Cisco Blogs
Share
tweet

Analytics and the Universal Pursuit of Insight

- June 15, 2017 - 27 Comments

What makes space so amazing? Every time human kind takes a deeper look, we find something that we didn’t know was there before. The same concept applies to analytics. Gathering deeper insight into any data set through automation allows you to make better and quicker decisions, expose gaps, and identify areas for cost reduction or investment.

However, before you buy the next seat on Virgin Galactic, let’s explore how you can maximize and quantify value from your investments in analytics. The first step is a detailed understanding of your organization’s business plan. For example, if your analytics strategy is focused on your infrastructure, you should ask: “How does my infrastructure provide value to the top priorities in my organization’s business plan?”

Based on these insights, the next step is to identify and establish key performance indicators (KPIs) to measure results and action items for continuous improvement. Your results will depend on the ability to track KPIs, determine the amount of data available, and identify and/or correlate trends across multiple data sources.

On this journey to the outer space of analytics, Cisco Advanced Services has built the experience and framework to help enterprises and services providers navigate and maximize the value from their data.

The value of analytics comes in multiple shapes. In addition to using an analytics approach to drive insights, another meaningful and often unrecognized side effect is internal efficiency. Tasks that used to be done manually in days can now be automated with information instantly available. This benefit reinforces fishing for value in a digital world across the entire lifecycle of IT services.

Many organizations see the analytics engine as the biggest hurdle of implementing the analytics strategy. And, initially, it will be. However, once selected, there are multiple components that require unique skills and efforts beyond simply coding the use case. To gain a better understanding and the ability to identify and engage the right types of resources, Cisco Advanced Services created an analytics reference model.

During the implementation of our analytics model, we looked at many different alternatives. We picked a solution where the learning curve to code the use cases into the analytics engine was the shortest.

What happened? We were presented with a pleasant surprise! Our consultants embraced the opportunity to get trained on analytics. They started to use the platform and explored new use cases based on their interactions with customers. Typically, we would have had a team of 10 developers. Now, we have over 1,000 consultants who are all enabled to build and contribute in a true crowdsourcing model.

Data acquisition is the most important area to have a dedicated team. Data availability through easy-to-consume APIs is the key for a model like this to work—effectively and efficiently.

With six months invested in the program, we have data from 2.7 million infrastructure devices (routers and switches) in customer networks around the world. These are correlated with a number of our internal data sources, like reactive cases, ‘Mean Time Between Failure’ calculations, and our software defects tracker.

Key takeaways for you to consider:

  1. Build value from analytics (number of sources, ability to correlate, number of use cases, etc)
  2. Drive external insights AND internal efficiency
  3. Establish an analytics reference model to determine the analytics engine
  4. Create an engine that addresses your organization’s objectives and needs—today and future
  5. Shoot for the moon

Want to hear more? Join us at Cisco Live in Las Vegas from June 25th to 29th and attend my Critical Insights: Industry-leading Analytics Platform session to learn more about Cisco Advanced Services’ latest analytics platform. Don’t forget to check out Matt Lewis and Joel Prothro explain how to prepare network operations for the digital future. See you there!

Tags:
Leave a comment

We'd love to hear from you! To earn points and badges for participating in the conversation, join Cisco Social Rewards. Your comment(s) will appear instantly on the live site. Spam, promotional and derogatory comments will be removed.

27 Comments

  1. Most people/organizations undertake analytics programs because they are looking for better answers. Soon they realize that the best analytics don't provide better answers, but better questions. Just like the space analogy, each time you gain an insight with analytics, it usually leads to new avenues for investigation. What are some the new and better questions our consultants have uncovered Ulf?

      Spot on Kevin. Many questions coming from the analytics capability. One of the more interesting ones is around crashes. With all this data and capability, and the challenge that most customers have very high performing stable network, and some encounter challenges with crashes, what is the pattern or the trigger to have an infrastructure with crashes, is it the hardware combination, the fact that some customers run old software, the combination of specific features or even how the device being located in the network. Any key questions you see when you look at the capabilities that analytics provide for a complex infrastructure?

  2. To me the most exciting aspect of this development is that with this framework, we are empowering hundreds of smart innovative Cisco engineers to solve customer problems and develop new ways to report the KPIs that are important to a specific customers -- by developing custom apps and custom reporting solutions.

      Paul, we already see how the work of our consultants starts to change. Things that used to take them days, sometimes weeks of work to manually analyze a large amount of data can now be done through analytics by coding an app in <1h. This is an app that automatically gets updated as soon as there is new data, and that can be re-used by other consultants. Most importantly, this allows our consultants to spend even more time with our clients to provide more value as the manual tasks they used to do back office can be eliminated.

  3. Capturing the data is step 1, yet then there is a massive quantity to sort through and analyze to try to extract the value. Having the tools and expertise to process and analyze the data, find trends, issues, improvements, etc., is next. But then expanding that capability to 1000 SMEs and beyond is truly an explosive way to rapidly accelerate finding these benefits "hidden in plain sight" within all that data. Very exciting to think about where this can and will lead...

