My Five Key Takeaways from the Big Data & Analytics Innovation Summit Dublin
This post was originally written by Helena Schwenk
As far as events go, the Big Data & Analytics Innovation Summit held in Dublin recently certainly proved to be one of the better ones I’ve attended. The conference is designed to help business professionals across all industries share best practices and drive success through big data and analytics. Overall there was an excellent selection of presentations, and the presenters all were diverse in terms of industry and topic focus as well as the level of detail they were willing to go into. As you may imagine the range of conversation topics was wide and far reaching; but a number of them did gravitate around some key themes and action points. Whilst these may not be entirely new in terms of thought leadership, they certainly underline the practicalities and real-life experiences today of organizations looking to harness the value of big data and analytics.
I left the event with five key takeaways.
1. Connect the dots: Several of the presentations during the conference focused on the need to eliminate data silos in order to drive value from disparate enterprise data sources. Customer loyalty management company Aimia, for example, spoke about how it was working on behalf of its client Sainsbury’s to bring together Nectar loyalty card data with its transactional point of sale data to improve the customer acquisition process for Sainsbury’s Bank. Similarly, Riot Games (the makers of League of Legends) also demonstrated the benefits of de-duplicating and matching ID’s across multiple databases to give them a greater understanding of complex player behaviors and interactions to help pinpoint areas for game performance and design improvements. Similarly both presentations promoted the idea of tapping into granular or atomic level data to get that richer and deeper view of customer or player behaviors.
2. Be creative with data: Another takeaway from several of the speakers centered on the idea of ‘thinking outside the box’ when it came to utilizing big data. Yahoo! provided an eloquent example of how it was using big data to drive innovations around human-centered computing. The company is using some pretty sophisticated and advanced analytic techniques to provide a deeper insight into online behavior. It had discovered, for example, that people tend to demonstrate different behaviors when web searching compared to image searching (people often browse rather than search). Understanding these discreet nuances and variances in behavior by different user types has been used to help change the design user experience, features, algorithms and metrics used on screen at yahoo.com for example.
On a related theme, similar to Aimia’s presentation, Yahoo! also focused on the need to blend quantitative as well as qualitative data (such as informal interviews, focus groups, surveys and crowdsourcing) in order provide a more holistic and contextual view of behavior. This is something I believe is often lost in big data talk when a lot of the conversation is about tera- and peta-bytes of information.
3. Keep a watching brief on privacy: It seems fairly typical nowadays that an event around Big Data would also dedicate some of its time to the topic of privacy. So no real surprises here except to say that although there was a high degree of awareness amongst the audience and speakers about the need to use data ethically and to protect privacy where it made sense, very few of the delegates appeared to find it was stopping or restricting them from doing “business as usual”. The exception to this came from those representing the online gaming sector who are impacted by the US Children’s Online Privacy Protection Act. Overall though, this seems to suggest that despite its high profile, privacy is not proving to be a barrier in the use of big data. Well, not as yet anyway.
4. In-memory platforms provide analytical leverage: I found it interesting that two of the presenters made mention of the use of in-memory platforms. Although not quite a statistically significant trend (!) it underlined the benefits of speed of thought analysis and interactivity enabled through in-memory technology. In particular, both Aimia and Paddy Power were using in-memory to power operational reporting applications and analysis to get the answers to key business questions that previously hadn’t been possible before. In the case of Paddy Power, the opportunities opened up by in-memory were huge, enabling them to understand the impact of special offers, how to plan free bet special offers and to enable risk traders to hone in on certain bets all within a much shorter time frame.
5. Benefits of sentiment analysis still to be decided: During the panel debate at the end of day one, there was a healthy discussion about the merits of using social media data and sentiment analysis. The consensus seemed to be that sentiment worked well in fairly niche areas such as analyzing editorial articles where there’s an imposed structure. However, the analysis technique certainly wasn’t foolproof when it came to analyzing tweets and social media posts for instance; most people agreed that a human element was still needed to counteract the nuances in language such as irony and sarcasm in order to get a better view on sentiment. On a similar note, one of the more interesting use cases of social media data came from Unilever who are analyzing Chatter conversations (Salesforce’s enterprise social network platform) to harness the ‘wisdom of crowds’ and access employee opinions they hadn’t had before. That insight has proved a valuable tool in helping the company ‘get the word out’ and generate ideas around certain initiatives as well as helping them increase employee engagement without the need to publicize or pay external consultants.
So there you have it. My five key takeaways from the Big Data and Analytics summit. If any of these ring true for you too, I’d love to hear your thoughts. Leave a comment here or drop me a line at firstname.lastname@example.org.
(Note: Opinions expressed in this article and its replies are the opinions of their respective authors and not those of DZone, Inc.)