Data-Driven Transformation

As the data-driven transformation should serve the needs business rather than being a goal in itself, careful reflection and alignment with the essence of the business are essential. When starting, the leadership is advised to gain absolute clarity as to how exactly this move will generate benefits and add value.

• What are our current and future necessities and how can data-driven measures support these?
• What specific advantages can this cultural change bring?

Failing to link the desired data-driven transformation to the overall strategy can easily result in wasted resources as the implemented measures will offer little value to actual business.

In our experience, adopting a data-driven culture in your organization is generally pursued for one of two rather distinct goals – the first targeting internal efficiency gains and the other seeking to turbocharge external growth.

Inwards-facing investments tend to serve as an “entry point” for many companies setting out on their data-driven journey. This approach brings advancements in processes with a focus on sales performance, planning, costs, time, movements, service level, availability, efficiency or even risk planning. dashboards. These improvements ultimately facilitate quicker and more effective strategic decision-makin. As the data required for these improvements is often already available, the investment demand is relatively low. The downside, however, is that inward investments tend to only offer specific control measures and they aren’t exactly game-changing for the overall business.

On the other hand, investments can be pursued with an outward focus: to grow, capture market share, gain a better understanding of clients or more quickly capture opportunities. Opting for an outward focus means collecting and managing data to guide both strategic and tactical decisions – based on information about demand, prices, channels, points of sale, traffic, competitors, trends, customer behavior, social media etc. This sort of information will be very relevant for managing business areas such as marketing, sales, trade marketing and innovation. At the same time, it can also offer strategic insights for “fulfiller” areas such as supply chain and industry.

The challenge in this outward focus relates to gathering the required information while ensuring sufficiently high quality and frequency. In our experience, the available sources of this data are often dispersed, poorly structured and lacking a tofrom with the internal areas.

Compared to the inward focus, the potential benefits of an outward focus are greater and can offer unquestionable value for the business down the road. This means that investments into an outwardsfacing data-driven journey will demand a longer-term view with more technically advanced professionals supporting this transformation - such as data scientists and engineers as well as data providers.

The second trade-off invites the leadership to consider whether a disruptive approach targeting innovation or an incremental approach targeting sustained improvements will be a better fit the current state of the business. These are distinct pathways that will demand very different approaches to the data-driven transformation.

Through the disruptive approach, leaders grant engineers and data scientists freedom to generate insights and identify opportunities for the business and alter the status quo in the most objective, bottom-up way possible. The guiding question here is: “What does the data say?” – irrespective of the structures that already exist internally. By implementing creative new teams and formats with access to data, value is added through innovation, data products, discoveries and new indicators that deliver efficiency and new business paths. We should point out that this more technical approach requires a higher degree of maturity, and value will only be effectively generated once there is already a sufficient amount of data available for technical experts to develop advanced solutions.

The incremental approach is more suited to businesses with clearly defined core projects already in place that would benefit from more refined project management through data-driven indicators, dashboards and rituals. The guiding question here is: “What short-to-mid-term improvements do we seek?” – with more focus on driving value generation based on the current business reality and objectives. Also, rather than relying on professionals with a technical profile, the incremental approach calls on the involvement of internal users such as business analysts and managers. In general, the incremental approach is a more concrete call to action that can be put directly into practice.

In either case, the “right” trade-off to choose will become clear once you’ve defined the specific role that the data-driven transformation will play in furthering your business strategy.

Initial proposals to transform a company by leveraging data for promoting the business and decision-making may be inspired by technical considerations from IT. In other cases, they can also be a response to specific business needs brought up by the leadership. Either way, deciding which area should lead this transformation is not necessarily obvious.

While some organizations may believe that professionals already specialized in IT should guide the rest of the company along this journey, others will want to select leaders from one of the business areas, with their more developed abilities to manage and execute projects. Both approaches offer distinct advantages, but each also comes with its own tradeoffs.

Business leaders

+ greater speed, autonomy and solutions better tailored to the needs of the business
- less ability to replicate and create scalable solutions

IT experts

+ more technical refinement in solutions and ability to replicate
- risk of reduced applicability across business areas and “over-technicalization”

The clear benefit of selecting leaders with experience in the business itself is that they tend to be more attuned with the overall strategy and the needs at hand. With a finger on the pulse of the business, they are aware of the priorities and importance of the indicators. The areas know how to apply indicators and, as these are made for supporting the business in the best way possible, they can also better define how indicators should be read. Having leaders from the business area in the driver’s seat will facilitate the process of setting critical priorities for the first phase of this transformation as well as optimally customizing the solutions.

