How Splunk could make the vision of a seamless user experience in the IoT world come true??
Splunk today helps organizations to streamline their operations and to provide a better end user experience via machine data management. It could be the enabler of the Cognitive Devices-as-a-Service future by its intelligent capture, indexing and analytics of machine data.
In a decade from now, in a world of connected devices, controlled by a variety of human inputs (voice / touch / gesture), Splunk could be the brain behind these connected machines providing an efficient, synchronous and seamless experience to the end user.
One of the things, Splunk owes its immense potential is to its singular focus on its mission to make machine data accessible, usable and valuable to everyone. Let us deep dive into how does this amazing company create value, how does it capture value, it’s amazing growth story and most importantly what does the future hold in store for Splunk.
How does Splunk create value?
Before that let us first understand what is machine data and what makes its interesting to handle. For any interaction that occurs in running a business, there is data generated by the website / app / servers / sensors / devices involved. This “machine data” is inherently unstructured, huge in size and the format is unsuitable for analyzing or visualization. [1] And not to mention, the data format varies based on the type of machine generating the data. Splunk also provides use case specific machine learning capabilities for better predictive analytics, anomaly detection and business optimization. [2]
Let us look at customer examples from three different industries-
Myriad Genetics: It uses Splunk to get results for its patients faster. The content and the format of the data (the primary data here is biological) changes constantly. Myriad sees Splunk as a platform for operational intelligence. For them, being able to constantly react and readjust to the questions for drawing new conclusions without going back to the start of the development pipeline is very helpful (this is achieved since Splunk has the capability to modify the schema at search time). [3]
Intuit: It uses Splunk to collect, monitor and visualize the data that it generates from a variety of financial management solutions. By enabling Intuit’s employees to gain the right insights across the entirety of business, Splunk helps Intuit to get another level of understanding about its customers, thus helping Intuit serve customers better. [4]
City of Los Angeles: It uses Splunk to consolidate and analyze data from a variety of sources (firewall logs, threat intelligence feeds, switches and routers etc.) Using Splunk, the city has brought the activities of 40 different agencies under a single umbrella, helping it gain visibility into suspicious activities, assess citywide risks and prioritize and mitigate threats. Apart from being able to get real time threat intelligence the city has reduced its operational costs and has created a cohesive and singular city level cyber security strategy thus better protecting its digital assets. [5]
As we see Splunk is being used by organizations in very different industries to analyze data (varying in format, total size and rate of generation) from disparate sources to gain operational intelligence about their operations, make them more secure and provide better service. Splunk also has apps and add-ons which can run on top of the Splunk platform for various use cases – these have been created by Splunk or members of the Splunk community. [6]
How does Splunk capture value?
The core products of Splunk are offered in two ways: an on-premise enterprise offering and a cloud offering and the pricing is based on the amount (Gb) of data you index per day and the rate is per Gb indexed. The pricing for the Splunk Premium Solutions (for security, service intelligence and user behavior analytics) is decided on a case by case basis by interacting with the Splunk team.
Splunk Growth Story
Splunk, founded in 2003, today provides operational intelligence to 13,000 customers in over 110 countries enabling their digital transformation. Its revenue has grown by over 3 times in the last 3 years demonstrating its value in solving an acute pain point for organizations. [7]
What does the future hold in store?
The world is moving towards more powerful and connected devices in a variety of places. From our homes, to offices in high rise buildings, factory floors to vehicles on roads and ships in sea, the devices will become intelligent and much more integrated with each other.
Two key points to note for understanding the potential for Splunk in this world:
- Significant computing will happen at the edge (made possible by various technological advancements) at an individual device level and between devices, to bring about a great experience for the end user and we will see a shift from cloud to edge computing. [8] [9] For better understanding please watch this amazing talk by Peter Levine (a seasoned tech professional and currently a partner at a16z) goo.gl/QU7aPT. Edge computing would be necessitated by the real time criticality of the operations and the goal of optimum utilization of network bandwidth.
