One of the great joys of leading organizations is to create a vision, define a strategy, and do their implementation, including obviously the organizational change aspects. Very often, once I created this vision, I am confronted with either proposing a strategy that speeds up the creation of immediate value for the business while making shortcuts, or take more time to build the right foundations while doing the right thing, and consequently delay the business value creation. In other words, having on one side the short-term impact, and on the other the long-term sustainability.
As we all know, nothing is black or white, and we always end up picking a strategy that tries to take the best of both worlds. And I believe this dilemma is even more pronounced in Data & Analytics (D&A).
This post will try to explain the situation numerous companies are facing today and propose 7 areas D&A leaders should focus on to ease our journey.“A dream doesn't become reality through magic; it takes sweat, determination and hard work” (Colin Powell)
A complex environment
- Deal with the pressure of the stakeholders for immediate results,
- Navigate through the politics of D&A ownership,
- Absorb budgets cuts whereas exceptional investments have to happen,
- Struggle to hire the right skills and talents in a sparse D&A job market,
- Suffer from a lack of understanding on the importance of their mission across the company,
- Need to drive huge change management efforts that sometimes go so deep it can even interfere with company culture,
- Face the inertia of organizations that have also been through so many company-wide transformations in the past decades,
- Juggle with compliance and evolving regulations (e.g. GxP, HIPAA, GDPR…),
- Ensure the decisions made for today do not put at risk the overall long-term IT strategy,
- Balance the usage of specialized tools and solutions with the overall IT architecture,
- Embrace the pace of technology evolution, especially when it comes to Cloud computing.
There is no silver bullet
1. Foundations, foundations, foundations
2. There is no such thing as a free lunch
3. Crawling before running
4. Technology is no longer the problem to solve
5. To be data driven, data need to become a company asset
6. A flexible organization
7. Educate
Jacques (Rene) Rancourt
x
Unless you are a newly created company embracing digital at the core of your business
model where managing, governing and analyzing data are foundational capabilities from day one, you are currently facing a common issue: trying to become a ‘data driven company’. To achieve what is behind these 3 words when you have been operating outside of the right model, takes a humongous effort that is probably an exponential function of the company size and age. From readings to exchanges with peers in other industries, a common theme emerges: all big corporations that have been leading their businesses for decades without anticipating the rise of the digital era are lagging behind today in D&A. The lack of on-going investments in data and analytics for years results in a very painful situation. To catch up, they now have to climb Mt Everest at the pace of a 100m race. And I would add, without the right preparation!! Leaders of D&A need to be the Sherpas of their respective companies, trying to place the cursor of their effort and strategy at the right spot between the ‘black’ and the ‘white’.
To make the situation even more difficult to manage, D&A leaders have to:
By the way, I believe it is what makes this job so interesting, rewarding and motivating. For sure, it is not boring!!!
So, what should we do? Here are some of my personal thoughts on the way the situation could be handled. Each company being different due to its history, culture, D&A ambitions, size..., we will all agree that there is no ‘one size fits all’ solution to the problem. That being said, let’s now look at 7 ideas that can help.
For those I have worked with, they all know I am a true believer of the ‘do the right thing first’. And in this context, it is for me building strong foundations with regards to solutions and data architectures, processes, governance, organizations, technology standards (for sustainable and long-term solutions), data cataloging, master and meta data management, just to name the key topics. Basically, when a house is built, we need to make sure our foundation is solid before building the next floors. In D&A, if we don’t do this properly, we will end up patching systems, create expensive interfaces between them, create complex data related processes, and lower the efficiency of our overall D&A ecosystem, while increasing its Total Cost of Ownership.
Let's be clear. ‘Foundations’ doesn’t mean ‘rigidity’. We have to design our solutions and architecture so it can evolve easily. The usage of API and microservices is a mandatory step as a best design practice (when relevant).
One of the motto we need to keep in mind at all time is 'simplicity, pragmatism and common sense'. It is very easy to fall into the trap of creating a D&A ecosystem that is more complex than needed. Having in the teams experienced Enterprise & Solution Architects, CyberSecurity and Data Privacy experts is mandatory to alleviate this risk. These people will engineer and design the ecosystem so it not only looks good from the outside, but most importantly from the inside. What's 'under the hood' is critical in D&A. We cannot stress this enough.
Business value needs to be at the core of everything being designed. A Value Proposition needs to be clearly articulated, understood and agreed, avoiding at all price the 'nice to have'.
As we build our foundations, we need to create in parallel one-off solutions that will help deliver the first business values. And we need to be explicit that we will most probably have to throw away those when the sustainable long-term ecosystem is ready. Let’s be transparent and all agree as a team it will happen that way.
To climb this mountain, huge investments will be made. We are talking about dozens or even hundreds of million dollars. It has to be acknowledged from the beginning and the reasons why explained. What I observed too often is that the top management cannot easily comprehend why such investments need to happen. We will come back later on this ‘education’ topic. These budgets will be used not only for the final ecosystem, but also for the interim one-off solutions that will be either destroyed or re-purposed later on.
