Arrayo spoke with industry executives and experts about keys to success in data governance. We talked to Chief Data Officers, Chief Digital Officers, CFOs, and experts in finance, pharma, biotech and healthcare. Our interest is to find an answer to one simple question: What works?

The people we spoke with come from a variety of organizations — some companies are young, agile startups, while others are mature fortune-500 corporations. Despite these substantial differences, everyone we interviewed is struggling with some of the same issues: regulatory pressure, low productivity in analytics, operational inefficiencies, and frustrated data consumers.

A common factor is your firm’s data culture.

Understand the Levels of Data Culture

There are three levels of data culture.

Level 1 — Data Heroes Like the “IT Heroes” before them, they take “raw” data and perform on-demand, manual steps to get the data into a usable shape for their clients. Immense amounts of time are spent joining, aggregating, enriching, and reconciling datasets before the data is delivered to the final user. To do this, you need experts who understand how to extract data, what it means, and how to prepare it. This takes time and effort, and it’s not easy to repeat or reproduce across use cases.

Level 2 — Data Guardians The Data Guardians know where the data is, know what it can be used for, understand how to extract it, and know how to make use of the data. Data Guardians are the gatekeepers of the data. Although manual processes still exist, most data related processes are automated or semi-automated. Data consumers know which data guardian to go to for their specific data needs.

Level 3 — Data Nirvana Data is democratized, understood, robust, well-defined, and available in a self-serve model to all who are entitled to use it. There’s one source of data truth for the world.

Your organization may be well on the way to “Data Nirvana”, or you may be just starting the journey. What does Data Nirvana look like in practice? As one of our respondents put it, it’s “straight through processing without manual manipulation. Clear recons. Data lineage that shows you the technical underpinnings. Reduced time to insight in analytics. The right testing and monitoring to ensure it is working, and all of it certified in blood.”

How close are you to Data Nirvana? Much of it depends on one key factor: your data culture.

Here are 7 tips to help drive your company’s culture towards “Data Nirvana”.

Assess Data Literacy

How data literate is your organization? Are your key people data-centric, or document-centric?

If leadership has been doing it the same way for 20 years, they don’t have enough practical examples under their belt to see how new concepts will work. One way to do this is to show a quick win. For example, in marketing you need to work internally with traditional marketers and with new media marketers. Marketing is one area that is undergoing a massive cultural shift, and you can leverage this shift to bring data awareness into the picture.

One of our respondents created data literacy “Lunch and Learn Data 101”, an educational series for their internal marketers. The series was open to everyone and started with the basics. In training sessions of this kind, it is imperative to discuss real-world examples that are familiar to your audience, and to use those examples to present old problems in a new light.

Next, get a handle on practice areas and overall culture.

• What practice areas are driving revenue and what is their data pain?
• What groups depend most on data instead of documents?
• Are you working within an engineering culture or a sales culture?

These questions are key to crafting and fine-tuning your message. For example, make sure that a sales culture understands that what they are selling isn’t a product, it’s data. A Financial Services or HealthTech company isn’t selling technology, it’s selling the inherent value in the data they provide.

Find your Allies

If you don’t already have them, start searching for allies outside of your immediate group. Find the people who feel the data pain and enlist their support to create a data-centric culture. People who tend to feel the pain are data consumers — anyone with a report to file, a customer to reach, a patient to treat, or an analytical problem to solve. These are the folks who will rally behind you when it comes to fundamentally changing data culture. When they speak to data producers, you’ll find that their voices will help shift the conversation towards best data practice. When meetings are set up between key data consumers and data producers, results can be amazing. Bring specific examples of how data in one area is affecting key functional groups such as marketing, finance, analytics, and sales.

Another source of leverage is to find synergies with other groups. Reach out to your colleagues at the enterprise or other regional levels. By combining forces, you can get the message out simultaneously across many levels. If you’re in a branch, find symbiotics with the head office. Enterprise level will be thinking about analytics, but they may have a blinkered view of what’s going on. If you can close control gaps and eliminate operational inefficiencies, enterprise will always listen.

On the tech side, you need to identify the people who are closest to the operational data. Find the ones who understand the data and work tirelessly to deliver this data to data consumers. This tends to be people working extensively with the specific data being utilized (operations, commercial, scientific, informatics, clinical etc.). Look for the people who know the data as part of their daily business, and who can validate it across silos. These key personnel know where the disconnects are and can help you plan to bridge the gaps. If you use a lot of vendor data, find people who can keep on top of trends and understand the vendors in each space.