  4. Great blog! Analytics promising value for Enterprise and Service Providers (SP) business alike is a fantastic development. For SPs with large scale and complex networks, analytics may be one of the biggest business strategies game changer. It is also exciting that Cisco Advanced Services has focused on analytics to help its clients navigate and maximize the value from their data. For many of us who are familiar with the SP business, we know that IT, network engineering, and operations investments are some of SPs biggest expenditures. Analytics may help SPs with targeted and timely investment decisions for infrastructure and support systems; competitive business policies and rules such as proactively notifying customers of network performance issue and bandwidth limitation, etc. The sky’s the limit when it comes to data analytics.

    The pursuit of new objetives, it´s the way to improve our world.

  5. Great Blog - Love that model. You make a key point here about "...explored new use cases based on their interactions with customers". You can place various Cisco and Customer skillsets in Data Platform, ETL, Data Science, Domain Expertise, and Customer Business Understanding along the model icons, and it is the collaboration of this combined Customer-Cisco team that we use to uncover insights with the highest impact.

      Great comments John, any thoughts on how advanced AI fits onto this model?

    • IMO, AI can be defined as using machine learning to support automated decision making for maximizing the benefit of the network to the business. Today we use the data and machine learning to uncover insights and to provide recommendations. We can also fully automate many of the remediation tasks resulting from these recommendations, such as keeping configurations compliant..(Anyone heard of intent base networking?) So to answer your question about the model - when we can go from analytics engine to upstream action by skipping (or just notifying) the User Interface - we are providing AI.

        Totally agree, the question is if the industry is ready for this yet? Technically the capability exists, most customers I talk to have a vision around this, but are not really prepared yet to have an engine automatically configure their infrastructure, they are afraid of skynet to become self-aware...

  6. Great article and I would add that as important as creating a powerful, scalable, open analytics platform is to define the right service model for customers to consume the outputs of that platform. The goal should not be to the platform itself but the outcomes customer can achieve by using it.

      Most people agree that the IT infrastructure is foundational for their organization's success, at the same time they struggle to measure the impact and contribution to the bottom line. Building out a well thought through analytics platform using many different data sources not only allow you to better understand your infrastructure but really serve as the first step towards quantifying that value.

      Spot on Daniel, this is not about building a tool, this is about the methodology of tomorrow's consulting services, providing advice based on informed insights coming from advanced analytics across a large number of components, not just within one customer's environment.

  7. Timely article...there is a lot of value locked away within this 2.7 million infrastructure devices in customers' network. If a customer can process and analyze all this data and gain insight, it can make a huge difference to their operation and they can really transform their business. Clearly, Cisco Advanced Services with our unique intellectual property in our platform, knowledge, and data expertise can help customers fully realize the value of their data.

      2.7m devices is just the start, and in addition to this we build value by correlating data from multiple data sources, providing a true ability to move from yesterday's "proactive" service to today's "predictive" and tomorrow's "preemptive"!

  8. Timely article...there is a lot of value locked away within this 2.7 million infrastructure devices in customers' network. If a customer can process and analyze all this data and gain insight, it can make a huge difference to their operation and they can really transform their business. Clearly, Cisco Advanced Services with our unique intellectual property in our platform, knowledge, and data expertise can help customers fully realize the value of their data.

  9. A very relevant post Ulf, especially in these times when customers struggle with data overload and the real message is often lost. We have looked at networks and logs for years and it's great to now put that expertise into our products and tools as well. Analytics is the way forward, great post!

      Lovya, you point out a real problem. In today's infrastructure, moving towards telemetry and increasing the amount of data it becomes very challenging for an operations team to stay in control of their environment. Analytics, on the other hand, gets better the more data you have access to when you start to look for patterns, anomalies, outliers and in addition to this correlate multiple data sources.

  10. The pursuit of insights and clarity with analytics, will provision the ability to tie Business Objectives to Success Measures and without equivocation, prove worth. We have seen in the past that proving a value message is challenging. By embracing this analytical pursuit, will strengthen the ability to define, state and validate successes and iterate a fair return of investment, in a clear and demonstrative methodology. This is the "New Frontier".

      spot on Patrick and I would like to add: you ain't seen nothing yet, this is just the beginning!

  11. Defining appropriate "Use Case"using analytics reference model is very critical to ensure customers achieves business-intent by adopting analytics service and solutions. Most importantly, analytics help customers to measure the outcomes with speed and accuracy.

      Most it departments trying to find ways to measure the value they provide back to the business. Understanding the company business model, defining KPI's (Key performance indicators) and using analytics to provide insights is a first step to quantifying the value of the infrastructure back to the business.

  12. It is difficult/time-consuming to understand data as-is without supplemental information or expertise, so it is exciting that Cisco Advanced Services now has the ability to aid enterprise and service providers in navigating through their data, both through an established framework AND SMEs. And this is just the beginning!

      Combining the three unicorns the trick, leading data scientists, industry SMEs, lots of data... On top of this, empower the community that uses the analytic applications to clone, modify or build their own as part of their day to day interactions with customers drives continuous innovation!

  13. The space analogy is good. I think the co-dev aspects of successful advanced analytics is often overlooked, so it's great to read about the idea of 1000 developers instead of some people locked in a room asking "how's this?". The role of the SMEs in analytics is critical to real breakthroughs.

      With space travel through the last 50 years, there has been a lot of trial and error, with analytics we do the same, where Artificial Intelligence adding the self-learning part. We are only touching the beginning of using analytics

Share
tweet