A Q&A with Tyler Johnson
Tyler Johnson is Founder and CEO of Convergent Technology Advisors, a consulting firm specializing in helping companies make the transition to a digital enterprise. The firm helps enterprises navigate the emerging technology space of cloud, social, mobile and big data integration by helping companies develop a transformation roadmap, evaluate and manage technologies and partners and measure success. Johnson has consulted and worked with global IT firms such as Rackspace, HP, CSC, EMC, NetApp, VMware, Dell, EMC and many others on product strategy, business and alliance development and business/IT alignment.
For progressive CIOs whom are trying to bring together social media, big data analytics, cloud infrastructure and mobile apps into a cohesive strategy, what are the keys to success?
Johnson: Today, most companies are spending between 3% and 5% of their revenues on IT. There is still too much focus on reducing cost and not enough on business transformation. Look at a company like Uber, a ground transportation upstart that owns no taxis. Netflix is the largest provider of video rentals but their business model has evolved away from providing physical videos. Airbnb has the highest number of listings on the planet for rentals but doesn’t own a single property. New business models are transforming many industries and with consumer expectations so high, companies need to focus on customer engagement, content and micro-targeting on any computing platform. It requires integrating all of these pieces – social, cloud, big data analytics and mobile. That’s going to take a lot more investment in infrastructure than what companies have been committing to in the past.
Are there infrastructure best practices that seem to work best for these kinds of environments, or are you seeing several emerging models right now?
Johnson: Technically, this transformation will happen over time. You don’t want technical debt creation, which is the need to rip and replace systems every time you make another change to your environment. One change shouldn’t require changes across all these other complex systems. The next step is to develop your hybrid cloud, which is the foundation for everything. It involves many elements including standardized APIs for monitoring capacity management, automated management tasks, integration with multiple IT vendors, data visualization dashboards and converged infrastructure providers. As recently as three years ago, people were claiming that everything would end up in the public cloud. Yet hybrid cloud makes sense because you don’t have to rebuild everything from the ground up when you still have systems that work fine. Just build the new applications in the cloud and maintain the legacy systems where you can maintain efficiency, agility and availability.
What is the best way to move toward a hybrid cloud model without introducing needless complexity?
Johnson: There are two sides to this. There is what to move and when and how to move it. Change is not just the technology, but you need buy-in from the rest of the company. Target applications with the best ROI to improve the business. Typically those are customer-facing apps or apps that when combined with other data sets like web services, location data and internal data, deliver business value. Next you will look at application architecture and suitability for public cloud compared with dedicated infrastructure. The real key is to think about ROI in business terms not just cost terms. Can we change our processes to create more customer intimacy as opposed to merely running a business process faster or cheaper? The overall theme is that IT needs a deeper grasp of the business.
In hybrid cloud environments, how should IT measure and monitor for both cost and performance needs?
Johnson: IT needs to have a tool that can measure actual consumption of infrastructure across multiple environments whether that is public, private, on premises or off premises. Using tools like StrataCloud for converged infrastructure (CI) monitoring will make it easier to compare apples to oranges. Companies need to move from traditional measurements of Capex revolving around uptime to business outcome metrics. This entails using data integration tools like Flume to integrate cost and performance data in a Hadoop cluster or using MapReduce to tie costs to social media outcomes, like tweets about your product. Imagine correlating changes in your website user interface to Facebook likes. Think about how powerful that would be. This requires a lot of manual backend work to make these connections, but companies need to get there because that’s where the competition is going. Choosing the right tools up front is critical to perform these integrations, because you don’t want to be at a permanent disadvantage in the market.