Thoughtspot 2016 Anti-Predictions for anti Analytics

2016 is here, and with it comes the traditional barrage of prophecies for the year ahead. Year after year, we hear leaders in the tech industry make predictions for what’s to come, all with an uncanny alignment between each vendor’s predictions and their marketing pitch.

When our marketing team asked me what my predictions for the BI industry might be, I thought it would be fun to un-hype the hype with some “anti-predictions.” So here’s my list, looking through my shiny crystal ball:

1. Big data will not be big after all

The term big data will soon be history, and this year officially marks its death. Gartner has removed Big Data as a theme from its hype cycle reports. At the same time, Big Data projects aren't getting any better. Gartner Predicts 2015 reported that failure rates can be as high as 60% for some big data projects; yet another reason big data isn't living up to its name.

2. You won’t hire enough data analysts or data scientists

The reporting backlog is only going to grow for companies that rely on elite teams of BI analysts. McKinsey predicts that by 2018 there will be a shortage of 1.5M data experts, and a recent TDWI survey found that the average business user is waiting 3-4 days just to get a report updated. In fact, Gartner says that the number of citizen data scientists, aka non technical folks who want to experiment on their own with data, will grow five times faster than the number of expert data scientists. So expect that backlog to keep growing.

3. Search will not become the standard for enterprise analytics

While we at ThoughtSpot and a few other vendors (Microsoft PowerBI, IBM Watson) are betting the farm on search being the technology that will eventually make data truly accessible, we don’t think 2016 is the year it catches fire in the enterprise. Sure, more vendors will add a search box to their solutions over time, most customers still don't believe search can fix their BI problems. Search-based BI has been in conversations for at least a decade, but no vendor has been able to make it work. Vendors that are betting on search need to overcome years of accumulated disbelief before search will be adopted widely for enterprise analytics.

4. BI adoption won’t climb above 25%

According to the annual BI Scorecard survey, only 22% of employees actually use BI tools. That's because despite all the hype around self-service BI, these tools still require data experts to act as the middleman between data and the average business user. The result is that products keep getting more technical and adoption rates remain low.

5. All your data won’t move to the cloud

Despite the proliferation of cloud services, there is still roughly 99 percent of data sitting on premise today. For analysis across enterprise data sources, it’s easier to pull down the small fraction of data from the cloud than it is to pull all on-premise data into the cloud. If that doesn’t convince you, consider your WAN bandwidth: WAN bandwidth continues to be bought and sold in megabits, while data volumes are approaching petabytes. Moving this amount of data to the cloud is not practical, let alone secure.

6. Prettier visualizations won’t make smarter decisions

Your charts and images are only as good as the data underneath them and the decisions people make because of them. While they may look good and answer your first answer, how do you answer the follow-up questions? Many traditional BI tools nail the data visualization piece of the puzzle, but fail to guide the average, non-technical user to meaningful answers.

7. A.I. won’t steal your job

There’s been a lot of talk about the disappearance of humans in the workplace with Elon Musk, Stephen Hawking and others claiming that computers will eventually overpower humans. But don’t expect a machine takeover any time soon. According to Forrester, only 9 million jobs are actually at risk of being replaced by machines over the next 10 years. That’s only 6% of all US jobs. A man vs. machine Armageddon isn’t on the cards for 2016. Not for business analytics, anyway.

8. Cognitive computing won’t materialize beyond a science project

While the technology behind Watson and other efforts is amazing for suggesting exotic food recipes, it isn’t anywhere close to being ready for crunching numbers for business users in a trustworthy manner. The problem with such approaches is accuracy, or the lack thereof. Would you make financial decisions based on an analysis that read, “There is a 30% chance that suppliers from Japan are the biggest contributors to your costs”? Or one that said, “There’s a 50% chance that you sold 3 million dollars worth of Nike’s in California last quarter”? We didn't think so. It’s not that useful.

Only time will tell whether any of these predictions hold weight. Yet it’s important, either way, to question the buzz and hold ourselves accountable. We need to look at the facts and trends that exist beyond our marketing pitches and product roadmaps, even if they don’t always align (for now).