Seven pitfalls to avoid when using big data to power digital transformation
As more companies seek out digital transformation projects, big data offers an opportunity to gain a competitive edge -- if leveraged correctly.
"Big data can help with digital transformation," said Gartner analyst Douglas Laney. "It's incumbent upon anyone who has data to design ways to leverage that information asset in a multitude of ways, in order to better cover their costs of acquiring, administering, and applying the data."
As of 2017, 41 percent of firms have either implemented big data, are in the process of implementing it, or are expanding their existing implementation, according to a Forrester study. An additional 26 percent of firms are planning to do something in the next 12 months, while 29 percent have no plans, and 3 percent of firms do not know if they have plans to do so.
The previous year, more companies said they would be taking on a big data project, so researchers expected that 41 percent figure to be over 50 percent, according to Forrester analyst Brian Hopkins. It appears that many executives ended up deferring big data projects, he said.
SEE: Big data policy (Tech Pro Research)
For digital transformation success, it's key to not let big data become a project unto itself, Hopkins said. "Don't have a data strategy -- have a business strategy, where you create competitive advantage through data analysis and insights," he added.
Several common challenges tend to arise for companies seeking to inform digital transformation with big data. Here are seven pitfalls to avoid in your enterprise's journey.
1. Forgetting the purpose of the data
Businesses that are most successful in digital transformation efforts tend to make big data investments, Hopkins said. However, "firms that do this right don't do it just for the sake of doing it -- they do it because they understand how to use these technologies to create differentiation and competitive advantage," he added.
Using big data properly is about generating actionable insights that can inform and improve decision-making. If a firm loses sight of that, they may end up wasting their time and resources, Hopkins said.
Companies have to know what advantage they want to gain, and how they can use insights from big data to reach it, Hopkins said. "All that comes ahead of deciding to buy big data technology," he added.
2. Failing to monetize your data
Most organizations assume that they use a data set once, and throw it away. "You need to think really broadly about the ways to leverage any drop of data, because you can use it over and over again, and you want to do that before it becomes stale," said Laney.
Data monetization is not just selling or licensing your data, but is any way that you are generating measurable economic benefits from it, Laney said. Creating larger reports for the sake of creating them isn't effective, Laney said. Instead, businesses should focus on monetizing their data.
Gartner has analyzed hundreds of big data use cases. When focusing on those that generated economic benefits, "almost none of them have to do with building a pretty pie chart," Laney said. "They all have to do with targeting big data at a business solution to drive some economic benefit."
3. Looking only to your own industry
Enterprises often look only to what others in their industry are doing with big data when it comes to digital transformation.
"My flippant response is always, 'Why do you want to be in second place?'" said Laney. Instead, executives should look to other industries ranging from retail to healthcare and consider how they can replicate those ideas within their own industry. That way, you could be the first, and get a jump on the competition, Laney said.
"Companies need to stop staring at their own labels when it comes to their industry and the data they have," Laney said.
4. Ignoring external data
"Folks are aware that there are other sources of data out there, but they're consumed with the data they collect internally," Laney said. "They don't realize the correlative benefits of bringing in data from social media, from syndicated data providers, or from harvesting web content."
Additionally, most companies work with several third-party partners from whom they could negotiate data. "The companies that are really thinking out of the box are the ones that are going out and grabbing all this freely available content," Laney said.
SEE: Digital transformation: A CXO's guide (TechRepublic/ZDNet)
5. Failing to inventory your data
Many companies claim they want to manage their information as an asset, Laney said. However, they don't include data in inventories, so they don't actually know what information they have. "They don't measure it, so even if they know what they have, they don't know what its potential economic value is," Laney said. (Gartner has published a model on how to measure a dozen different data qualities, and how to quantify information as if it were a balance sheet asset.)
While working with an energy company on a data strategy, Laney was told the company only inventories "major assets" such as transformers and generators. After the meeting, he went into the men's room and saw that the urinals were all inventory tagged. "This comes from a company that says, 'We only inventory our major assets. We don't include information, but do include our toilets,'" he said. "That's the kind of disconnect I see often. Most companies track their laptops, but don't track the data on their laptops."
"When companies tell me, 'We want to manage our information as an asset,' job number one is to inventory it with the same care that you do your other assets," Laney said.
6. Centralizing ownership
As part of a daily governance practice, most companies will establish owners of data, Laney said. While it's important to have accountability and responsibility for information assets, ascribing one owner can perpetuate "information hoarding" in the business, he added.
"We're bigger fans of terms like a 'trustee', or someone like that," Laney said.
Another pitfall is when data is seen to be owned by IT exclusively, Laney said. "Unlike decades ago, when data was tightly coupled with the technology where it was used, today, data and technology are very decoupled from an architectural standpoint," he said. "From an organizational standpoint, they should also be decoupled."
7. Failing to plan for the future
Big data technology is moving at an extremely fast pace, Hopkins said. Many companies choose a best practice architecture from the year before, and then spend a year implementing it. "By the time they get it, it's two years out of date -- that's a lifetime at the speed digital is moving," Hopkins said. "When I talk to tech buyers about how to invest for digital and insights, I tell them, 'The only thing you should be investing in is making yourself able to move at the speed of digital.'"
If you choose to get locked into a architecture because it's more cost-effective, you won't be able to keep up, Hopkins said. "You need to build architectures that are very flexible, so as your business figures out what it wants to do, or pursues a new market or product, you have the flexibility to adapt and keep up," he added.
For this reason, big data in the public cloud is going to fully eclipse on-premises data within the next five years, Hopkins predicted. "Everyone moves to the cloud for the cost, but the real reason will be to keep pace with innovation," he said. Advances in artificial intelligence, machine learning, and quantum computing will eventually impact big data stored in the cloud, Hopkins added.
Also see
- Learn these 3 languages now if you want to become a data scientist (TechRepublic)
- Data analytics: Five ways to make it work for your business (ZDNet)
- Building a big data solution? Here's why it's important to consult users first (TechRepublic)
- ASPI urges government to see beyond the 'hype' of big data (ZDNet)
- IT leader's guide to achieving digital transformation (Tech Pro Research)