IndustryWeek: How to Handle Big Data, Before It Handles You
When done right Software-as-a-Service systems make data analysis much easier, but getting it right requires deliberate action. Here are some tips.
Remember the old days on the job … say, 10 or 15 years ago? Back then, you probably logged into one or two business systems each day, typically ERP or finance programs, and then augmented that with Microsoft’s suite of applications.
How times have changed. We’re now living in a world where technology is evolving at an absolutely crazy pace, and Software as a Service (SaaS) subscription apps are running the asylum. The 2017 State of the SaaS-Powered Workplace Report highlighted that companies use 16+ SaaS apps on average, up 33% from 2016.
That sounds like great news. It shows that businesses are moving fast, increasing agility, and focusing more on customers, right?
Well… not so fast.
When done right, SaaS systems make data processing and analysis infinitely easier, quickly generating sexy reports and dashboards, accessing company data from anywhere in the world, and gaining insight through advanced algorithms. But getting it right takes a lot of thought, work, and ongoing attention.
For starters, the ease of subscribing to and introducing these products within your organization makes it easy for your employees to bypass the traditional technology strategists in your company. This can lead to scattered and duplicative efforts that lack the benefit of oversight and thoughtful consideration across the enterprise.
Take, for example, cloud storage and file sharing. Chances are good that if you poll employees across your company, you’ll find that they are using a variety of systems such as Dropbox, Google Drive, Box—and on and on. While these habits may seem innocuous, they can easily lead to more silos, a variety of subscription costs that are hard to trace, and having company data sprawled all over the place. Not to mention security risks and the potential to misplace or lose access to key content, information or other company assets.
Those are big problems, but by far the biggest challenge (and benefit) SaaS systems introduce is their capacity to generate and support massive amounts of data.
Remember how I said companies are using 16+ apps on average? Consider that every single login generates new data, and multiply those changes and additions across dozens, hundreds or even thousands of company employees. That’s a lot of information to keep track and make sense of.
Let’s start with the root of the problem: when we put garbage in, garbage comes out. I’m talking about data quality here, and it’s probably one of the most widespread issues my team sees in the world of Industry 4.0.
Sloppy or incomplete user input, misaligned third-party integrations, and poorly defined (or entirely absent) data governance policies can pollute datasets to the point of obsolescence. And that amounts to a huge waste of resources.
In fact, an article in the Harvard Business Review recently pointed out that poor data quality is enemy number one preventing the widespread and profitable use of machine learning. And one study found that business and data analysts end up spending 50 to 80% of their time cleaning up datasets. Given that these individuals are typically highly qualified — and highly paid —- experts, do you really want them spending their time fixing typos and deleting duplicates?
Many of these problems can be prevented by taking a strategic approach to data management. As with any other critical aspect of your business, this will require a comprehensive and well-executed plan supported by resources and a commitment from leadership.
Start by assembling a cross-functional data management team. This should include your technology specialists, but it shouldn’t be limited to the IT department. SaaS has empowered everyday users to employ technology to work smarter and more collaboratively. Our goal here isn’t to squash this engagement, but rather to offer support and unite your entire company toward a common strategic purpose. Try to have members of most major departments represented on the team, and be sure to include a C-level sponsor who will own this initiative and drive buy-in and action.
Once your team is ready, I recommend focusing their efforts on three fundamental data management priorities: process and system alignment, user adoption, and maintenance and governance.
Process and System (Application) Alignment
Start by taking a comprehensive look at your business processes and understanding how these processes generate data. Approach this step with two goals in mind: 1) reducing barriers to data input and maintenance and 2) enhancing and encouraging data use.
Reducing barriers to sound data management
Start by identifying a common set of SaaS tools that will both effectively and efficiently support the work of your team. If you can find tools that people in your company have already independently adopted, even better, since this shows that your employees find these resources to be helpful and user-friendly. Use these core apps to build out a company-endorsed toolkit and develop mini rollout strategies to promote their widespread use and adoption. As with driving adoption (see below) for your overall data management program, the key here is to communicate the benefits, provide training, and hold people accountable. Make it clear that team members are expected to use these resources collaboratively and that using alternative systems is unacceptable. For example, if you’ve decided to include DropBox in your toolkit, then no one should be storing work files on Box.
Next up, tackle the paper and spreadsheets. If you’re still conducting any part of your business on these, your team needs to develop a plan to bring this process into the 21st century. Once you’ve stopped the bleeding and have implemented a digital replacement, work on transferring legacy processes and information into your new system(s) as needed.