Remember — data does not live on its own. The people in the various business lines know data as part of their daily lives and use data on a real-time basis. If you can keep them on the front lines of the fight for best practices in data governance, data management and the data processes surrounding it, you will begin to understand points of entry where your problems originate, and valuable solutions can be implemented. It’s far cheaper to catch data problems at the source than it is to fix at quarter close. The business line people are the ones who can help you with data and processes required as it enters your ecosphere.

Another key to success, of course, is your sponsors. Do you have the right ones?

• The Chief Risk Officer, the Chief Admin Officer, Chief Scientific Officer and the Chief Digital Officer can be good sponsors for data initiatives
• The CFO (as long as you can show ROI) is a powerful sponsor for data projects too, since they are adept at tying data performance to financial performance

Look intensively at the strategy of the company from the CEO or division head’s perspective: where does data fit as part of strategy? You need a mandate to act. How transformational does the C-suite expect the Chief Data Officer to be? Using this knowledge, ask yourself how best to build expectations surrounding data initiatives and timelines.

Remember, there will be some people who won’t go on the journey to “Data Nirvana” with you. It’s painful, but you have to get a groundswell going. Utilize the ones who will come with you and who understand the power of data.

Increase Buy-In with Data Confidence

Once everyone starts to understand where the pain lies, it’s time to show what a good data culture looks like in practice. Now is the time to increase buy-in across the organization by building confidence in data. It will take time and focus. Here are some tips:

• Find the critical processes that could benefit from automation, but which have many manual touches due to a lack of confidence in the data
• Go through enough cycles of reducing manual touches so that the people on the ground are confident in the data they are using
• Keep in mind that an iterative process is key and be prepared to invest time
• Prepare for lots of communication strategy, lots of handholding and a strong understanding of how good data practices can integrate into business as usual
• Early in the process, the Chief Data Officer can place people from her group to sit with the data consumers and prove out the data for each reporting cycle or each analytic pass so that confidence in the data becomes a foundational element

All of the above are great ways to build out a portfolio of data projects that will increase data stability and confidence.

Another success factor is to build consensus as you go. Start by getting your own team’s input first: make sure you have your signposts lined up, and you know where you are going. Then, take it on the road for refinement.

Transparency increases confidence, too. People want to know where the data comes from. They want to see the data journey; they want to see that someone is measuring and monitoring data quality. Build confidence in the data at each level by building transparency. In analytics, if you know the values that are going in, your stakeholders will have more confidence in the algorithms and the decisions they drive. If you can show where in the data journey things are going wrong, that tends to open lots of eyes. Everyone wants artificial intelligence (AI) and machine learning (ML), but make sure people know you have to fix the data foundations before you can deliver creative technology on top of it.

Leverage Important Business Initiatives

One of the best ways to raise data awareness is to solve fundamental business problems. It’s hard to get people to see a benefit to their work if data quality is a hypothetical goal. The most successful data executives are able to integrate their goals into specific business initiatives to show value quickly. Find the business groups that have the most mature and refined data practices within the organization, and work with them first. As one respondent puts it, “you should make life easier for the people who make the money. Smooth the way for them.”

Once you have a short list of initiatives, make sure to define the change you’re trying to implement:

• What’s the mechanism of that change, and how will a business initiative be affected?
• Are you going to use data to improve patient care?
• Speed time to insight?
• Red-flag risk before it happens?
• Change patient experience?
• Create straight-through financials?

That puts boundaries on the problem — you know that variability in data will have a negative impact on the process, and you can articulate what can grow out of having robust data and how it ties to a specific initiative.

In the beginning, you’ll need to think like a startup. Do a lot of POCs (proof of concept). Once you find stakeholders who are hungry for data and eager to tie your work to their initiatives, get them involved in the POC from start to finish. Communicate outwards from your POC. Make sure everyone understands how good, robust, high quality data is critical to the success of the initiative and use the POC to lay out the plan to achieve this. Going forward, every initiative should include a data plan to achieve a similar success while building standard practices and methods that can be reused across projects.

When you show that what you do is working in the context of a business need, you can build advocacy. Leverage this to bring it to another group and show what you can do and communicate the critical role that your team performed.

