Deciding which enterprise intelligence (BI) answer to implement in 2022 isn’t any small feat. The business commonplace is increased than ever: As soon as extremely prioritized options comparable to drag-and-drop functionalities and fascinating chart visualizations at the moment are thought-about commonplace, thus carrying much less weight within the decision-making course of.
So, which options shouldn’t be missed when looking for the most effective BI device right now? What are essential concerns that may enhance the convenience of adoption for each information engineers and enterprise customers?
Let’s study the highest seven must-have options of a wonderful, modern-day BI device.
1. Multitenancy That Permits Analytics Throughout A number of Prospects
When distributing and doubtlessly promoting your information product to clients, it’s important that every buyer can solely see and entry its corresponding information.
In contrast to different BI instruments, GoodData provides easy and painless sharing of personalised information for a number of, completely different tenants. The principle benefit originates from one easy supply of reality: the prevailing logical information mannequin (LDM), which is loaded within the desired information construction. This supply is replicated for every of the affected tenants, whereas displaying solely user-specific information. (For extra particulars, dive into our documentation about multitenancy or a extra particular implementation use case shared inside our GoodData Neighborhood portal.)
So as to present GoodData’s multitenancy in motion, we created a Github repository. With the COVID-19 dataset, you possibly can section the specified info into a number of workspaces (user-specific areas the place metrics, dashboards, and stories are saved). For the COVID-19 information within the Czech Republic, the workspaces characterize particular person counties. Observe that every one insights, metrics, and LDM definitions apply to child-workspaces (workspaces with a dependency to a grasp workspace), as nicely, whereas filtered to point out solely user-specific info.
Moreover, your clients could wish to benchmark their information in opposition to information from different organizations. GoodData permits comparability between aggregated information of various shoppers or business requirements. Including a brand new dataset with aggregated information — to help your benchmarking wants — will do the trick. (Dive deeper by way of this GoodData Neighborhood thread.)
2. A Logical Information Mannequin That Speaks the Enterprise Language
Information modeling capabilities are vital for any aggressive BI device obtainable out there. The choice to visually and simply determine the relationships between designated entities is an important prerequisite for any BI undertaking.
Straight linked to the aforementioned multitenancy use case, one in all GoodData’s key advantages is its method to information modeling — as demonstrated by its LDM characteristic.
Primarily, an LDM describes units of used information in a significant (aka logical) approach. The mannequin might be arrange independently to a supply database that establishes a basis for elements of the semantic layer in information administration techniques. When created correctly, an LDM allows the creation of latest metrics, stories, and insights with out counting on difficult joins or lookups.
The creation of an LDM could require some stage of technical experience in addition to enough enterprise acumen. Nonetheless, as soon as created, it offers enterprise customers and BI analysts with an optimum background to create all desired metrics and insights with out the necessity of adjusting the prevailing information relationships. Primarily, it focuses on an important data-related job: deciphering information and utilizing it to assist your group. The metrics, insights, and dashboard sitting on prime of the prevailing LDM might afterward be redistributed to a number of shopper workspaces, the place particular person changes of your shoppers/finish customers can happen. (For additional info, please learn our weblog publish on logical information fashions.)
Right here is an instance of an already present LDM (ready within the GoodData.CN Neighborhood Version demo).
3. The Skill to Join Practically Any Information Supply
You in all probability know the story: The enterprise decides {that a} new software program needs to be applied, they usually wish to observe information coming from it … ideally beginning yesterday.
Our current introduction of Dremio integration aligns with GoodData’s imaginative and prescient: connecting just about any database to your BI answer shouldn’t be an issue. Dremio is the newest addition to a catalog that features Snowflake, Redshift, BigQuery, PostgreSQL, and Amazon S3, amongst others. Plus, not solely are these prepared out of the field, it’s additionally straightforward to arrange. Via the GoodData API, you possibly can simply create a connection to any most well-liked schema within the present database, all whereas making certain you get your information at close to real-time velocity.
Do that for your self through the use of the GoodData.CN version and replicating this POST command:
{
"information": {
"attributes": {
"title": "prod-db",
"url": "jdbc:postgresql://localhost:5432/prod",
"schema": "public",
"kind": "POSTGRESQL"
},
"id": "prod-ds",
"kind": "data-source"
}
}
4. A Metric Editor With Clever Question Completion
The flexibility to carry out complicated calculations and aggregations is essential to any BI device. GoodData’s metric editor offers the top person with the power to create customized metrics for reporting. The Multi-Dimension Analytical Question Language (MAQL) is the engine of the machine.
The important thing benefits of MAQL embody the next:
- No joins or sub-joins as MAQL works on prime of LDMs and its queries are context-aware.
- Any metric might be instantly used for reporting, reused once more, or deployed to assemble different metrics.
- MAQL makes multidimensional evaluation easy by abstracting any information complexities. You shouldn’t have to specify the very fact or attribute origin as it’s finished routinely for you.
After the creation of a set metric, the power to format appropriately is essential. GoodData offers you with out-of-the-box options to pick whether or not your metric must be a forex, a quantity with a number of decimal factors, or some other format you want. (For extra info, go to GoodData College.)
5. The Skill to Drill to URL
Normally, driving engagement together with your information product might be tough. One foolproof approach to take action is by maintaining person expertise prime of thoughts. For instance, your finish person could spot an irregularity or an fascinating growth in a report, after which wish to additional study this subject within the supply software program. By offering the power to drill to URL, you possibly can assist facilitate — and streamline — this course of.
Let’s examine how straightforward it’s to arrange in your dashboards:
6. The Skill to Attribute Filters That Are Metric-Particular
Making a report — which goes to comprise a metric filtered by some attribute values — is nothing new. When choosing a metric, you’d simply apply a filter within the Analytical Designer (GoodData’s atmosphere that permits customers to create their stories and visualizations, in addition to additional information exploration) after which alter primarily based in your preferences.
Nonetheless, what should you’d like to pick the identical metric a number of instances, however with every time filtered by one thing else? GoodData has you lined: As a substitute of making a brand new metric in MAQL with the filter lined there, you possibly can merely choose attribute values, which is able to impression solely the chosen metrics.
This makes it straightforward to create a brand new fast perception (with out the necessity of producing a brand new metric); it additionally offers the power to match a number of attribute values in opposition to one another.
The finalized perception could look one thing like this:
7. Completely different KPIs Affected by Completely different Dates
When creating a brand new dashboard, the BI analyst will notice that the stories added to the dashboard are more likely to be filtered by a number of completely different date dimensions.
Let’s take into account a report that exhibits what number of leases of DVDs your online business accrues, after which add a second report that demonstrates what number of DVD returns occurred monthly. We want to present each of these subsequent to one another on the identical dashboard, however each will need to have a unique date dimension. GoodData permits the person to simply choose which date dimension (current within the information mannequin) goes to have an effect on the filtering’s finish consequence.
Each of the stories outlined above are current in the identical dashboard, and customers can filter them by completely different information dimensions. The identical goes for filtering the dashboard by attribute values.
Subsequent Steps
When contemplating the mixing of a brand new BI device, one ought to pay attention to its technical necessities and options. Understanding the present and, extra importantly, future use instances on your information is essential, too. Falling into the entice of tantalizing visualization widgets and different superficial options could show expensive; the precedence at all times must be to offer the enterprise with correct and quick info whereas making the lives of information engineers and BI professionals as straightforward as attainable.
Header photograph by Tara Winstead from Pexels
