The Importance Of Business Intelligence
There was once a time where people in the workforce were treated as human processors who execute a predefined process that took data inputs and transformed them to data outputs. This was time when metal computers were too expensive, and paper instructions were cheaper to execute than machine code. Humans who crunched numbers were literally called "computers".
There are still companies that treat humans as data processors or computers, even though the cost of automation has dropped precipitously. In these companies, some managers may tell themselves: "It's easier to pay the analyst salary than to invest in automation to free up people's time. Free time is dangerous. Free time means unproductivity, unproductivity means layoffs. A layoff means that my budget shrinks, and I could be next on the chopping block. No, my ability to predict my costs and define precisely what I will and won’t do is key to my job security. What would people do with their free time anyways? It's not like the problems to solve at the company are infinite."
This mindset is so incredibly dangerous to the company. An influential manager may be able to keep themselves and their team employed for decades, but they will not contribute to the growth of the company. In fact, their beliefs will cause them to take action, conscious or not, to subvert and sabotage the company's growth, because change represents a danger to their budget. And when budgets get shaky, the team will be sacrificed, causing brain drain and further cynicism and detachment.
Hopefully you are not at one of these companies. But you might work at a company where the price of automation has been objectively, historically, and irrefutably expensive. You might benefit from reading the Power Plant story to understand how a company ends up in a place where cheap technology actually can result in very expensive projects.
Secretaries versus Executive Assistants. Translators versus Collaborators
The most common paradigm for how companies treat their business intelligence and reporting is what I call the "data secretary" model. In these companies, the intelligence of the analyst is respected. The methods of the analyst seem so foreign that they get stuck doing the work of stakeholders who can't do it themselves. And should they do a good job, stakeholder needs continuously become more complicated.
Now, I understand that people have secretaries to communicate ideas and manage information. Back in the day, many computers were literally too difficult, too expensive, or too heavy to practically fit into the lives of executives. Zoom and unlimited phone calls were not a thing, so executives literally had to move themselves through space to get work done. Secretaries played an important role in bringing information out of inaccessible formats and integrating it into the work lives of their bosses. But the secretary is usually not responsible for creating any key communications themselves.
As personal computers became cheaper, lighter, and easier to use, and as phones began to resemble computers, even those in the highest echelons of a company started to type their own letters. Some did it willingly. For them, investing to learn new technology was a blessing and allowed them to turn their ideas into action faster. Some did it less willingly, but capitulated in order to keep up. Motivations exist on a spectrum.
If you met an executive today who absolutely refused to read or write digital communications and demanded a secretary to write their texts, you would be stunned. You wouldn't hire them. Now if an executive asks for an assistant to help with planning and facilitation, there is respect and autonomy in the design of both roles. All parties can leverage technology and their skills to create better communications required at their level. They get aligned first, and then they fan out to do their part of the job. The executive and the assistant can both benefit from the wave of technology that is available to them to communicate shared ideas.
What is the level at which an executive should interact with data? We are currently in a wave of technology that allows ordinary, non-technical people to extract information from data and form their own insights. These technologies are nothing new. Business intelligence (BI) tools have been around for decades. They are portable, quick to learn, and becoming easier to use every day. But just like the adoption of personal computers took time, adoption of business intelligence exists on a spectrum.
Companies That Embrace Enterprise Wide Business Intelligence Achieve New Frontiers
There are companies out there where every single employee is data literate and has no problem using data to answer questions and make better decisions. In companies like this, operators can take some degree of process design into their own hands by streamlining their own processes by configuring dashboards and setting up simple automations that are available in all modern BI tools. Middle managers spend very little time tracking down information or statuses. Instead, they create systems and processes that regularly and predictably democratize information. They make it available on demand and work to improve the structure of that information as the role of their team evolves.
Because of the work of the entire organization to make information predictably available, leaders at these companies spend very little time hunting down status updates or reviewing vanity metrics like percentage growth. Instead, they look for what drove that growth. They can review organized information to see patterns and form new insights that help them discover what works. Armed with knowledge, these executives can take bigger, more informed risks and redirect their energies towards expanding their frontiers.
