Seven Typical Mistakes in Digital Transformation

For some companies, “big data” and “business intelligence” are still just beautiful words. The reason is a misunderstanding of the essence of these processes and stereotypical thinking.

“I already know everything much better than your reports”

This stereotype is often found among managers of the old school. They are used to making decisions personally, guided by their own experience, intuition and interpretation of information received from their subordinates in the framework of personal communication - face-to-face or conference calls.

On the one hand, if a person is an expert in his business, then he is the best at interpreting the data around this business. He is aware of what is happening, how production is organized, knows the context in which his subordinates make reports and transmit information.

As practice shows, this approach works only up to a certain limit.

The main problem is that instead of objective evidence and factual data (numbers, metrics and parameters), the leader receives subjective opinions and assessments of other people. As a result, the boss first of all hears only what he wants to hear. And the subordinates, for their part, do everything so as not to upset the management: they keenly feel the expectations of the boss and for the most part begin to play along with these expectations, telling not everything, not that, etc.

Digitalization solves this problem, it shows the real picture. Based on objective data, it is much easier to make the right management decisions.

“I will set up analytics, and it will immediately become clear how to manage a business”

The second misconception is in many ways the opposite of the first. In this case, the decision maker relies solely on analytics, data and reporting that are collected from corporate systems. The manager sincerely believes that digitization will help solve absolutely all business problems and put things in order: understand what is happening in each of the departments; build effective sales; plan a production program, etc. But this is an illusion.

In fact, analytics and reporting are just tools for extracting the right information from a large amount of data. And here the requirements for these data, which are formed by the entrepreneur himself, come to the fore.

For example, using analytics, you can get information about how sales are going in the company, what is the conversion rate at each sales stage: how many calls employees made, how many commercial offers were sent, how many meetings were held, how many sales were closed. All of this data is presented in some form, such as a bar chart with the stages of a sales funnel. The graph shows that the decrease in conversion occurs between the stages of sending a commercial proposal and organizing a meeting. This is useful information, but what exactly does it say? How to interpret this data?

Interpretation is the task of the decision maker. It is he who must correlate the facts from the report with what tasks the business faces and what management decision should be made. That is why BI systems are called management decision support systems. Their job is to provide data.

“We will build our own analytical system, we have programmers”

It is not uncommon for a director or IT manager to believe that a company can build its own analytics system on its own, without the involvement of a contractor or consultants. This organization of processes provides for the active use of Excel by employees for their work. The enterprise also has a solid staff of IT specialists who know (learn) Power BI or similar tools, they are ready to do integration on behalf of the manager. Nevertheless, in 80% of cases, such projects end in failure. The remaining 20% falls on companies with a truly unique team of IT specialists (for example, it can be IT companies, including telecom and fintech) and stagnant enterprises that have extremely little need for analytical systems.

In general, BI systems are implemented in companies for development: they allow you to conduct certain experiments on the business and look at the result (worse or better). At the same time, the systems themselves must also develop along with the business: architecturally, technically, in terms of approaches to working with data (reporting, analytics, forecasting).

The analytical system must be designed in the right way. First, it must not be extensible. Developers must be knowledgeable about modeling because any business data contains a specific model. This model technically needs to be able to fit into the BI system pointwise, structure the data warehouse, create the correct connections, organize the collection and updating of data, etc. correctly. As a rule, IT specialists do not have such competencies even in very large IT companies.

“We will merge a lot of different data from different systems into one database, and everything will become clear”

This misconception is directly related to the quality of the data. It is not uncommon for a company to be completely confused: the same entities (for example, commodity items and product names) in different corporate systems can be called differently, separate unrelated product directories are maintained, etc. At the same time, CIOs have an opinion that if all this disparate and “dirty” data is merged into a single database, everything will immediately become clear: reporting will become transparent and of high quality.

In fact, combining data from different systems will not solve the main problem. At its core, the question here lies not in the technical part, but in the methodological one. It is important to understand whether the company maintains a single product directory for sales, marketing, advertising, customer service, up to printing price tags in stores. If not, then even before the stage of implementing analytics, you should create a single directory that will become the source of standardized names for all other corporate systems.

“Analytics is expensive, and you need to invest a lot of money right away”

Quite often, companies that are thinking about digitalization believe that the implementation of analytical systems is insanely expensive. And partly it is. At least the integrators themselves try by hook or by crook to maintain the high price of projects, explaining this by the complexity of immersion in the internal business processes. They explore data sources, storage and upload rules, compare names, calculation formulas, etc. Entrepreneurs, on the other hand, have a fear of investing in this direction, because so far they have little idea of what exactly they need.

Basic business intelligence without creating a separate IT infrastructure can be built on your own, uploading data from corporate systems yourself. In this case, the costs will be much less. At the same time, business analysts or department heads can do the work. If necessary, they can ask IT specialists to develop automated data upload tools and set up analytics on top of them: platforms, visualization, dashboards, etc. For companies that are taking their first steps in BI, this is a great experience and a very cheap option. In addition, specialists can touch the data with their own hands, understand what requirements they have for an analytical system, and only then think about a full-fledged infrastructure.

“The main benefit of analytics for business is to save labor costs for employees”

As a rule, many digitalization projects are associated with the automation of existing reporting and getting rid of manual labor. At the same time, there is a need for data analysis. And quite often, CIOs insist on evaluating the cost-benefit of implementing a new system based on labor savings. But in fact, the impact of the BI system is not limited to payroll savings. This is just the visible part of the iceberg.

By accessing data, discovering patterns, and finding unknown insights, an analytics system brings new business opportunities. It allows the manager to make fundamentally different quality management decisions. For example, identify cities for opening branches or replace real employee meetings with online video conferencing. And it is quite possible that in these improvements there will be a much greater economic effect for the company than simply saving on the labor of one person who was involved in the preparation of reports. As a result, the company can start earning much more, get a new impetus in development.

“I will hire an integrator and he will bring me the best data management practices”

A very common misconception of customers is the complete distancing from the process of developing and implementing an analytical system. In this case, the company fully and completely hopes for the expertise of the system integrator. As a result, she issues a general technical task and does not participate in further stages of work.

In fact, it often turns out that the performer does not have the slightest idea about the essence and principles of solving customer problems, he sees only an opportunity to make good money. As a result, the contractor will experiment at the expense of the customer, and this will negatively affect the quality and success of projects. An equally important aspect is the experience of the team in terms of the subject area. It is not enough to be able to build BI systems according to a template, it is important to have industry competencies. Each industry has its own best practices, its own set of tools, its related systems (for example, SAP is used in the form of ERP), it is important that the performer knows it very well.

For this reason, the customer company needs to clearly monitor the work of the contractor, and if the parties have discrepancies, and the dialogue goes into a formal plane, there will be no success.