Data visualization problems I face in almost every project

My name is Georgy Tsygankov, I am engaged in / specialize in building BI solutions in the IT company Tieto. We are engaged in business analytics and data visualization. For about 7 years I have been working with various BI platforms and have a fairly diverse experience of projects in several industries: logistics, manufacturing, retail and the field of providing services to the population. I am interested in comparing platforms from the point of view of developers, analysts, business users, as well as analyzing business pains and complexities on projects.

In this article I will tell you how visualization affects the perception of information, what is important when choosing BI systems, and why almost no implementation project is complete without external consultants.

Beauty vs consistency

Visualization is not just beautiful reports, but clear business logic. And those who have just embarked on the path of introducing and regularly using BI tools may confuse these concepts.

BI analytics is often confused with the concept of “infographics”, or drawing a beautiful report, which reflects the dependencies of different indicators. The main task of such a report is to attract attention, including through the interactivity of elements, where one flows into another, moves and turns. However, all of these animated features and drawings serve more of a marketing function.

Real BI-analytics are solutions of the DSS class, or decision-making assistance systems. Its main task is to provide the user with information in such a way as to reduce his time spent on analysis and making management decisions. Therefore, the foreground is not visual perfection, but the transparency of the reflection of information.

In practice, the difference between these concepts is great. For example, with an incorrect selection of the visualization option (3d Piechart), the perception of the ratio of sectors is disturbed, and the user perceives the values ​​as smaller, but in fact they are larger in absolute terms. By using the color and arrangement of elements in infographics, you can place accents so that the data is presented in the right way, for example, without emphasizing on insufficiently good indicators. Therefore, in BI analytics, visualization is standardized and has the usual designation: if we see a tall column in a chart, then without looking at the numbers we understand that this is the largest indicator.

Now there are practically no companies that do not have their own standards for the visualization of all materials: documentation, presentations, analytical and financial reporting. As a rule, companies with a developed corporate culture have a brand book in which all these points are described and regulated. At the same time, it is allowed to create dashboards that differ from generally accepted standards for internal use among divisions. For example, an analyst or a head of one of the business areas can prescribe a color scheme in the technical assignment that is convenient for him specifically. The main task is to give the user the ability to quickly navigate the analytics. For companies that do not have a corporate identity, or want to move away from it, consultants and contractors usually offer the best solutions.

But this disregard for standards ends at the moment when the departments are tasked with preparing consolidated statements for the top management of the company. In this case, one cannot do without standardizing the approach to visualizing analytics.

The quintessence of BI analytics is to reduce decision-making time and improve their quality. And the goal of standardization is to make it easy for management to understand the report of each department. When there are requirements for a general color scheme, report structure and categories, even a quick glance at graphs and charts is enough to immediately understand the essence of the report.

Indicators of indicators in any analytical dashboard are always the same, the standard option is “traffic light”, this is a traditional color palette. However, the indication changes when BI becomes a vehicle for making operational and strategic decisions. This is more common for digital transformation companies whose employees are involved in the digitalization process. Imagine a business with over 100,000 people. Each has its own KPI, which is carefully monitored. The indicators of each employee can be viewed on a separate dashboard. That is why companies do not limit themselves to indicating “traffic lights” and use other colors close to the natural palette, which still do not interfere with immediate perception of information and even evoke positive emotions.

In most cases, BI platforms differ in the depth and level of settings for the appearance of visualizations. The most advanced solutions allow you to implement even the most specific tasks that are not covered by the standard set of capabilities. At the same time, you do not need to go to marketplaces or create a new type of visualization in supported programming languages.

But the parameters for comparing BI platforms are not limited to these. For example, employee involvement is important for the management of companies. As part of training the system, each business user is explained how to customize and benefit from the generated reports in their daily work. It is also important for managers that these reports are built not once a month on demand, but are a full-fledged tool for making decisions. As a rule, companies build a KPI tree, and employees at each level track the performance of their indicators, which “add up” to the generalized performance indicators of the entire organization.

The ideal picture for a business – employees themselves understand how to apply and customize visualization, they see what indicators need to be improved, and there is no need to hire external consultants, because users themselves cope with everything. However, this is extremely rarely achieved – in reality, a business cannot cope without the help of outside experts.

Why business can’t do it on its own

Despite the fact that the number of companies implementing BI is growing, and some of these players are implementing even more than one such system, internal expertise is not enough. Engaging external Business Intelligence experts early in the deployment of IT solutions is a must, and for good reason.

· An abundance of variegated tasks

There is a popular misconception that it is enough to just buy and run a BI system, and then everything will be fine: users themselves will create beautiful and convenient dashboards. In practice, however, this will not work.

It is not enough just to implement BI, even if it is a self-service analytical system. At the same time, business needs to solve many related issues, including organizational ones. Among them are a design system, data certification and access to them, dashboard templates, training for new users, and much more. Unfortunately, not all internal BI teams are ready for such a huge number of diverse tasks, some of which are almost impossible to understand without the involvement of external experts.

· Staff shortage

Experts of the domestic IT market have noted the increased demand for IT specialists in the past couple of years, which, of course, determined the growth in the cost of projects for the implementation, development and support of BI systems. The shortage of personnel is increasing, which sometimes can even lead to the freezing of IT projects.

And here the issue of qualifications of specialists in the field of Business Intelligence is just as important. Even with their own IT team, few people do without external experts. Firstly, often internal specialists do not close the tasks of implementing and configuring analytics – they simply do not have enough competencies and experience on different projects. And secondly, many people need help choosing the most suitable system from those who know the market, are independent of vendors and are based on the real needs of the business, and not on personal interest. This choice bears fruit – as a business significantly reduces the risk of errors and reduces the time for project implementation.

Another popular scenario: the company already has a BI solution, and you need to develop it, increase functionality. In this case, the in-house IT team copes with only minor tasks of finalizing or configuring the system. And in the case of self-service solutions, business users themselves can take the necessary actions: IT specialists prepare a data mart, and department employees choose how this data will be displayed. However, most often the task is more difficult, you need to know the intricacies of the platform, and here they are already attracting external consultants.

· High cost of mistakes

As practice shows, saving the budget for outsourcing experts translates into costs for the company for additional training of internal specialists, correcting errors and compensating for the time spent.

Moreover, the involvement of external IT partners is safe – any modification of the system can be carried out without access to data. Consultants propose a data structure and, as a result, provide users with the requirements for that mart or information source structure.

· Labor intensity

The render build is the top of the pyramid. Its most resource-intensive stage is working with data: collecting data from sources, “cleaning” it, the need to “fold” it correctly for storage, and checking versioning. These tasks do not depend on the platform, however, they make it possible to understand how difficult it will be with the construction of reports and visualization. For this, an external consultant is involved in order to determine whether there will be enough downloads from source systems and whether it is possible not to ask storage questions.

Based on my project experience, I can say that these problems were encountered most often in practice. But I am sure there are much more such “pains”. If you have encountered similar problems – both in-house and as an external expert – I would be grateful if you would share your experience. I would be glad to discuss such cases in the comments!

Similar Posts

Leave a Reply Cancel reply