To avoid a tunnel effect, resulting in a big bang approach, it is critical to plan for a modular architecture and deliverables. Some could argue that it could lead to a higher cost but based on my experience it is often very expensive to try to deliver too much at the same time due to numerous dependencies in the deliverables.
In addition, delivering more smaller systems and solutions along the way help creating business trust for the future. One of the problems of D&A is that we are dealing very often with intangible assets. What we build is not always easy to perceive, ‘touch’, experience. Showing evidence that our approach and strategy are the right ones is an area we need to work on to ensure on-going buy-in. It is key to the success.
Agreeing on a list of prioritized deliverables needs to happen in concert with the various stakeholders. And during this exercise it has to be agreed by all that not everything will be done right away. Obvious, but not always easy to do!!
With innovation that happened in the past years (and still happening) in Analytics, Data Platform and Cloud Computing, technology is no longer a showstopper or a roadblock. I am not implying that it is easy to do such implementations, but the capabilities that are now offered can answer to most of our D&A needs. The D&A technology landscape is very rich, and it is up to us to pick what fits best for our respective situations.
As far as Cloud Computing, I am sure we all agree that the D&A future resides in the Cloud. There is no way a traditional on-premise data center can respond to the elasticity and reactivity required by D&A solutions. If you have not embarked yet on this journey, you better start now!! The breath and pace of innovation happening there is mind blowing. The 3 key actors (GCP, AWS and Azur) are competing with each other to our entire benefits!!
On a side note, I am a true believer that data lakes and data warehouses are mandatory complementary capabilities for a ‘digital company’ (among other things obviously). Too often I heard that data warehouses were dead when data lakes came to life. Absolutely not. Long live data warehouses!! Similarly, there is a lot of noise regarding the decline of Business Intelligence (BI). Again, and without going too far, BI has a lot of benefits with regards to industrializing reporting and dashboarding. It has to be a key component of our D&A ecosystem helping to reduce operational costs.
Finally, we need to make sure we identify which technologies are candidates for quick renewal and short life cycles, versus those that are the foundational building blocks with long life cycles. Each technology cannot be changed every other year. This would be way too expensive and risky. Here again we are going back to the architecture topic mentioned before.
In 2017 the Economist published a report titled “The world's most valuable resource is no longer oil, but data”. I fully agree with this. Companies need to embrace this principle. To improve productivity, customer satisfaction, research, manufacturing… data needs to be considered as a key company asset. When we do this, we start allocating necessary resources, assess data value, decide on ownership and usage… In other words, we manage our data as a value creation asset for our company.
Another important point to consider is to avoid mixing data ownership with data usage. Because today many companies are still organized by function, the various departments tend to take ownership of the data they use, resulting in duplication of data and data silos creation. This is a common mistake that needs to be solved with the creation of official data ownership, which is obviously tied to the Data Governance topic. Once this is achieved, data owners become fully accountable for their data life cycle, access, privacy... This then translates into internal data providers, serving the other functions.
One of the toughest parts is to determine which organization should be created to take charge of D&A in a company. I am not presumptuous enough to answer to this question in this section since it is a very complex one with dozens of parameters for each company. Though, from my point of view, and because data should be considered as an official company asset, there should be a major component of the D&A organization that reports directly to the CEO, having transverse responsibility across the company. I don’t believe a specific business should own this responsibility. Unfortunately, we very often see a Global Chief Data Office, or Global Data & Analytics team embedded into another function for the pure reason the head of this unit has an experience or a demonstrated appetite for this topic. If it is an intermediate step, maybe. Though, it shouldn’t be the long-term model.
Another question that needs to be answered is the split of responsibilities (if ever) between the D&A Head and the Chief Digital Officer. Again, not a simple question, rather a big one!!
As far as the operating model of the D&A organization, we should foster as much as possible project-oriented organizations bringing flexibility and reactivity. In other words, instead of a static monolithic organization, it is better to assemble teams according to business needs. These teams should be multi-disciplinary (with members from any relevant function) and be in place just for the duration of a project. It is obviously easier to write than to do but we should aim at that model too.
Last but not least on this topic, it is crucial Roles & Responsibilities (R&R) between IT and Business are clearly defined and communicated, during the implementation phase but also for the operating model. This is another key aspect. Too often we see D&A initiatives fail or dramatically slow down because of blur R&R areas among teams and people. To me, it cannot be IT or (exclusive) Business owning D&A in a company. It has to be a joined ownership, leveraging everyone's strengths and potential, and in a true partnership mindset.
Being data driven requires a change of mindset and I believe this is the most complex topic to address. People aspects are always the toughest ones. In the context of D&A, major change management efforts need to take place at all level of the company for the reasons explained earlier. And we should start with the top management. A way to achieve this is to plan for D&A individual educational sessions with ExCom members in order to explain simply what is at stake, the ‘buzz words’, the concepts, what data driven means, the benefits, and the reason for the cost… With all due respect, I have too often seen top managers too embarrassed or shy to express their lack of understanding, resulting in wrong decisions at the end of the day.
This education should cascade down into the company, so everybody understands and take ownership of the topic.
