DataWeek 2013 Conference & Festival

In this talk, I will discuss how we use data at Etsy to inform product development, from prioritization to experimentation to launch communication, and everything in between. I will discuss the challenges of planning with continuous experimentation and deployment, as well as the evolution of Etsy's data culture. I'll answer questions like: How did we get everyone to care about experiments? How do we combat the instincts to make the data "look good"? How do we manage running dozens of experiments simultaneously? 

Etsy is the world’s online marketplace for handmade and vintage goods. Etsy's mission is to enable people anywhere to easily build and directly support independent, creative businesses. At Etsy, we push code over 30 times a day and we have terabytes of data generated by the tens of millions of members of our community.

Changing user behavior is either the metric of achievement for a technology startup - or a sure sign of early failure. Will users ever fundamentally change how they live or work in order to properly use your product? If the goal of your startup is to acquire new users with a bare minimum of change to their work / life behavior - can you really solve interesting problems? "Changing user behavior" may take on different meanings in different contexts - such as influencing behavior through UX design or gamification - or changing behavior with financial incentives. But for most of us tech entrepreneurs- can we honestly aim to change user behavior?

Liberating and integrating data siloes within an enterprise is proving just as vital to the big data revolution as publishing data to the public. This panel explores topics in viewing the enterprise as a "data hub" that requires data standardization, quality control, integration, and analysis across the entire enterprise of data. Is the advent of "big data" and the exponential growth of available data making cross-enterprise data integration more difficult? Normalization, standardization, cleaning and quality control, data scientist work-hours: what is the biggest bottleneck in the vision of the enterprise as an integrated data hub?

Crowdsourcing and outsourcing have become increasingly popular solutions for engineers and data scientists to aggregate, augment, and preprocess large datasets. Crowds can label, categorize, and enhance all kinds of data, from records to photos, to help data scientists get better quality results. Choosing the right crowd or workforce is critical, and mistakes in the selection process can lead to disappointing results and cost overruns. Bypass the marketing jargon and learn what workforce to use, for what purpose, for how much and why.

Xignite offers financial market data solutions on the cloud to make it easy and inexpensive to integrate real-time market data into your application, eliminating the heavy investments in infrastructure, significant software engineering efforts, and on-going operational costs of going with a traditional data provider. In an industry where latency matters, how can a cloud solution deliver real-time market data? How do you scale a cloud solution to handle millions of requests per day? How do you implement redundancy to mitigate hardware, network, and even data-related issues? What architectural trade-offs do you need to balance?

Data is important but increasingly misunderstood. Over the past few years we've seen the cult of data move out of the dark sub-basements of IT, the nausea inducing confusion of ad-tech or the white starched cubicles of finance into the more airy suites of marketing, sales and human resources. This shift has been fast, sudden and unexpected to most. Like the shift from early scientific exploration through alchemy to our modern scientific method the actual act of data interpretation has been clouded with a mix of absolute faith by its practitioners and fear from those it threatens.

The problem is that the current masters of the dark arts frequently provide guidance that may have a core of truth, but can be applied in ways that will, at a minimum, make your hair fall out and, in the worst case, guide you down a disastrous path. We've seen the birth of data driven cults.

But data does open tremendous new opportunities for business to create new value and streamline existing operations. To do so, however, we need to migrate from looking at data to actually understanding the information produced. We need to stop trying to turn lead into gold, and start trying to discover oxygen.

How do we get there? Data without meaning, context, or evaluation is just, well data - not information. Data must first exist, have structure, depth and most importantly meet objectives and KPIs to drive value into an organization. We need to understand the limits of our current systems. It's tempting to always look for patterns and causality, but sometimes they simply don't exist, no matter how much we will it. Data is not pure. The common concept of n=all is a fallacy built on belief that everything can be measured. Understanding these key aspects of running a data-driven business will help you get out of the dark shadows and outside the ritual of big data. We can move away from the data driven cult of information paralyses and create information systems that are pragmatic, scientific and ultimately drivers of creativity and innovation.

Semantic technology has been around for years and was supposed to save us from information overload. So far, it failed. The Semantic Web or Web 3.0 is still Tim Berners-Lee's dream, and good old Web 2.0 keeps drowning us in oceans of content. But while social media is certainly the cause of this deluge of information, it can also be the solution: first, as it provides us with a huge amount of data that we can use to qualify this information through big data technology; second, because it educated and created a need for millions to become human curators. By combining algorithms and humans, we reinvent media while bringing the meaning back to the Web.

There are tons of specific big data solutions in the market, with more being slated all the time. As a business, its hard to keep up! Correlate is the last business intelligence platform you'll ever need. We takes in all of the best of breed solutions, allow you to produce your own and customizes all of it in a way your employees can actually use and understand. 

Bring your own device (BYOD) has been redefining both the SMB and enterprise workplace for a few years as employees, including C-level executives, increasingly work and exchange company data on personal devices, whether approved or not. Now, with bring-your-own-cloud (BYOC), employees are using free personal cloud options as well, like Google Drive and DropBox, to easily store and access company documents and data from anywhere and everywhere. But, what’s the catch? 

Like personal devices, personal cloud application environments are often incompatible, meaning business processes become disjointed, and employee collaboration and efficiency suffer as a result – not to mention the data security risks associated with unsanctioned cloud tools. 

In order to maintain business continuity and recover employee collaboration, it’s essential for IT departments to implement bothBYOD and cloud solutions that not only offer employees consumer-grade simplicity, but also the ability to seamlessly and securely collaborate. This panel session will explore how companies can manage both BYOD and BYOC in a way that allows employees the collaboration, mobility and data accessibility they have come to expect while simultaneously meeting business continuity and secure data storage requirements. 

Attendees will learn how to strike a balance between employees and IT, so companies can ensure data is protected, and collaboration is easy. They will hear tips and tricks on how to seamlessly integrate BYOD and BYOC into their IT strategies, which includes incorporating mobile device management, conducting security audits, integrating with a central management tool and providing simple solutions. Through the implementation of employee-friendly BYOD and BYOC strategies, attendees will learn how to secure a competitive advantage over those organizations less prepared for the cloud and personal devicedriven workplace.

API’s are powerful tools to help businesses transform.  So how can you utilize them to leverage company, contacts, financial data and analytics, social, vertical-specific data, industry research and more to drive better business results?  How can you combine these information sources to deliver real-time intelligence into any application, such as ERP and CRM systems, corporate intranets, webforms, mobile devices and more?  Join this workshop to hear best practices for utilizing API’s to develop new and innovative use cases.

As the buzz of Big Data transitions into the realities of costs, efficiencies, risks, and planning, security becomes one of the key components of the 2013 big data discussion. What new topics in data security need to be addressed as companies store, analyze, and move exponentially increasing amounts of data? What are the emerging concerns that will shape the next decade of data security solutions?