And finally, look at everything on a small screen. Chances are, your employees are logging quite a few of their work hours on a mobile device. If your data capture and retrieval processes aren’t mobile-friendly, you’re making it unnecessarily hard on your team… and that leads to sloppy, missed or incomplete data input and lower levels of user adoption.
Enhancing and encouraging data use
Once you’ve built out your new systems and processes, consider how your SaaS applications can support your overall data management and analysis goals. Can you mine data created as a result of your workflows to better understand your company, your customers, or your team? Can you build API connections between services to streamline information sharing? If your company has implemented an enterprise-level business platform, merging your data within a single system can give you a more comprehensive picture of your business. But be judicious when deciding what data to capture or integrate. The last thing you want to do is bog down or muddy up the very tool that is supposed to offer you agility and clarity.
Next, brainstorm ways your company can effectively use the data you’re mining and analyzing, and weave these practices into your day-to-day operations. One of the best strategies I’ve seen is incorporating dashboards into team meetings. By creating dashboards showing progress toward KPI’s, you can both keep your meetings on track and offer valuable insight toward the work that matters most to your organization. Plus, when employees know the data they’re entering will be reviewed and discussed publicly by their colleagues, you’d better believe they care about data quality.
Let’s start with what success looks like. Organizations that have achieved meaningful user adoption make data maintenance, analysis, and use a common part of day-to-day workflows. Everyone in the organization understands that maintaining data integrity is a core responsibility of his or her job. And individuals across the company know how to use data to work smarter. Achieving this level of user adoption relies on three primary factors: communication, training, and accountability.
Ensuring your team knows how to maintain and use data is only one half of the equation. You must also communicate why it’s important. Be sure to share outcomes and insight made possible through data use so that employees understand the benefits of their work. Conversely, illustrate the pitfalls or problems created through sloppy data entry or management.
Data management and information literacy should be integrated into your company’s training initiatives. To keep these issues top-of-mind, I’d recommend holding at least two webinars, lunch and learns, or workshops on these topics per year. Additionally, you should supplement these sessions with regular informational updates. Sharing a weekly tip or article will remind team members of your company’s focus on data use and quality and help them continue growing their skills.
Finally, as with any other aspect of your company’s operations, you must be prepared to provide oversight and hold your team accountable. Set clear standards and goals for data use and monitor employees’ performance toward meeting these benchmarks. Incorporate these metrics into regular feedback or review sessions. And, most importantly, ensure that users who create data hygiene problems are responsible for cleaning them up.
Maintenance and Governance
If process alignment and user adoption are your goals, your organization’s data maintenance and governance plan is your “how to” manual for accomplishing them. Your data management team should be responsible for developing a practical, actionable roadmap for preserving data quality, encouraging user adoption, ensuring responsible and safe data use, and promoting ongoing alignment between your business goals and your data management practices.
This guide should be specific. At a minimum, it should outline clear roles and responsibilities for data hygiene, security, and strategy and provide detailed instructions for how users should enter, interact with and use data within your organization. For example, who is responsible for performing routine data hygiene efforts? How often and how do they do it? Who holds your team accountable for maintaining data quality? What are the repercussions for failing to adhere to data standards?
This guide should also outline your data security model, which should be designed to ensure that team members have access to exactly as much data as they need to successfully do their jobs (and no more). It should also cover database design specifications and incorporate tools such as automation, validation rules, help text, and data dictionaries to limit the instance of user error and uncertainty when entering or data or building integrations.
If all of this sounds like a full-time job, it is. That’s why companies that are serious about digital transformation are hiring staff members with titles like “business analyst,” “data scientist,” and even “chief data officer.” They know that tackling these issues will take a great deal of time, attention and resources.
However, the costs of poor data management can be devastating. At best, your investments in SaaS tools and systems will be diminished. At worst, poor data management can lead to inaccurate pictures of your business’s overall health, lost inventory or assets, costly mistakes, poor customer service, or missed opportunities to capitalize on or react to emerging trends like AI and IoT.
SaaS will continue to serve as a major driver for digital transformation. Companies that are able to effectively incorporate, orchestrate and leverage these systems—and the mountain of data they can generate and process—will have a crucial competitive advantage in the marketplace. Those that don’t may lose big.
Andrew Rieser is President and Co-Founder of Mountain Point, a digital transformation consulting firm specializing in the manufacturing sector. He has nearly two decades of experience in designing and implementing digital business processes.