Connect the Dots

Operationally, data is scattered between systems, files, and someone’s spreadsheet. Your goal is to get people to think of data as a continuum, not as a chunk. Inevitably, every organization will have to start managing the data lifecycle: that means knowing where new tools and IT systems fit into the entire data landscape. However, people who are putting out fires aren’t interested in the view from a thousand feet up. You can change this if you take a specific example and map it out in a one-page picture. When there is transparency in the data life cycle for even one small piece of data, you’ll get them to see where they fit into the larger picture.

It’s vital to connect the idea of good data to people’s daily business. Depending on where you sit in the organization, you can demonstrate the value of good data practice by speaking to pain points in a specific group within the organization. What do they need in order to…

• Reduce regulator pressure?
• Make the connection between poor quality data and higher operational costs?
• Increase operational efficiency in clinical trials?
• Understand what key assets are in the pipeline?
• Determine how many times an identical data source has been procured?
• Ensure cross functional data sharing to enable drug discovery?
• Enroll the correct patients to ensure success in a clinical trial?

Lay out a clear vision for your colleagues and make sure you’re speaking their language. As one person we spoke to says, “when you say the word data, their eyes glaze over.” If you can tie the concept of robust data to specific goals, you can get the attention and buy-in of your colleagues. Don’t spend all your time and budget on documentation — use it to make a real difference and then communicate that difference to everyone, so they can see how it works for their world.

Connect across silos. For example,

• Marketing data needs to be married to sales and distribution
• Preclinical data needs to be married with clinical trial data for biomarker discovery and companion diagnostics
• Trade-desk transactional data has to be married to counterparty attributes for risk and regulatory reporting

That’s a change to the mindset. It means different groups will have to talk to each other about data meaning, procurement, partnering, and data strategies. If you can use a POC to link disparate datasets in service of a business solution, people will start to connect good data to business possibilities.

Use the carrot and the stick. Some of the most powerful tools you have in changing data culture is the use of accountability. Make the data producers — often, that’s the front office — understand they are accountable for any inspections that find data gaps. Can you fold accountability into reviews? Have a dialogue with key people who aren’t monitoring their data as they should, and make data quality a goal for them. Get culture and conduct reviews involved if there’s no change.

On the other hand, you can’t just whack people on the hand if they do wrong. You have to support them in their efforts to do the right thing. Give them good tools — they can’t behave well if they don’t have the right tools. Support efforts with lots of personal attention and hand holding. Keep the attention on post-implementation, too — don’t just build and bolt.

Show, Don’t Tell

Show the potential of data power. For example, what do you do when new datapoints come in? In Health Sciences, new devices such as wearables (think Fitbit) are generating new data types that need to be integrated into an existing picture. There are a lot of sources, and you need to show how this data can be democratized and used reliably.

Show your impact. You can see immediate financial impact that senior management will recognize if you can show how data quality impacts capital. Carrying more capital than you need to, because of bad data, is a real expense. Show the exact figures. If you can prove out the financial impact or the regulatory impact, you can build a solid case. Once you’ve done that, you can go to the heart of what the key business users want from data — how to reach more customers, close more deals, or sell more products.

Show the improvement. The goal is to make life better, right? Show how you do that. Show how your work has bridged data silos to improve a business process or solve a business problem. Don’t be afraid to show missed opportunities that were due to data being hard to use, misunderstood, or slow to production. Use past fails to show how to improve the data culture, and the benefits that can be delivered.

Sell yourself. Make sure you communicate that there are projects that people want to do, but they can’t achieve success without good, trusted data behind the plan. Make sure everyone understands what that looks like.

Show senior executives what you do. You must have the stamina to sell yourself constantly, because you are shifting something that is a subconscious resistance to a new way of doing things. That means never letting up. To sell effectively, don’t lose sight of what you’re selling. Keep the message clear and simple.

Conclusion

Smart companies understand the value of data. Successful companies know how to effectively utilize all of their data assets. In the age of big data, distributed systems, interconnectivity, analytics, machine learning and artificial intelligence, and regulatory and risk assessment requirements, it is becoming increasingly clear that good data governance is essential for successful data use. In financial services as well as life science, data culture affects the success of your data enterprise. The strategies discussed in this article are employed by successful companies across industries, leading to better data governance models and ultimately to better and more efficient data utilization.

We hope you have enjoyed this article and would love to hear any thoughts or reactions you have to our article. Feel free to reach out to contact@teamarrayo.com or write a response down below.

*An electronic version of this article can be downloaded from our website.

*This article was written for SteepConsult Inc. dba Arrayo by Renée Colwell.