Not long ago, big data was a niche topic. But, fast-forward a few years, and today it’s the fuel driving most data-driven businesses. What's more, big data is informing strategy, boosting operational efficiency and accelerating growth.
Financial investment has meanwhile kept pace with the data boom. Some 75 percent of organizations say they’ve spent on advanced big data infrastructure or plan to in the next couple of years. And a flood of new and exciting big data-focused startups have arrived to satisfy the growing appetites of enterprise customers.
But as attractive as the big data space is, it can be a tricky sector for new entrepreneurs. Consider that 66 percent of all startups fail within the first 12 months. Though big data businesses may feel confident that they’re entering a hot space, stumbling blocks await. They need to be anticipated and considered.
Here are five of these critical challenges facing big data startups today.
1. The lack of talent
Big data is the definition of a growth market: Six out of ten IT decision-makers expect their organizations to spend more on big data initiatives this year, with only 5 percent anticipating any type of budget decrease. But, while the category has ballooned, the biggest question is whether this growth will outpace the talent needed to implement it and scale its adoption. And the answer is yes, that problem is real.
According to the McKinsey Global Institute, the market will be short 1.7 million capable big data professionals in the United States by 2018 -- about the same time that the U.S. data market is set to be worth $41.5 billion. As the category expands, the skills gap will widen. Startup staffing and personnel -- the key to competitive advantage in an aggressively growing space -- will become even more challenging. There is no easy solution, as the only true fix is years away, when the talent pool will have naturally deepened to meet the market demand. (There’s also a hint of irony here because many big data startups, trying to help mitigate the lack of talent in the market via their own software, still face the same problem.)
2. The cost of talent
Some 71 percent of businesses and IT organizations consider themselves either average or lagging when it comes to leveraging data. There’s clearly a need to improve overall talent capabilities and education for the existing workforce. However, when staff training is paired with the shallow talent pool that currently exists, to keep up with newly developed products and approaches entails a significant cost.
Collectively, such operational expenses ran companies over $130 billion worldwide in 2013; and, given the fast-paced nature of the data business and the subsequent need for more personnel and continuous training, these costs will only skyrocket for entrepreneurs hoping to plant a flag.
3. Fighting mis-education
In a recent piece about Hadoop, the Wall Street Journal’s Deborah Gage said, “Few buzz phrases are freighted with higher expectations than big data.” The rising “hype” and anticipation of big data’s potential that Gage described have pushed many organizations into blind adoption: They're eagerly embracing tools that often don’t meet their needs, simply because those tools seem to be the most popular (Hadoop being an example).
Further complicating matters is the fact that big data platforms are inherently thick. This makes it challenging for vendors to articulate functionality and benefits and even more difficult for adoptees to understand the platform. This is why, according to Gartner, “Through 2017, 60 percent of big-data projects will fail to go beyond piloting and experimentation and will be abandoned.” Mis-education is ultimately stalling true scale and success.
4. Funding obstacles
The data world is blessed with VC attention and eye-popping funding, as proven by Hortonworks and Dataminr’s recent $100 million rounds. But in many ways, the race for cash is becoming detrimental to new players.
As the sector has grown crowded, it’s more challenging for the typical entrepreneur to separate his or her company from the rest of the pack, because so many companies like Palantir, MongoDB and Mu Sigma (all of which have at least $200 million in funding) are already well known by the investor community. And, because funding has increased, at some point we can expect investors to become more tentative in their commitments to earlier-stage ventures and instead invest in more established up-and-comers.
5. Big competition
With global spending on big data predicted to hit $125 billion in 2015, startups aren’t alone; they face brutal competition from billion-dollar legacy companies like SAP, Microsoft and IBM. These giants can release feature updates to products that cannibalize entire startups. Their capital is infinite, while startups have to focus their efforts more narrowly simply to maintain their cash-burn rate.
And this is actually a good thing. The best way for startups to find success and maintain leads over massive competitors is to focus on a niche and do it well. Larger companies are always on the prowl for ways to gain the competitive edge. So when startups prove themselves to be a standout in a particular niche, across storage, preparation or analytics, bigger players that may be considering entering that area will typically find acquisition a more attractive alternative than building those capabilities in-house from the ground up.
The lesson here: Building a successful big data business isn’t for the fainthearted. But if you anticipate and prepare for the five challenges described, you can play a major part in the category’s continued and impressive growth.