- Splunk today can access a variety of data types from disparate sources to help its users solve a myriad of problems. [10] A pictorial depiction of the same can be found here – goo.gl/uQmbcA.
If we combine the two points, Splunk, by analyzing the data being received by one device from another device and providing intelligent responses, can enable a world of Cognitive Devices-as-a-Service (CDaaS) (intelligent and coordinated experience from one’s devices as per one’s need).
A small illustrative example of how would this world operate: suppose 10 years from now just as your self-driving car is nearing a traffic signal, your car’s brakes fail and the car cannot stop abruptly to avoid serious injuries to you by shutting down the engine. The Splunk instance in your car recognizes that (based on data about your body and the current speed of the car) and via close area network the Splunk instance on the traffic signal also recognizes that your car cannot stop. This instance accordingly directs the lights so that your car can pass through and then shut down more gradually.
In this Cognitive Devices-as-a-Service world, devices would communicate within themselves to ensure that a human being gets an orchestrated, thoughtful, secure and delightful experience as depicted in the picture.
Source of images: [11] to [20]
References:
[1] https://www.edureka.co/blog/what-is-splunk/
[2] https://www.splunk.com/en_us/resources/machine-learning.html
[3] https://www.splunk.com/en_us/resources/video.c1bGw5ODE6rlFH2qWdzw6ewgrZ5UL5qQ.html
[5] https://www.splunk.com/en_us/customers/success-stories/city-of-los-angeles.html
[6] https://splunkbase.splunk.com/
[7] http://csimarket.com/stocks/single_growth_rates.php?code=SPLK&rev
[8] http://www.businessinsider.com/edge-computing-is-the-next-multi-billion-tech-market-2016-12
[9] https://softwareengineeringdaily.com/2017/02/03/the-end-of-cloud-computing-with-peter-levine/
[10] https://www.splunk.com/en_us/resources/machine-data.html
[11] https://thecliparts.com/wp-content/uploads/2016/04/free-3d-clipart-1.jpg
[12] http://cdn.bgr.com/2012/11/mobile-icon.jpg?quality=98&strip=all
[13] https://cdn.pixabay.com/photo/2013/07/12/18/58/laptop-154091_960_720.png
[14] http://www.radio.net/inc/v2/images/landing/hero/devices/single/ipad.png
[15] https://www.technobuffalo.com/wp-content/uploads/2017/03/Fitbit-Alta-HR_Front_Black_TimeForBed.jpg
[16] http://kingofwallpapers.com/vehicle/vehicle-005.jpg
[17] http://wavelengths.freewave.com/wp-content/uploads/2016/11/Traffic-IoT-Blog-image.jpg
[19] https://i.ytimg.com/vi/NhlS6WmN7T4/hqdefault.jpg
Hi Sidharth. Thanks for the great post. I cannot help comparing Splunk to our GE case. I have 2 questions:
1. How does Splunk collect all this data? GE owned the machines and, hence, it was easy for them to integrate sensors in the machines to get the data. Does Splunk have partnerships with various machine manufacturers like server manufacturers like IBM? Or is it able to ingest a data dump from any machine and then organize it for easier visualization?
2. In the GE case, GE was trying to capture value by promising to clients that their insights can lead to cost savings and then taking a cut of these cost savings. Splunk has a different value capture model in which they charge for the data that is uploaded to their platform by its size. Do you think that if they move to a GE type value sharing and capture model, it will help them to scale faster?
Hi Bipul,
Regarding the first question – Splunk can ingest data from several sources and then organize it for various use cases.
For the current use case of machine data ingestion and analytics I think Splunk’s value is proven in the market and hence the current model would suffice. However, once Splunk’s instances start getting deployed on the edge a more customized pricing solution directly tied to benefit (which could be cost saving, reduced time of decision making or better experience for the end user – which could be quantified in several ways) might be needed. In that Splunk would have to solve for how to keep the pricing model simple yet accurate.