Blog post

Symposium Notes – Day Two Jumps in the (Data) Lake

By Merv Adrian | October 19, 2016 | 0 Comments

InformaticaGartnerDBMSdata warehousedata lakeData IntegrationApache HDFSApache HadoopAnalyst RelationsData and Analytics Strategies

My second day of Symposium 1:1 meetings continued the “security of big data” theme (4 of the day’s 15 conversations – usually, but not always, about HDFS-based data), with a data lake flavor. The concerns were retroactive – often driven by an internal audit. “We built it, now how do we secure it?” is a common question. And “it’s almost all structured data so far,” confirming what Gartner found in the 2016 big data survey. Vendor conversations (4 of the day’s 1:1s) also included a look at security – “how much is this going to matter to my customers? Who can I partner with?” has been a typical thread, and I met with a security consultancy whose practice seems to be ramping rapidly.

Another recurring theme was data ingest, and increasingly, I’m seeing clients turning to their existing commercial offerings like Informatica rather than cobbling together strategies out of multiple open source components. Typically the focus is on managing and documenting the process better. An increasing focus on existing strategic technology suppliers was accompanied by several questions about the continuing viability of Hadoop specialists – “are the big guys going to close their window?” An investor, a vendor and a user all asked, showing a unanimity of concern I’ve been seeing for a few months now.

I’m always struck by how many clients begin with “we don’t have a data warehouse, but we have concluded it’s time to build one.” Hearing this because of our direct connection with users is a great reminder that, as William Gibson said, “The future is already here, it’s just not very evenly distributed.” Organizations span the spectrum from early to late adoption.

Four of Tuesday’s meetings included discussions of the  liaison role between “business” and “IT” – the distinction persists, despite or perhaps because of the increased focus on self-service. And for the first time, I was asked about “AIBI” – intelligent, conversational front ends for analytics that do more than just provide natural language but also advise, and when they will dominate the market. “Cortana, are margins growing in the Northeast on our online purchases? What’s the best model for improving that?”

And, of course, cloud. One intriguing discussion challenged my usual economic expectations. A real estate investment trust pointed out that for their business model, “Capex is actually OK with us – we don’t pay taxes the same way. We have to pay 90-95% of revenues in dividends; we don’t get retained earnings. We’re measured on income from operations.” So their perspective on the economic benefits of cloud computing was quite different; they don’t mind spending on a data center.

This is why analysts love Symposium – we get challenged, schooled and surprised. It’s easy to fall into well-travelled intellectual ruts. We get at least as much out of this as you do.



Comments are closed