COVER STORY
Splunk as the roots, supporting the appealing flower of the user interface, which drives and supports the desire to shop. “Splunk ensures the cheques that are written by our front-end people are cashed by our back-end people, so we need to ensure that when you click a button that says ‘add to basket,’ it does what it says on the tin. By having those buttons and interactions
working at all times, by monitoring where there’s a shortfall and having actionable tasks, it means that we can maintain service.” Splunk has also enabled John Lewis to more
accurately model the journey of the customer through its website, providing every possible permutation and combination of the scenario a customer can go through at checkout, for example. “With Splunk you don’t need architects, you
have the data which shows reality and goes back ad infinitum,” said Cummins. The team quickly realised that customers’ interaction with the John Lewis website was not necessarily as linear as they had previously believed. During Valentine’s Day last year, Splunk
illustrated the differences between the John Lewis model of how customers would interact with the website and the reality of how people actually wanted to use it. “That’s very exciting and potentially alarming,
©John Lewis, Stratford
our war room, our command centre with a huge amount of screens presenting Splunk dashboards, primarily, alongside other dashboards too.” Based on those metrics, John Lewis is in a
position to make immediate, real-time, actionable decisions about marketing activities or technical equipment. “The key here is that the data is ‘actionable’,” continued Cummins. “We can pinpoint lost opportunities and ensure our service level remains constant. We get to maintain the customer experience
and if that’s maintained, people are going to get to the checkout more successfully and have a good experience and that’s going to show in the conversion rates.” Splunk has enabled John Lewis to monitor
drop-offs and payment failures, due to technical issues, fraud detection or incorrect card details. Now, whenever drop-offs are greater than a baseline figure, an alert triggered in Splunk Enterprise allows the John Lewis team to investigate root cause. Equating the situation to a plant, Cummins sees
12 Autumn 2014
but it’s a great opportunity because it means we now know where to focus our attention for success. There’s no point in having the crown jewels in the wrong place, it’s about spreading the crown jewels where people are.” But how does Splunk and log analysis actually
work? Cummins explained that at a very basic level, machine data is being generated everywhere through all sorts of actions. “If someone’s on our website, clicking on a
button, adding to the basket, checking out, all of that generates data. If you’re on a mobile phone it’s sending signals to masts, everything around us is generating machine data.” John Lewis collects its machine data from a
variety of logs including e-commerce applications, web servers, middleware, database logs and so on. The company operates a fully load balanced, multi- site IT architecture across two data centres. Each site uses Splunk forwarders to collect the
data to be indexed by Splunk software. This indexed data is then accessed via a number of Splunk search heads with job servers for scheduled tasks and processes. John Lewis also uses Splunk DBConnect
www.retailtechnology.co.uk
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45 |
Page 46 |
Page 47 |
Page 48 |
Page 49 |
Page 50 |
Page 51 |
Page 52 |
Page 53 |
Page 54 |
Page 55 |
Page 56 |
Page 57 |
Page 58 |
Page 59 |
Page 60