The Perils Of Partial Adoption = The Disintegration Of Questions And Answers
The adoption of company-wide data literacy is not nearly as universal as company-wide email or messaging. There are still many executives who believe in delegating data-related tasks far from their team. In these companies, there is usually a reporting service desk that centrally produces the business intelligence that all teams need. This sounds good, people can specialize. Specialization creates speed. But the issue with analytics is that it's not a technical task -- it requires context and customization. It's not changing the oil in your car, it's creating a custom suit from custom cloth. Fast analytics is mass produced analytics, completely devoid of context, meaning, or insight.
There is tremendous communication overhead delegating the search for insight. The problem with getting data or metrics as a receiver (or stakeholder) is that once you see item 1, you want item 1-a, then 1-b. Receivers of metrics see their job as asking questions but lack the skill or will to actively participate in generating the answer. Data secretaries have to work with data that comes from across the company at various quality levels and have to piece together the business context that created the data. Receivers are usually out of their depth at describing what they need because they lack experience to actively explore and touch data past the surface level. So they just describe more and more, and wonder why it takes so long for the analysts to "get it right". The fury and frustration that exists in service desk reporting teams and in the stakeholder is palpable.
Thriving Business Intelligence Is A Group Effort
Thriving business intelligence requires the active participation of all to create context that allows data teams to structure for the data that is important to the enterprise. The structure of this data, also called the enterprise data model, mirrors everyone's understanding of how the company works. In fact, this activity often generates its own form of insight worth retaining.
Imagine if you discovered that a manager only asked questions but refused to Google or check the company dashboards to see if an answer already exists. Instead, they delegated finding an answer and refused to contribute to the company's wiki once they found a solution. There are some positions that can sustain this level of delegation, but this is not a beneficial pattern. Google is actually REALLY good. Wikis are actually REALLY searchable. BI tools are actually REALLY accessible. How quickly do you think this executive will fall behind? And what example do you think they set? What will the impact be on the people who work for them and with them?
Shared Data Literacy Trumps Data Technology
The secret is that business intelligence is not really about technology, but about an attitude towards what it means for a company to be intelligent. The mark of intelligence is being able to retain information, compress information into insights that can be applied in new situations where you cannot analyze all the data, and update your insights based on new information.
Company intelligence is different from stakeholder intelligence. It's not the rate at which you can ask questions, but the rate at which the whole organization can structure data, share information, generate insights, and update their understanding. Business intelligence is about the quality and speed of sharing information. A company can have excellent business intelligence with very little technology if it is understood that every week, all parties updated their communication log on a structured document so that information can be accessed on demand and retained for all.
Make no mistake: the only difference that technology makes is that it can automate certain predictable information sharing activities and allow the company to express more complex structures. The contribution of a BI tool is a drill down to inspect the underlying data that created an insight. The contribution of a database allows you to link data from different parts of the organization together in a structured way. But these benefits from technology can only be unlocked if all members of the organization can differentiate between predictable and custom data activities. The former can be mass produced and automated. The latter takes collaboration. But this collaboration is never wasted: it is the activity that enables the company to express more complex information structures and deeper insights. And over time, even these become predictable for the company as they move onto new analytical questions.
In this age, ignorance is expensive. Human intelligence should never be fractured in which the question asker is not the answer seeker. There is a reason that the word "slow" is synonymous with stupid. Every human being who can use a browser already has the ability to explore data, process information, and update their beliefs. In a business intelligence company, everyone then has the responsibility to share what they know to contribute to the knowledge structure of their community. That is all business intelligence is. The associated technologies (BI tools, data marts, data warehouses, ETL pipelines) are simply tools to be able to handle larger and more diverse sets of data and model their relationships to match human understanding. Business intelligence is a mindset. As analytical leaders, we must see part of our job here to cultivate this mindset. We have to give everyone who wants to solve problems the skills to solve more diverse problems, larger problems, and critically interrelated problems.
If not, the data secretary desk will always have a seat for you. But we don't guarantee that the company that designs that desk into their business plan will be around for much longer.
