SD-WAN: Beyond Cost Savings, Productivity, and Security
Having considered some of the business advantages of software-defined wide area networks (SD-WAN) over traditional hardware-based wide area networks (WAN) as well as several use cases, we now turn our attention to the question: what else can SD-WANs do for customers after deploying the technology?
Beyond what has been explained, SD-WAN is able to monitor and collect data from all the network appliances within the network. A traditional network can’t do this easily as the intelligence of the network is hardware-based, and the data collected resides individually in each appliance and remains in silos. In such networks, administrators would need to probe the network or run complicated command-line scripts to gather information.
Here is where SD-WAN gives users greater network visibility and intelligence, which allow users to view the entire network's metrics. So it isn't surprising that one of the most important SD-WAN features for network administrators is the ability to have full visibility of the kind of applications as well as the capability to control network parameters on the network. These criteria could include enhancing security, optimising resource allocation, boosting application and bandwidth performance, and even deriving location-based and mobility information.
Below are three examples of how an SD-WAN provides the necessary tools to analyse the network using Cisco's Meraki products.
- Traffic shaping
When a branch experiences slow network performance, the administrator can easily log in via a web browser and within a few clicks discover any user using an unauthorised application that is sapping the bandwidth. An administrator is able to detect the user ID, where exactly the offender is located within the branch, the kind of device that is being used, and what kind of applications are monopolising the bandwidth. A traffic shaping policy can be applied to throttle down the user's bandwidth usage and to limit or cut off the user or device from using the network. By using traffic shaping, rogue applications or users will not waste network resources indiscriminately. Traffic shaping is also useful for any enterprise that wants to prioritise critical applications by directing these to the links that have the most bandwidth so that these applications will not suffer performance degradation.
- Optimising network parameters
With SD-WAN built-in analytical tools, network administrators can easily detect and gain very granular views as to what network problems are. These include traffic congestion, throughput rates, packet losses, application anomalies, jitter (inconsistent data stream arriving at the destination), and latency (data delays) problems.
Aside from network parameters, users can measure at the application level and look into transactions and response times, server response times, and mean opinion scores, to name a few. Along with this is the ability to identify security threats such as rogue traffic, suspicious connections, distributed-denial-of-service (DDoS) attacks, and data exfiltration (unauthorised copying or retrieval of data).
- Location analytics
Perhaps the most useful analytics SD-WAN provides relates to location and spatial information. This feature is being used by brick-and-mortar enterprises, especially in the retail and the hospitality sector. These companies have begun experimenting with tracking foot traffic patterns, customer visits, on-site dwell time, and repeat visit rates just like how online stores are able to. Some examples of enterprises already using these features are the International Hotel Group, Prada and Ladbrokes.
So how does this happen?
Knowing your customer
All Cisco Meraki Wi-Fi access points are equipped with a Wi-Fi scanning radio which is able to detect mobile devices searching for hotspots to connect to, regardless of whether or not those devices are actually connected to the broadcasting access point. By collecting this data of mobiles within a confined area, the system is able to classify users based on three profiles: passersby, casual visitors, and connected visitors.
So a mobile that passes a Wi-Fi access point briefly is considered a passerby. If another mobile lingers past a certain time, it can be considered as a casual visitor. And finally, when a mobile stays put for a long time and logs in to the Wi-Fi network, it can be considered a connected visitor. From these classifications, further inferences can be made about the owners of these mobiles such as proximity, engagement, and loyalty.
If, for example, a restaurant in a mall knows there is an increase in foot traffic (passersby and visitors) at a specific time, it can reasonably surmise that there are people lingering in the vicinity of the shop at that said time. Knowing this, a restaurant could offer special deals or open the store earlier to those around the area through an advertisement on the window in order to capture this otherwise unreached foot traffic.
For those who have chosen to connect to the free Wi-Fi (connected visitors), data can be broken down further to investigate who is actually patronising the restaurant (engagement). Using Meraki's analytics tools, the restaurant would be able to determine if they are first-time, occasional, or returning visitors, as well as how long returning visitors engage every time they do so (dwell time). Combining this with point-of-sale, inventory, and customer behaviour data, the restaurant would be able to reasonably determine a composite profile of that connected visitor. Such actionable insights could help understand customers' buying patterns and determine who are the restaurant's loyal customers so that they can be rewarded through other means.
For example, if a customer comes in every day of the working week at 8 a.m., buys a cup of coffee and stays for 15 minutes, the restaurant could offer a special breakfast set (coffee and food) to that customer for a small additional sum. If this same customer accepts the offer three times in that week, the restaurant could offer a 50% discount for breakfast during the weekend. Such insights would be a boon for any enterprise, especially those in the retail and hospitality sectors because it opens up new revenue and loyalty possibilities for them.
Creating location maps
Location data from these smart Wi-Fi access points could also be correlated to indoor location maps. For instance, the same proximity, engagement, and loyalty data collected by the analytics tool could help a mall map out where its most popular hotspots are, where shoppers are likely to gather throughout the day. Such data could be used to guide retailers as to where to place advertisements that would be eye-catching to those passersby. In another scenario, a retailer, such as a supermarket, can optimise store space and place special offers and promotional items at the hotspots instead of having to guess the best spots to do so.
This functionality can be even more granular. By pairing the Wi-Fi access points to Cisco's Meraki smart cameras, users will be able to gain room-level visual analytics. Using high-end cameras with mobile-grade processors, users can even do person detection and people counting, where the cameras are able to classify people in recorded video anonymously.
Beyond the network
Although SD-WAN brings many benefits to the network itself as we have seen in previous chapters, the true power of such a software-defined networking approach is in the data collected from the network itself. Network analytics is about technical data collection, but today's systems are able to go beyond that. At the end of the day, modern enterprises demand much more actionable insight from their investment and SD-WAN is the answer to what these enterprises are looking for.