Increase your Data Quality
You can only rely on your analysis as you are sure that your data quality is sufficient. This is why LeanIX provides features such as Fact Sheet Completeness, Quality Seal or Surveys to keep track of your data quality and completeness. This section will introduce these features and show how it is connected to the tools collaborative approach.
Keep your data quality as high as possible is one of the crucial tasks for every Enterprise Architect. Our Whitepaper "Digital Transformation - A Case for Lean Enterprise Architecture" gives an overview on the most important aspects:
Digital transformation: A case for lean Enterprise Architecture Management
There are a couple of features in LeanIX to help you maintain high data quality. This page shows you:
- How to configure Fact Sheet completeness levels
- How to use the Quality Seal to combine collaborative editing with a rigid governance process
- How to use Surveys to actively approach stakeholders to provide and maintain data
Fact Sheet Completeness
Making sure the most important data of a Fact Sheet is completed is the first step toward high data quality, one of the decisive factors in the acceptance of EAM. In LeanIX the quality of data is computed automatically on the basis of a few indicators / automated values. You are even able to set your individual weight for each element.
Completeness Level: All relevant sections as defined in the Administrator area are included into completeness calculation. Completeness calculation is now adjustable in Pathfinder so that the overall completeness may reach 100 % even if certain fields are not filled.
Last Edit Date: The last edit date shows the actuality of the data.
Check Fact Sheet Quality Seal: The *bell symbol* shows that the Fact Sheet content maybe not quality assured. This could be either the case if the end of the reminder period has been reached or somebody not responsible for that Fact Sheet has edited some data. The Administrator per Fact Sheet Type can set a check reminder period.
How to configure completeness for Fact Sheet Types
The completeness can be configured by workspace Administrators without involving LeanIX support.
So if you think that certain information on a Fact Sheet is especially important, please contact your workspace administration for the appropriate changes.
There are two ways to configure this:
- Go to the administration menu > Meta Model Configuration
- Open the Fact Sheet you want to edit > More Actions > Configuration
Both ways bring up the same page:
Click on the Edit button in the header of the page for the right column to appear
In the right column click on the scale symbol
In the "Completion weights for subsections" tab, you can assign different weight values to each subsection of Fact Sheets.
Once you are finished click on the "Show changes" button, which will lead you to a "Summary of Changes"page, where you can review what you have done and proceed to save the changes by clicking on the "Apply" button
Completion weights impact on completion score
Completion weights have a direct impact on the completion score of the entity’s (field, section, subsection) parent (field → subsection, subsection → section, section → fact sheet). Weights are strictly hierarchical, so changing the weight of a field won't affect the weight of its parent subsection, only define how much this field affects the completion score of the subsection.
Predominantly using weights of either 0 or 1 on entities helps with keeping the calculation easily comprehensible and should cover the most common use cases.
Completion score for each entity is calculated using the below formula:
Completion Score = (Σ(CW S) / Σ(CW)) 100%, where:
CW: Child entity completion weight
- when calculating the completion score for a subsection, fields can only have 2 states: 0/1 - empty or completed
- when calculating the completion score for sections & Fact Sheets - we use the completion score of each child. This can be any float number.
When an admin does not set any weight (0) to a section, then an "optional" tag appears in the corresponding subsection in the Fact Sheet. If a field has a weight (1) the "optional" tag disappears.
In addition, admins can also define mandatory fields. You can explore the mandatory fields and the quality states here.
Once a field is set as mandatory, you can see the corresponding field marked with a (*) on the Fact Sheet itself.
Subsection completion score
Proceeding with the field → subsection example, it should be noted that weight is a relative number, which shows how much completing each field affects the completion score of the parent subsection.
Let's take a subsection with 3 following fields as an example:
- Name (weight: 5) | completed
- Description (weight: 3) | empty
- Product Category (weight: 2) | completed
- Alias: (weight: 0) | empty
For most cases, using completion weights of 0 and 1 is sufficient and is the simplest way.
Note that fields with no weight (0) do not have any impact on the completion score. So, in our example, the Alias is optional, and leaving it blank won’t hurt the completion score.
When calculating a completion score, each field's weight is multiplied by field state (1 - filled, 0 - empty), summed up, and divided by a sum of all fields' weights.
