6 min.
Why the usefulness of data starts with people
1L’art de la gestion de projet2Un projet à succès commence par une bonne gouvernance3Cascade, agilité, demandes de changement?

Why the usefulness of data starts with people

Digital transformation

This article is the second in a two-part series on inspiring discipline across your organization. You can find the first part here.

Remember our simple formula: Strategy + Discipline = Results. In a continually changing, competitive environment, leaders are constantly facing new challenges and must adapt both the organization and their leadership style to develop and foster new mindsets within their teams.

Here, we explore in more depth the discipline and focus of a data-driven mindset that centers its decision making on data, one of the digital transformation concepts addressed in our first piece.

How can leaders effectively guide their organizations into the future of the digital age?

Exploding quantities of data have the potential to fuel a new era of fact-based innovation in businesses, backing up new ideas and new business models with solid evidence. Stimulated by the hopes of improving customer satisfaction, streamlining operations, and clarifying strategy, companies are accumulating data, investing in technologies, and paying generously for analytical talent. Yet for many companies, a strong data-driven culture remains evasive. Data is rarely the coherent basis for efficient decision making.

Why is it so complicated?

Our experience working with clients in a range of industries demonstrates that the biggest obstacles to creating data-based decision-making in businesses aren’t technical; they’re cultural. It’s simple enough to describe how to introduce data into a decision-making process. It’s far more difficult to make this automatic for employees. The shift in mindset, the gap in skills and knowledge of legacy employees and the creation of new habits present a daunting challenge.

So, we’ve laid out some pointers to help sustain a culture that owns data at its core.

1. A data-driven culture starts at the very top.

Companies with strong data-driven cultures tend to have top managers who set an expectation that decisions must be anchored in data. They lead through example— this should be normal, not novel or exceptional. At leading tech firms, senior executives spend the start of every day reading detailed summaries of proposals and their supporting facts, so that they can take evidence-based actions. These practices propagate downwards, as employees who want to be taken seriously have to communicate with senior leaders on their terms and in their language. The example set by a few at the top can catalyze substantial shifts in company-wide norms.

2. The main goal of data in an organization is to support a business process.

A business process in simple terms is just a set of activities that allows an organization to create a service or a product for a customer. The data enables the organization to understand each activity of this business process to make it efficient and effective.

Generally speaking business processes support business goals. An example of a simple business process would be Acquiring a customer. Two important keywords to remember in this paragraph are efficiency and effectiveness. The data should enable an organization to generate actionable insights to optimize its core processes in order to reach its business goals. To sum up the data should help an organization make better decisions and improve its operations.

3. Start your value-driven analytics initiatives with a business problem.

It always starts with a business problem. In her book “Modern Analytics Methodologies: Driving Business Value With Analytics”, Michele Chambers explains that to drive business value with analytics you need to:

  • focus on a business objective (ex: minimize waste, improve efficiencies),
  • identify a business process from your value chain (Porter’s value chain) linked to this objective,
  • identify within this business process some core activities to improve.

Let’s take a concrete example:

  • Business objective: Decrease the cost per acquisition (CPA), the amount you spend to acquire a customer
  • Business process: Acquiring a customer (Marketing & Sales activities)
  • Core activities: Manage acquisition campaigns

Once these activities are identified, the goal is to analyze them and find areas of improvement using data and insights. Launching these kinds of initiatives often need a proof of concept, this proof of concept might need a business case with some realistic hypotheses. It is possible that you might not need to build a business case if your upper management is ready to sponsor your proof of concept.

4. Choose metrics with care — by design.

Leaders can exert a powerful effect on behavior by artfully choosing what to measure and what metrics they expect employees to use. Suppose a company can improve their reach and brand positioning by anticipating users’ engagement with the brand and its products. Well, there’s a metric for that: customer engagement score. Most companies measure interest in a product by viewing click performance online. While in fact, viewing engagement goes beyond just the click. It emcopasses all the steps within the customer journey that gets you to the product. So a team should continuously make explicit predictions about the magnitude and direction a user moves within his journey with the brand. It should also track the quality of those predictions – they will want to steadily improve.

5. Build a proof of concept (POC).

Building a proof of concept might be sometimes the only way to show how to bring value to your business using data & insights. Like running an experiment, you will need to formulate an hypothesis based on this hypothesis you need to infer some potential business benefits. Be realistic and don’t inflate the potential benefits linked to the results of your experiment.
This process will take the form of a formal business case.

An important step, before building your business case, is to assess your analytics capabilities like recommended by Thomas H Davenport in his book “Competing on analytics: The new science of winning”. Failing to complete this step might cause you trouble in the long run. Assessing your analytics capabilities means understanding if you have the right people, processes, platforms and understand well the culture of your organisation to run your experiment. Like any project you need to find a sponsor for your proof of concept. Once all these pieces are put together and the business case accepted you can launch your initiative. The result of the experiment (POC) might not yield the expected results, it is important to understand why and look for future opportunities.

6. Break silos between business leaders and data geeks.

Data scientists are often isolated within a company, resulting in business and data leaders knowing too little about each other. Analytics can’t survive or provide value if it operates separately from the rest of a business. Companies at the leading edge make sure data science is closer to the business. They pull the business toward data science by insisting that employees are conceptually fluent in data analytics topics. Senior leaders don’t need to be reborn as machine-learning engineers. But leaders of data-centric organizations cannot remain ignorant of the language of data.


7. Make the uncertain less uncertain.

Everyone accepts that absolute certainty is impossible. Yet managers continue to ask their teams for answers without a corresponding measure of confidence. Requiring teams to be explicit and quantify their levels of certainty or uncertainty has powerful effects. It forces decision makers to directly deal with potential sources of uncertainty. Analysts gain a deeper understanding of their models when they have to rigorously evaluate uncertainty. Is the data reliable? Are there too few examples for a reliable model? How can factors be incorporated when there are no data for them, such as emerging business model dynamics?

An emphasis on understanding uncertainty pushes organizations to run experiments. Conduct statistically rigorous, controlled trials of ideas before making widespread changes, such as prototyping of or proof of concepts (POC).

8. Analytics makes employees happier.

Empowering employees to argument data themselves enables self-accomplishment. Technology enables automating the boring stuff. If the immediate goals directly benefit them — by saving time, helping avoid rework, or fetching frequently-needed information — then a chore becomes a choice.

Some closing thoughts

It’s a cliché, but change really does start from within. Change should never start with technology, as businesses look to change it should all be focused on one thing, the “hearts and minds” of the organisation. Companies — the business units and the individuals that comprise them — often fall back on habit, because alternatives seem too risky and the shortfall to meet ambitions appear too great to overcome. Perhaps we need to add a new formula: Courage + Fear = Innovation.

While data can provide a form of evidence to back up hypotheses, and provide the confidence to jump into new areas without taking a leap in the dark; it has to come from the business who believes in the vision, strategy and impact onto them as individuals.

Simply aspiring to be data-driven is not enough. To be driven by data, companies need to develop cultures in which this mindset can flourish. Leaders can promote this shift through example, by practicing new habits, by including business units and individuals in the process and creating expectations for what it really means to root decisions in data — always putting people first.