How to keep up to date with your company on a daily basis

Keeping track of your business metrics is undoubtedly always beneficial. However, in times when so many changes are constantly occurring, it is necessary to look deeper and much more frequently. In this case, it is likely that you won't be able to do without a data collection system, especially one with the most up-to-date information and in a convenient format.

How does such a data collection system typically look?

Often it involves daily verbal and written reports prepared by employees: emails with attached Excel files or links to Google Sheets, as well as meetings and calls.

This method is unreasonably labor-intensive: you need to access the email, open the message, download the file or follow the link, and find the necessary table. Essentially, this is a pull action rather than a push. At large business scales, it becomes very difficult to see the details in the data reports, as they are usually presented in a non-intuitive manner.

Alternatively, the process may be as follows: the manager independently logs into all the company systems one by one. It's worth noting that there are usually several systems in each company, and the manager looks at numerous indicators from different departments to form a comprehensive overall picture in their mind.

Logging in yourself is also a viable option. In many systems, you can view indicators and even set up automated email distribution. However, this does not provide an understanding of the full range of indicators — only a fragmented representation. Moreover, logging in is also a pull action, and it involves even more clicks than in the first scenario.

***

But today we know that we can and should do better than these two solutions.

***

Business-functional requirements for this new solution:

1. Push Messages in Chat

Business owners are always busy individuals, but the priority of chats remains high for them. Therefore, metrics should be sent to them via Telegram, especially gaining popularity lately, or WhatsApp, which still remains the most popular messenger. There should be push notifications through chat messages: a report is sent as an image to a group or a specific user. This process can be automated or assigned to a responsible person.

Why is a chat needed? It's simple: otherwise, business data will lose the battle for attention against social media and news feeds. If they are only accessible through a PC screen, they have significantly fewer chances of receiving the attention they deserve. On average, each of us spends about 5 hours looking at a mobile phone screen. It makes sense if the information goes where attention is directed.

2. High Visibility

Perhaps, this is the most important condition. What makes up visibility?

• Configured visual hierarchy and the use of pre-attentive attributes (more on this later).
• Reduction of white noise (unnecessary clutter) to reduce cognitive load.

It's essential to understand that our brain reads visual elements in 200 milliseconds, processing them unconsciously. If we read a chart so quickly that we don't even realize it, then looking at a table, we "read" it. Here, a more verbal system is used, which is a conscious and relatively slow process. That's why simple but well-formatted graphics should be used.

3. Plan/Fact KPI

There should be a comparison with planned KPIs on a daily basis. Just looking at revenue is useless— it's better when you can compare day-to-day with the plan or with the previous month and year. Comparing actual data to the plan allows giving an evaluative character to the obtained information, bringing the user one step closer in the decision-making process.

At a minimum, there should be simultaneous comparison of MTD and YTD indicators. And further — more: you can add other three-letter abbreviations like WTD, QTD, YEE, LTM.

Presence of Secondary IndicatorsThey should change from quarter to quarter. According to KPI building methodologies, a company's focus indicators should change and simultaneously be balanced or distributed across levels (each hierarchical position having its own KPI, similar to OKRs).

There are two tools designed for the same purpose. They have different strengths through which an experienced manager can manage, achieving the best results. The Balanced Scorecard system is best used for communication and strategy implementation at the organizational level, while OKRs (Objectives and Key Results) are excellent for motivating people to set and achieve goals at their level.

Accessibility. Timeliness. Detailing.A business owner should always have the ability to enter a specific resource and interactively check and analyze the reasons for deviations from the plan. And there it is! That requirement, the most challenging to implement. If the previous four points can be realized only at the Excel level but with the involvement of a very skilled analyst with high data visualization capabilities, the last requirement will require a different tool.

***

You might ask: 'Perhaps, there is no need for this requirement? Is it really necessary for the owner to have the ability to look at the details of their business indicators daily?'

Here, there's no sense in convincing those who have learned to work relying on intuition, and at the same time, proving the obvious to those who have long understood the power of numbers.

I suggest leaving the question open because everything depends on the management style in each individual company.

***

But, let's say, the management style is data-driven. How then can these requirements be met, especially the last one?

Technically, there is a possibility to gather all the data for building the required and possible KPIs in one storage place — a Data Warehouse.

In simple terms, from the users' perspective, a Data Warehouse is a place where all the necessary data is stored in the form of tables that are interconnected by key columns to preserve analytical value. It is also necessary to ensure that it (the warehouse) is automatically updated on a schedule. Such updates between databases can be configured through scripts in Python or through its more structured reincarnation specifically for data orchestration - Apache Airflow. Often, during the update, it is necessary to transform (cleanse and/or supplement) the data, as they are mostly unsuitable for immediate analysis. If you're interested in reading more, all these processes of moving data from one place to another are called ETL (Extract, Transform, Load — extracting data from sources, transforming/cleansing them, and loading them into a data warehouse).

Having a warehouse, you can start visualizing data with well-known tools such as Power BI, Tableau, or QlikView. Unfortunately, these tools are becoming semi-closed, so more affordable or free and still open tools come to the rescue. These include Apache Superset, Metabase, or Redash — three free data visualization systems. Despite not being as functional as fully commercial products, they can handle the basic functionality. It's also important to mention the ability to work with Yandex Lens, a visualization platform from Yandex, which also, in general, visualizes data quite well.

Now, it remains to understand how to best display KPIs, linking data and configuring reports. Often, it is precisely due to the lack of clarity in reports and the simplicity of their use that users stop accessing them. And for the key user of reports, the business owner, these parameters are extremely important.

To become a professional in data visualization, it is necessary to acquire fundamental knowledge. Understanding why some data visualization methods work better than others is deeply rooted in psychology. That's why every time you deal with data visualization, you must apply the same principles of Gestalt or pre-attentive attributes. There is a lot of literature dedicated to these and other methods. For example, Cole Nussbaumer Knaflic's book 'Storytelling with Data' is a valuable resource. Thus, this is an extremely important aspect.

As for the time frame, development can take about 2-3 months, but this is assuming the presence of a ready team. I want to note that previously, all this was only available to large companies, and building a data warehouse took almost a year and cost around 1 million dollars.

Now that working with data is being democratized, even in spite of external constraints, companies of all sizes are just as likely to move to data-driven management as owners of multiple businesses...