In our example, calculations will look as follows:
- Name contributes 5 points, as it has weight 5 and is completed: 5x1=5
- Description contributes 0 points, as it is empty: 3x0=0
- Product Category contributes 2 points, as it has weight 2 and is completed: 2x1=2
- Alias contributes 0 points, no matter if it is completed or not, as its weight is 0
The sum of all points contributed by completed fields is 5+0+2+0 = 7.
The sum of all field weights is 5+3+2+0 = 10.
The total subsection completion score is then 70% → (7/10) * 100% = 70%.
Sections completion score
Once the subsections' completion scores are calculated, they are combined in the same way, to calculate the section’s completion score. So, if we have the following list of subsections within a section:
- Name & Description (weight 1, score 90%)
- Lifecycle (weight 1, score 50%)
- Successors (weight 0, score 50%)
- Predecessors (weight 1, score 0%)
In this case, calculations will look as follows:
- Name & Description contribute 0.9 points, as it has weight 1 and is 90% complete: 1*0.9=0.9
- Lifecycle contributes 0.5 points: 1*0.5=0.5
- Successors contributes 0 points since it’s weight is 0: 0*0.5=0
- Predecessors contributes 0 points since it is not filled: 1*0=0
In the above example, Successors subsection has no effect to completion score, as it’s weight is
0. But Predecessors subsection will lower the section’s completion score, since it has weight and is not completed.
The sum of all points contributed by subsections is 0.9+0.5+0+0=1.4.
The sum of all subsection weights is 1+1+0+1=3.
The total section completion score is then 46.6% → (1.4/3) * 100% = 46.6%.
The Fact Sheet completion score is calculated using the same approach as the section’s score.
Visibility of Sections
It is good practice to hide (deactivate) any section which is not in use in order to make life easier for your stakeholders. Those sections are still accessible, e.g. via API, and any data already stored will of course remain present.
To do so, on the same page where you can configure weights, you need to click on the section you want to hide and then in the right side menu, you can turn on and off the "Visible" toggle.
The quality seal is a great mechanism to combine collaborative editing with clear responsibilities and a rigid governance process. The main concept is very simple:
- As a Fact Sheet responsible or accountable user, it is your task to approve the quality of a Fact Sheet
- Every responsible/accountable user can edit information at a Fact Sheet without breaking the Quality Seal
- If somebody else who is not responsible (or administrator) changes your Fact Sheet (e.g. an attribute, a relation), the quality seal breaks and you get a mail.
- If a relation is edited, the Quality Seal at the related Fact Sheet break, too.
- Besides that, your workspace administrator can configure a certain time period (e.g. 30 days). This implies that after this period, your quality seal automatically breaks, helping you to continuously maintain the quality and up-to-dateness of your Fact Sheet.
How to configure the Quality Seal
As an administrator, go to the Fact Sheet Definition page.
Choose on one Fact Sheet, and under the Quality Seal tab, you can choose between the following options:
Enable/Disable ‘Broken': This option enables the Quality Seal for the state ‘Broken’ (as well known as 'check needed'. You can manually break the Quality Seal based on your permission. The Quality Seal breaks as well if you, which are not responsible, change fields of an 'Approved’ Fact Sheet. When this state is enabled, a renewal interval can be set to break the Quality Seal after a defined time as a reminder to do a revision.
Enable/Disable ‘Draft’: This option enables the Quality Seal for the state ‘Draft’. New Fact Sheets will be automatically set as 'Draft’ and its configuration is the precondition in order to define fields as mandatory.
Enable/Disable ‘Rejected’: This option extends the Quality Seal for the option ‘Rejected’.
How to set the Quality Seal
For each Fact Sheet where the Quality Seal is enabled, a little button on the top right shows the state and allows you to approve the Quality (if you are a responsible user).
If you do not have permission to change the quality seal, you will see this message instead.
What happens in case the seal breaks
In case the Seal breaks, all Fact Sheets responsibles and accountables receive a mail. Mails are aggregated if required, i.e. if 10 Seals break at the same day, you receive only one mail.
Also, the events (both approval and break) can be traced in the Fact Sheet audit log.
How to filter for Fact Sheets with broken Quality Seal
In the inventory and in the reporting, you can filter for the Quality Seal state. You can combine this filter options with all other Filters.
Further, the table view gives you a direct overview where quality seals are set. You can also export it to XLS in one click.
How to use Survey to collect and maintain data
Maintaining high data quality is a great use case for our Survey as well.
Updated 3 months ago