Tearing Down the Barriers to Insight with No-Code Analytics Solutions
In this post, I'm going to address a common situation that companies find themselves in when looking to expand the use of analytics at their organization. Not having the proper tools, capability and processes to take advantage of their data.
First, we'll explore a new type of software called business process analytics which looks to bring to light operational inefficiencies through process mining techniques. I'll review Celonis, but there are others in this category as well.
Finally, I'll explore the concept of data democratization and give a critical review of a category of tools that look to knock down the barriers to data science. I'll explore KNIME, but other solutions are very similar in capability.
Please note that this is not a recommendation of any specific platform, just a review of no / low code solutions with an emphasis on two solutions. The best solution for your use case may be different.
Let's get started!
No and Low Code Analytics Solutions
A variety of solutions claim that users can access data and analyze it with little or no coding experience. The promise of data democratization is valuable to businesses because more decision makers can have the data, they need to reduce the risk in potential decisions that may benefit from a more informed viewpoint. These solutions achieve no/low code solution by obfuscating the ETL, data mining, machine learning and statistics through their software. A drag and drop interface can replicate advanced SQL joins or run algorithms that help end users mine data for insights.
The two solutions we’re comparing in this review are Celonis (a low code process mining solution) and KNIME (a low/no code data workflow tool.) Both tools can give business users and decision makers direct access to the data that will assist in reducing the risk of decisions; but there are caveats for each. In this review, we’ll identify how they are related, but also explore the strengths and weaknesses of each tool.
Metaphorically, like a living organism, businesses have a variety of processes that help them achieve their goals. Revenue management processes like accounts receivable and resource management allow for the exchange of assets to generate profits. Processes like supply chain management allow for the acquisition of goods to sell to fulfill orders. These disparate processes are critical and are difficult to analyze in aggregate because they have different, specialized, purposes.
Celonis claims that their software can help companies analyze their processes by mining those transactions and identifying how they are working together to achieve organization-wide goals. Their claim is that the software is like a “brain for your processes.”
Their claim is that the software is like a “brain for your processes.”
Celonis appears to do this by connecting to these disparate process management systems and connecting common keys in events logs to monitor different outcomes that each step of a business process can take. Analyzing event time stamps, machine data – the tool creates a spaghetti diagram of all the paths that the process can take.
With Celonis, possible analyses include time series analytics, simulation, clustering analyses, and outlier detection to help teams understand the most common paths in a process, but also identify when processes go wrong.
Competitors to Celonis: IBM Process Mining, SAP Signavio Intelligence, Oracle BPM
Celonis is excellent for exposing the variation in processes and understanding the true cost of doing business. It enables decision makers to step back from a specific process and look at it in aggregate and how it affects all parts of any organization.
Celonis is a no / low code process analytics tools which decreases barriers to knowledge across and organization; and democratizes analytics capability beyond a traditional analytics team.
The tools graphs, charts and diagrams are easy to interpret and create strategies from.
Celonis requires careful setup and planning to get going. Environments can be difficult to manage which means that a customer will need technical support or potentially a consultant to assist in its implementation.
Once implemented, the solution will likely require dedicated staff to support and maintain.
The user base of this new software is smaller than other types of software. It can be difficult to source employees with experience using the tool.
The KNIME software is also a low/no code tool for helping a company solve an analytics problem. This software is targeted at the issue of access to data within an organization’s databases. Typically, analysts with SQL knowledge will mine those databases to produce reports that service the needs of the organization. KNIME works to help employees with less skill in SQL and database querying languages gain access to report creation capability. Ultimately, KNIME is a data workflow tool that allows analytics teams to democratize data by improving routine manual processes in the data workflow process.
KNIME works to help employees with less skill in SQL and coding skills gain access to report creation capability.
The software achieves this by using visual programming with drag and drop icons that represent different capabilities typically codified in an SQL query. An analyst can join two (or more) data sources together by literally drawing a line between the sources and specifying the type of join in the software. These types of joins, aggregations and data processing workflows can be automated, replicated and shared across a business.
In addition to workflow automation, the tool also allows for light modeling and data visualization within the tool to form an end-to-end analysis that has those same properties (automation, replication, and socialization.)
KNIME is targeted to solve for data democratization and better access to analytics across an organization.
Competitors to KNIME: Alteryx, dbt, RapidMiner
KNIME is easy to use, especially when a company’s data warehouse is relational / SQL based.
Truly a no coding software which empowers more users with access to data.
Generate quick analytics, “on-the-fly.”
Low code, no code systems appear to have a common drawback in that setup takes careful planning and management.
Due to its visual nature, KNIME is resource intensive as a querying tool. Complex queries may be difficult to scale in this tool. Users will consume more RAM in KNIME compared to “standard” techniques to access the data.
Investment in KNIME server is needed to benefit from the full democratization and automation features of the software.
KNIME doesn’t remove the need of a user to understand the data being queried to produce reports that are meaningful.
To better serve an organization, analytics teams should explore software that can expand the efficiency of their teams to produce new insights into operations. This takes the form of data democratization and the expansion of analytics into process mining.
Analytics can be a barrier to getting decision makers the data they need, when they need it. Analytics workflow software, like KNIME and Alteryx, can lower the barrier to data. These tools can give decision makers access to more data and lessen the time to insights.
Analytics can be applied to processes. The ability to identify operational inefficiencies and optimize resource allocation can improve a company's profitability. Both of these types of solutions are worthy to consider for your organization.