Hello All !
You remember last week ? We were speaking about ESB and how you can leverage this central component to drive your business in real-time ! Simply by using Splunk Stream to capture ESB traffic on the fly without any modification…
Today, I will focus on a potential use-case for this at the “Splunk Hotel”! SplunkHotel is a company that owns a few hotels but also references other hotels on its booking website. Those independent hotels pay to be on the website. In exchange we guarantee additional revenue.
Splunk already collects the data in Stream and as we saw last week we collect the business payload so we can get inputs regarding the business, including revenues and trends.
Let’s play with the data! One of my revenue channels is the service offered to independent hotels. So I have to make sure that I generate some revenues for them. Below, you will find an interesting KPI that provides a kind of room performance score and the ratio between availability and booking requests. We want every room to be in the top right, like Splunk on the last Gartner Magic Quadrant. That would mean there is high demand and lots of bookings!!
Basically, the x-axis represents the number of times a room appeared in the search results and the y-axis represents the number of times the rooms were booked.
I can see that there are a few rooms with high demand but low bookings. Typically those rooms are probably too expensive or maybe too far down in the search results. On the other side, I can see rooms that do not appear as much in the results and I have to move them further up the ranking so they start appearing on the booking website.
I created a table report listing all the rooms, a few details on each room and an action column. You can take this directly from Splunk and then add two actions: apply a discount and / or “rank up” a room so that the room goes up in the search results. You know what, applying the discount is a Web Service call!
As Splunk is capturing all the ESB exchanges, I will have this information (on the previous actions) in Splunk! I can see analytics on this like how are the bookings for this specific room performing since the discounts were implemented? Wow – that’s awesome isn’t it?
I built a few KPIs that allowed me to deep dive into each discounted / ranked up room to measure the efficiency of the action. Let’s click on room 22 to see the impact.
Since the last discount, I can see I have two new bookings, resulting in additional revenue of 1560 euros. This discount costs me the difference between the real price and the discounted price across the number of nights booked, 520 euros.
Measuring the efficiency in real-time helps you iterate and adapt your strategy in real time. It could be highly effective to do this on rooms that are still available for tomorrow to avoid missed revenue and unoccupied rooms …
As a quick summary:
We collected information from the ESB , analyzed the data and took a series of actions. We then instantly measured the impact of those actions and as a result could adapt our actions to find the best fit.
One of my customers does exactly this on top of his Camel ESB. He works for a global insurer with customers including railway companies that offer ticket insurance on their booking website. When someone is buying a train ticket, if that person asks for insurance, a web-service from the insurance company is called to provide a quote. If the insurance is bought, another web-service is called. Each call is Splunked and then measured in real-time with insight on the revenue, quote number, etc.
Hope this was clear and helpful!