On the other hand, placing IT leaders in charge is the best way to ensure excellence in executing the new solutions: from data governance and curation to ease of future replication. IT professionals are not, however, as familiar with the intricacies and demands of the business itself. While they may not be as apt in translating these needs with the same degree of fine-tuning, they will ensure that the data-driven transformation itself progresses based on technically robust solutions – an expertise that business leaders are unable to match.

Whichever path is chosen, some prerequisites must be in place first. Business areas will only achieve success if IT has provided them with a conducive tech environment along with training to ensure business analysts know how to use the tools provided. Setting the IT area up for success means providing sufficient resources for gaining a deeper understanding of the necessities of the business, for executing the data-driven model and for maintaining it.

As companies often don’t have sufficient resources to ensure IT can optimally provide high service levels across the board, business areas typically end up assuming the role of key tech sponsors in this transformation. One effective remedy here is to set up data lakes. These will only require IT teams to provide raw (structured) data on a consolidated platform while business areas are trained to “consume” this information to generate the needed KPIs, data visualizations and analyses.

Whenever considering changes that involve data or digital approaches to doing business, the matter of human resources can be a delicate topic. In virtually any place in the world, the qualified digital professionals needed to facilitate this are extremely expensive on account of their scarcity. For the solutions envisioned in your transformation to come to life, a slew of technical capabilities – from programming, data engineering and data governance to UI, UX, BI and cloud computing – must be joined together in a dedicated data-driven cell at the company.

As such, the formidable challenge faced by most companies starting this journey is: How do we attempt to bring together these extremely costly resources in a way that will allow for initial gains to be achieved without breaking the bank?

The trade-off that your data-driven model must address is how much to keep internal and how much to hand over:

• Outsource to third parties to gain speed and cut down on staff costs versus
• Invest in internal capabilities to gain precision and retain key expertise

A response to this challenge we have found to work well with clients is to only contract out necessities that are are highly specific and more operational.

At the same time, it is highly advisable to guarantee that any solutions demanding data intelligence remain in-house – from defining the logic behind calculations and understanding data sources to specifying how transformations are done. Losing out on these key sorts of expertise will certainly be felt in the future, e.g. when it’s time to do maintenance, evaluate the model in place, add additional indicators, address questions regarding the calculated indicators etc.

By outsourcing core knowledge, besides losing essential expertise, your business could also run the risk of the highly anticipated data model either failing altogether or falling into disuse, thereby losing its value and purpose.

Common sense dictates that effectively monitoring performance and outcomes should be done using “few but good” indicators. However, the complexity inherent in some business operations does not allow reality to be clearly defined a simple one-pager.

Take, for example, businesses in the retail apparel industry that must plan future collections, design new models, produce, sell, distribute and monitor the performance of current collections – all in parallel. Each of these complex processes is vital for the business and there is also an entire machinery of internal governance that needs to maintain visibility over each one.

Knowing how to prioritize among a long list of options is therefore a consistent challenge. Even when choosing to focus on a single area, ritual or process, there will still be an abundance of possibilities. Added to this are technical considerations that bring in another layer of complexity to supporting indicators. These can range from accessing primary information, automating, ensuring the correct rules of business, knowing how to do comparisons or presenting indicators an effective manner.

At the early stages of the data-driven transformation, leaders need to define which indicators to prioritize and, once again, decide whether speed or depth will be more suitable for achieving the objectives of the business and data strategy as well as the reality of the organization. This trade-off considers the following points:

• Prioritizing the elements that generate the most value for the business and its processes for the business area(s) versus
• Focusing on elements that are easier to implement for the technical support area.

In an ideal world, we would be able to identify the point at which both of these overlap most.

In many cases, however, we see the leadership at companies spend too much time and effort discussing and striving for an “ideal” set of indicators. A best practice we recommend to avoid this is taking indicators that are already available (even if not the most ideal) based on the data at hand and getting a head start – implementing them through the governance and using them to make progress in the company’s new data-driven approach.

Through the day-to-day practice of testing out this initial set of indicators, the organization will evolve on the transformation by learning how to extract insights and use these as a basis for decision making – while also evaluating effectiveness. In this context, the C-level plays a critical for setting meta indicators, defining priorities and objectives and, at the same time, providing the organization with the freedom to design a dashboard of indicators that are best suited to achieving them.

Some common errors can also arise when pursuing this trade-off.

After developing and implementing your transformation strategy, the next step tackles the matter of adoption: handing over the data-driven solutions that have been created to the end-users – the internal professionals who will “consume” the data being offered. Having a dashboard ready to use is one thing but actually getting your users accustomed to effectively using it in their day-to-day routines is a different challenge altogether: making sure they are looking at the indicators, making decisions based off of these and applying insights to guide the overall business.

