct reference group.
By connnecting a report period to a report group, the correct reference group will be selected in the report.
Automatically determining the period in report
It was already possible to determine a period in a report based on other report periods. This calculation has been made more dynamic so that period is already determined on-screen and can be made visible to the participant.
Order of trend diagram and trend table configurable
In our customer feedback the need has arisen to be able to change the order of trend tables and trend graphs. The order was automatically set as young to old but it’s now also possible to change the order from old to young.
Conditional elements in dashboard
Conditional elements could already be included in a dashboard. We have made it possible to show the elements of a dashboard based on a condition. This way the dashboard can be further customised to the data of a participant and their roles and memberships.
A large part of the e-mail workflows within a survey can now be configured per event. Per event type can be configured what action needs to be executed:
- Sending an email to a participants, at sign-up, starting a survey, handing in a survey, approving a survey, etc
- With a report
- With the copy of a survey
- With both
- With logincodes
- Sending an email to the benchmarking survey manager(s) at signup of a participant, starting a survey questionnaire, submitting a survey questionnaire, approving a survey questionnaire, requesting a membership, etc.
Certain changes are relatively simple to implement, others take longer in time. Thusly it has taken us a longer time before we decided to implement panels. This is because panels have a lot of difficulties and can have an influence on all executions and calculations, including impact on the performance.
Panels are groups of participants which all have participated in two or more periods. This concerning validated surveys. There are no checks inside the surveys, if a certain question has already been answered in the selected periods, or if an out-of-group value condition applies to a question.
If a panel is chosen by the participant or selected by the benchmarking survey manager (e.g. when exporting groupdata), then the group numbers will be calculated for all the participants that have participated in the selected panel period.
x = validated questionnaire.
Panel P1 to P8 exists out of four participants, B, C, E and H. You can use this panel to for a trendgraph of P1 to P8.
Panel P1 to P4 exists out of eight participants, B, C, E and H. You can use this panel to for a trendgraph of P1 to P4.
Panel P7 to P8 exists out of five participants , B,C,E,G and H. You can use this panel to gain an insight into the development between these periods.
The benefit of this is that there is no need anymore to create separate reference groups for this. Create the panel and select which periods you need. Enable the panel for use by participants in the panel settings.
In the dashboard, analysis can be displayed to compare the numbers of panels and non-panels.
Multiple panels can be used parallel to eachother, where the panels can be determined on the basis of which periods are used in the report. Use the following formula for this: getPanelWithPeriods(#period1,#period2) (for two periods.)
Returning to function after completing survey.
Per request, the tool will remember where a function is started from, so that when the function is finished, the tool puts you back in the last known position. This makes it possible for example, to make full use of the home page. After filling in a questionnaire, the participant will be returned to the homepage instead of the list with periods.
“Thank you” page after incomplete survey.
Per request, the thank you page that participants see after handing in an incomplete survey, has been returned.
Prior periods of texts and survey questions
Next to numeric and arthmetic questions, it is now also possible to display the value of text questions and list questions of the previous periods, to participants in the surveys and when validating to benchmarking survey manager s.
Deviant URL’s now have more configuration options per page. Even the language can be different per url.
Exporting the same selection.
With all forms of export, the benchmarking survey manager can determine the same way which periods, reference groups and panels can be selected by the participant. This way it is easier to use all forms of export, because they can be limited in the same way.
Expressions calculated on-screen, incl. text calculation
Working with two screens is now better protected against overwriting data
It was possible that problems could arise when a participant was working on two screens concurrently. This especially if the survey questionnaire happens to be a particularly long one. In one screen the questionnaire would be submitted, whilst other screen displayed the questionnaire as still open.
This could lead to loss of data. We have resolved this by making sure that when a questionnaire is being submitted and processed, the data can no longer be modified, even if another questionnaire is open in a different screen.
Front-end pages fully dynamically configureable
The menu and submenu structure are now dynamically generated. A page will be visible in the menu structure if it is configured that way. A page is linked under a submenu if it’s under a menu structure.
It is possible to make a page publicly acessible without the need to log in. With this you can make a full website (an extranet) that is accessible for everyone.
Survey pages no longer have to be linked under the menu item survey, but can now linked under other pages, e.g. “Home”.
Questionnaires in a performance benchmarking survey now directly accessible through a link
You can configure an input page of a survey on ‘public’ or ‘logged off users’. This makes the input page of a performance benchmarking survey even easier to access as users do not need to be logged in. Just send them the direct link to their questionnaire of a period.
Checking signups on address and zipcode.
You can configure an extra check for the signup process where you can check for duplicate users by checking email adresses or the combination of address and zipcode. This way you prevent users from signing up more than once with the same address.
Participants with multiple locations can this way signup with multiple locations and make use of the same email address. In a performance benchmark survey, different users are created for every location and the benchmarking survey manager can then connect the locations to one user. In the other surveys no users are created by the participants.
Signing up retroactively.
Signing up before a survey can be filled in, can provide a threshold for a some participants. Because of this we have made it possible to signup retroactively. After handing in the the survey, the user can still decide to sign up, for example, to receive the report.
Should the participant decide not to sign up, they remain an anonymous participant. Should they choose to sign up, the participant data is overwritten and a user is created.
To enable retroactive signups, there have been extra settings added to the “thank you” page.
With this you can make certain fields in the “thank you” page mandatory or non-mandatory. There is no extra check on duplicated emailaddresses or adres+zipcode combinations.
For our plans for 2020, we want to add new types of survey, such as employee survey, customer survey, member survey (customer survey for associations), culture survey, salary survey and job reward survey. The surveys for this are generally not too complicated to make. It is usually about a specific functionality for a survey, such as being able to to create employees, customers and members by the consumer of a survey or just being able to create standard functions.
Functionality for these types of survey will be necessary in reports as well.
We are very curious to hear your thoughts. If you have any wishes, remarks, ideas, references, potential customers, please relay them to us. We value your inputand will do our best to include your wishes in a following release.
Can you recommend us to your associates? Please let them know how to contact us and state that we were recommended by you. We would truly appreciate this.
Other articles you might like
Now one of the things that many associations are going to run into is errors or differences in results when compiling benchmarking results...
One of the biggest challenges in our experience for most associations is providing market reports and effectively communicating...
How we support you
We believe that benchmarking could be much simpler, cheaper and more efficient through professional automation. And we believe you can add much more value for your clients with intelligent automation.
- More then 15 years experience with benchmarking
- Help on implementing the questionnaire with you
- Support on analysing and reporting the results
- Training you to run your own benchmarks