Beyond speedy proof of concepts, the reality of cloud BI has never been as sexy as the idea. Historically, most of the data intended to provide 'intelligence' lives in centralized data warehouses, and those warehouses don't live in the cloud. Enter the fun of data latency and integration.
But is the market for cloud BI changing? And if so - why? I went looking for some satisfying answers.
Market drivers fueling cloud BI
On a recent TIBCO Spotfire webcast, Top Ten Best Practices in Cloud BI, EMA Vice President of Research, Shawn Rogers made his case for what he called the 'unstoppable drivers' of Cloud BI. Though Rogers mentioned the obvious - cloud BI enables companies to 'fail faster' and test new scenarios quickly - he outlined new trends that are giving teeth to cloud BI:
- Lines of business are pushing their own BI agenda and pursuing their own solutions - often cloud-based.
- BI users are becoming more diverse and more demanding in their user experience, cloud can provide a better (and more mobile) UX.
- User communities are maturing. Consumers of BI are a 'new breed of knowledge worker', less timid about technology and more sophisticated.
- New technology, from in-memory to big data (NoSQL, Hadoop) is forcing a re-evaluation of existing BI infrastructure.
- The economics of cloud enable more companies to get involved with BI, allowing them to widen the scope of BI projects utilize cloud to widen scope economically.
- Companies are realizing the value of new types of data from new data feeds.
- On-premise data warehouses are under strain from these new requirements and data sources.
- Data warehouses are decentralizing - cloud is an ideal platform for decentralized BI architectures.
- Cloud can minimize the pain of traditional BI projects.
- The CAPEX-based cloud pricing model makes cloud BI projects easier to fund and pursue.
Rogers also argues that a more diverse BI constituency will lead to better organizational decisions. It's a pretty compelling list, but what do the numbers say?
Cloud BI adoption trends - by the numbers
One major cloud BI study comes by way of Howard Dresner (here's the 2014 version, licensed by GoodData). The 2014 version of Dresner's Cloud BI survey polled 853 respondents, a bit more than half from North America, 30 percent from Europe, Middle East and Africa, and smaller samples from Asia Pacific and Latin America. Now in its third year, the study is loaded with more enough stats and figures for a very long lunch break (a vendor review and analysis is also included).
Here's a few findings that stood out:
- More than 75 percent of respondents rated cloud BI models at least 'somewhat important'. Small organizations report the highest existing and planned adoption of cloud BI in 2014.
- Cloud BI has shown visible gains in perceived importance for three consecutive survey years.
- Functional areas rating cloud BI as important are led by sales, marketing and executive management.
- The percentage of enterprises currently using public cloud BI increased by more than 17 percent since last year and more than 53 percent since 2012.
- Cloud BI plans through 2015 are 'noticeably skewed' towards private over public clouds.
Security is always the cloud question we compulsively linger on - the security findings of this study were interesting. As it turns out, the most commonly-cited security requirement cited by respondents is 'none.' Dresner believes this reflects 'ongoing non-critical, unaudited, or unregulated use of cloud BI'. However, certain industries did cite regulatory concerns such as ISO 27001 and HIPAA.
On the other hand, the most commonly cited barriers to cloud BI adoption were security and data management related hassles. This contradiction deserved more analysis. My gut reaction was that it may reflect a tension between using cloud BI for lighter use cases versus mission critical BI projects with sensitive data.
Benefits of Cloud BI
When it comes to cloud BI benefits, Dresner's survey found that cost and ease of access continue to outrank scalability, deployment, and maintenance as the highest ranked cloud BI benefits. On a Birst-sponsored webinar analyzing the 2013 version of the survey, Dresner made a broader case for cloud BI: 'Users want BI solutions that are easy to buy, deploy and use, modestly expensive and offer a quick time to value. The cloud-based BI solutions clearly do that for them... Inevitably, cloud business intelligence is where we will all end up.'
During a March 2014 presentation on cloud BI, Tableau's Senior Director of Analyst Relations, Suzanne Hoffman, provided some adoptions numbers to chew on. As per Hoffman, 76 percent of those who have chosen Tableau's cloud BI offerings have done so because of 'speed to deploy.' Hoffman cited the ability to get up and running 'inside of 5 hours' - or even five minutes - 'if all your data is in the cloud.'
Tableau's online offerings, Tableau Online and Tableau Public, are similar except for a major distinction: Tableau Public is intended for public data sharing and is a free product, without the security features in the Tableau Online product.
At the time of the webinar, Hoffman cited 200,000,000 impressions and 30,000 authors on Tableau Public. Mobile BI has also pushed cloud adoption: accessing the right data on any device has been a big Tableau Online selling point.
The typical users of Tableau Online are the types of early adopters you'd expect:
- Companies with little or no IT support
- Need for an extranet deployment
- Already a heavy cloud user (e.g. bulk of data is already in the cloud)
- Want a flexible deployment: fast startup, option to go on-premise later
- Are seeking a quick trial process
Cloud BI adoption data implies a bright future for cloud BI, but surveys such as Dresner's point to gradual growth, rather than an explosion. Size of company, industry, and geography are all factors to consider before making overly-generalized cloud BI statements.
Cloud BI doesn't magically erase the data obstacles that must be conquered to achieve BI ROI. Though cloud security fears are overhyped, cloud BI does pose its own set of data challenges, and not just integration. IT managers must now content with data uptime, service level agreements, data ownership, threat management, contract termination, and other cloud-based contractual issues.
For SMEs or companies without IT staffs, cloud BI has already taken hold. Same goes for proof of concepts and skunkworks experiments. Plus: cloud early adopters have a ready-made reason to move to cloud BI. Yes, I expect the same line of business trends dragging ERP into the cloud to do the same for cloud BI - but it's going to take a bit longer than it takes for marketing teams to add 'cloud' before BI on the slide deck.
Image credit: Woman changing reality © Sergey Nivens - Fotolia.com
Disclosure: diginomica has no financial ties to Tableau, TIBCO or Birst.