The general difficulty that businesses encounter here is that this transition is not necessarily 100% natural for users to adopt. Many may already be used to conducting business using a pre-existing set of indicators – which may be much more basic. Despite having successfully advanced through all of the previous steps, this switch over among users may still require special attention and actions.

If you’ve already ensured prior engagement, chosen the right teams and set out an implementation and pilot strategy, this will certainly go a long way in making the process of adoption much smoother and more natural for your users. The challenge posed by this step, therefore, has a lot to do with how well you’ve ensured a connection with your professionals during the previous steps. So now, how do you accelerate engagement?

Implementing new dashboards will often require making other associated changes. It may mean introducing new indicators or forms of monitoring the business that did not previously exist or even replacing the way professionals are used to approaching indicators with one based on a platform that’s more modern, user friendly, automated or efficient.

The trade-off that companies must consider in the first part of the adoption phase is:

• introduce an entirely new set of indicators or
• update the way that you deal with the existing set of indicators.

Based on our own experience, opting for a mixture of both is optimal. Moreover, best practice suggests that you should initially start out by updating the approach and subsequently introduce new indicators. The risk of “going all out” with a new set of indicators is that this may cause an unmanageable disruption to the day-to-day work of your professionals and negatively impact the way they analyze results.

Dealing with a new format, a new platform or unfamiliar indicators can easily be interpreted as being overly complicated. You’ll need to do much more convincing and offer much more training to get everyone on board and aligned. While striking a balance between new and merely updated indicators is important here, initially, simplicity should be prioritized during the adoption phase. This will ensure a sufficient degree of continuity with the way things were generally done before.

What does it mean to empower areas of your organization? In the context of the data-driven transformation, it means allowing your intelligence teams – analysts, coordinators or managers that analyze data on a daily basis – to have “premium” access to the data solutions you’ve provided and have the skills to use them beyond what they were used to. This entails granting access to raw data that will allow them to create new visualizations, carry out deeper-level analyses, understand the underlying cause of a poor indicator and more.

Empowering your business areas is about access and capabilities: unlocking the potential embedded in more profound contact with the data being introduced through this transformation journey. There are, however, different approaches and modalities of empowerment that will vary depending on the needs of the businesses and your organization.

The trade-off here is between spreading this empowerment among a broader range of professionals or focusing on specialists to gain agility. More specifically:

1. making a technical capability a prerequisite and skill expected of the entire team of business analysts, e.g. being able to program or
2. creating internal specialists in the area responsible for managing data.

While both paths are certainly possible, they each come with their own set of trade-offs. Deciding that a technical capability will be a requirement for an entire area is inherently a longer process, requiring you to implement programs for professional training and the employee journey, to adjust your hiring criteria and update your model for performance evaluations. At the same time, going broad will usher in a greater and more profound transformation for the long term while also increasing the organization’s ability to innovate.

Opting to simply assign a team of specialists to the task will be much faster, facilitating internal communication and governance with the team responsible for the overall data-driven transformation. On the other hand, this entails setting a relationship of dependency on certain resources, which can lead to bottlenecks in the future.

A balance must also be struck between

1. centralizing this process to ensure sufficient data governance and scale or
2. granting a higher amount of freedom so that the business areas can advance and adapt.

We’ve experienced both extremes in past projects, along with the impacts that each entails. More complex organizations with a large number of areas, diverse business units and international operations are generally advised to grant their business areas a greater degree of liberty to act in order to avoid a rebound effect that can incentivize these areas to create their own parallel data channels. On the other hand, this decision can very well lead the area leading the data-driven transformation, e.g. centralized business intelligence, to become less influential and more operational, thereby reducing its power to transform the organization as a whole.

The essence between these two trade-offs lies in allowing your business areas to adopt the newly developed solutions. It also means that these business areas should feel that they are the owners of these new solutions, opening the field to collaboration, co-creation and faster adoption.

One of the success factors for ensuring perpetuity involves incorporating the data-driven transformation as an ongoing process. This means creating a new area and fixed team of professionals with a distinct set of roles, responsibilities, career paths etc., guaranteeing that you have the control towers in place to accompany and support the business along this journey.

Companies that view the data-driven transformation as a project with a start, middle and end point seriously heighten the risk of ending up with an unsustainable and short-lived initiative. While this transformation does, of course, demand initial capital expenditures – to get things started, make strategic choices about which paths to take and execute your initial ramp-up – this transformation does not simply come to a tidy conclusion the day that the final indicator has been implemented.