Dimitar Dekov, founder and CEO of B EYE, shared with us the good practices in launching, managing and developing projects related to data analytics.
Dimitar Dekov, founder and CEO of B EYE
What are the most common factors that determine whether a Data / Business Intelligence (BI) project is successful?
In many cases, the success rate depends entirely on the users, as well as how they use their data and BI. It is difficult to say whether more low-end users are better than fewer, but higher-ranked ones. This is also determined by the structure of the organization. The main criteria for measuring a positive result can be presented as two variables: 1. whether processes are optimized and 2. whether there are qualitatively new results, such as new processes and business niches.
What conditions should be in place to start a Business Intelligence project?
I would identify the processing and management culture in the organization as the most crucial factors. If the people in the company, at different levels, have the necessary management culture, then this is a good prerequisite for starting the project. As one of our customers said: “Any company can produce and sell electricity, shoes or any other product. The difference between successful and unsuccessful companies comes from how they use an asset called “Data”.” There must be a critical mass of people involved in this culture. It is also essential to consider the organizational structure since dashboards and KPIs are created in accordance with it. For example, it doesn’t make sense to have a dashboard measuring a company’s digital presence, if you don’t have a digital marketing specialist.
What kind of people should be involved in the project? What qualifications are needed for the staff?
It is a good practice to involve both external and internal resources. Rarely can a company achieve the necessary innovation with only the internally available resources. The following members should be participating from the client-side:
- Business Champion/Business Sponsor – the one under whose auspices and sponsorship the project is created;
- Process/domain owners – the experts with knowledge of the relevant processes in the company, such as sales, deliveries, production, etc.;
- People who know the data and know-how and where the data is stored in the various systems. Unfortunately, in many cases, this is not recorded anywhere in the organization and this kind of knowledge is only spread by word of mouth. That is why it is known as “tribal knowledge”. This could be the case even in developed organizations, but it is essential to combat such bad practices by creating appropriate documentation;
- Business analysts to collect the requirements;
- Key users who will actively use the created dashboards and data.
Of course, some of these roles may also be filled in by an external provider, but not the Business Sponsor and the Process / Domain Owners.
From the vendors’ side, the following roles may be provided: a Project manager, a Visual analytics designer, a Front-end/back-end developer, a Technical documentation Writer, a QA, a Coach. It is good if the IT department is also involved in the project, yet not vital for its success. They play more of a partner role, but they should not make key decisions since the owner here is the business.
“We often see companies choosing a cheap tool or one that comes as a bonus when they buy another software system. These types of “bonus” tools often neither meet the needs of the company nor offer the necessary functionalities. Thus, when you draw the line, they turn out to be very expensive.”
To what extent does the management need to be involved?
Both the idea and sponsorship of such a project come from high or mid-level management. These types of data projects bring managerial changes. Thus, management plays a crucial role in the data project. Very often, the main business sponsor does not deal with operational issues, of course, it also depends on their management culture. However, if in my company I do not support and promote the changes, they will not happen the way I would like them to happen.
How can management promote the project within the organization?
Usually, the people involved in the project are the main promoters of the idea. This includes us, as an external company, together with the project stakeholders. A successful story is created, on which management can then step and build upon. Often, initially, there is resistance to change. This has happened even in our most successful projects. Hence, we need a Business champion – someone who encourages others to “bear the pain” and actively promote the benefits and results of the project.
And what is the most appropriate methodology for implementation?
I would say the Agile methodology – an implementation in small steps. This way, the risk is controlled, and the best results are achieved for the company. The Waterfall methodology has not been used in these types of projects for a long time.
How important is the choice of a software tool?
In the long run, the choice is crucial. We often see companies choosing a cheap tool or one that comes as a bonus when they buy another software system. These types of “bonus” tools often neither meet the needs of the company nor offer the necessary functionalities. Thus, when you draw the line, they turn out to be very expensive. Equally important, however, is the vendor. Even the best tool, combined with a bad implementer, will bring poor results.
If you want to choose a suitable software, check it out in action. Not based on tables or other ways software was evaluated 15 years ago but based on a Proof of concept (POC) project. Keep in mind that these POC projects are not for free. In the West, quality POC projects are paid for. In my opinion, expecting them for free is the reason why many IT projects have failed. The practice of requesting free POC projects is vicious and repels good integration companies. Good quality does not come for free.
What happens after the Business Intelligence project is completed? Another BI or AI project, or something else?
In most cases another Business Intelligence project or an extension of the current one. You should keep in mind that Business Intelligence is a long-term task. The project is being developed together with the company. AI (Artificial Intelligence) is something different. At some point, mature companies that have a good data culture may start thinking about advanced analytics and AI.