This is a story about transformation. About radically changing the way an organisation thinks about data and self-service. About embarking on a journey to leverage data to achieve business goals – in this case, providing a better customer experience and managing expenses.
It’s a story about Citigroup, one of the world’s leading financial services company, providing consumers, corporations, governments and institutions with financial products and services around the world.
Before their transformation journey began, Citigroup struggled to understand what fundamentally drives expenses, questioning whether they should even be in the business of certain products without truly knowing and understanding the driving forces behind the cost to serve. Their journey began in the employee help desk, with a view to provide the tools desperately need to run more efficient operations, provide decision support, and get closer to the voice of the customer.
“Help desks are tricky to optimize, and tricky to lean.” explains Kenski. The main tool is human capital. At the core they are driven by people, which means driving change fundamentally impacts the human condition. There is an ongoing struggle to find the sweet spot between customer experience and expense – and that’s the exact challenge Citigroup decided to tackle head on.
The Citigroup call center is no shrinking violet. With over 300,000 clients across 93 countries, supporting 20 languages and processing more than 100,000 tickets each and every month, agents across eight regional help desk locations work tirelessly to serve and support a growing global workforce.
Culturally, Citigroup employees had a low tolerance for the existing self-help solutions. With a propensity to throw work over the wall to level two support, more than 50% tickets opened were moved to higher cost service providers, even though 60% were not actually experiencing technology-related problems. They very much worked with a “One Stop Shop” mentality, depending heavily on someone else helping them find a solution.
This is what they wanted, needed, to change. Kenski describes it as giving them “something intriguing to go to. We want them to think ‘It’s just like how I get help in my real life.’ We’re not shoving work on you, we’re empowering you.”
And so they turned to IBM to partner with them on this transformation journey.
In 2013, IBM and Citigroup began work to build out an analytical platform, bringing data in from key sources including ResolveIT, ServiceNow and Remedy, transforming it into useable form, augmenting with predictive and advanced insight and presenting key valuable insights to the business to help them make more effective decisions and drive change.
Beth Rudden, an emerging technical leader at IBM and all-round analytical rock star, spearheaded the technical solution for the Citigroup call center. She humorously compares the solution to a turbine engine – “you suck things in, you compress, you ignite, then you blow it out.” Most organisations are getting pretty good at the suck things in, compress, and blow it out phases. But Rudden really wanted to focus in on the ignition point – what’s going to ignite the analytical engine that fundamentally transforms data to insight? In her view, it was the injection of SPSS predictive and advanced analytics that took Citigroup’s data to the next level and ignited change within the business.
But that wasn’t the only critical piece. She also stresses the importance of ensuring insights are visualized in a way that business users can consume and understand. Given that 90% of information comes into our brain through the visual vortex, as humans, we are created to find and recognize patterns. Close your eyes and imagine a million records in a database – there’s no understanding in that. But imagine it on a map, and you can instantly recognize that one particular country has 20 tickets open and needs more focus. Within the analytic platform, they measured the past in order to predict the future, and visualized it using a range of visualization techniques including geo-spatial, word clouds, correlation and causation charts. “It’s important to put it all together and tell stories about the data“, Rudden points out, “to make people aware of what this means to them. Give them the context in which to make the decision.”
The inclusion of both structured and unstructured call data is also critical in understanding not just how the call center works, but how business works, and what the customer really thinks.
Barrett Touhy and Tom Lydon pulled together a team of specialists from across IBM to design and build an underlying analytics platform that could ingest both structured and unstructured data, and deliver on the business objectives. From a technology perspective, Rudden breaks the analytics platform down into four fundamental parts:
- The hard part: This is the “suck and muck”. Combine, cleanse and match data from multiple structured and unstructured data sources. In Citigroup’s case, DB2 and Softlayer were leveraged to create a system that could easily and quickly scale out as needed.
- The really hard part: Apply business rules to make sense of the data and add context, powered by SPSS predictive and advanced analytics, and InfoSphere data transformation tools.
- The really really hard part: Re-frame the data as facts and dimensions in terms that make sense to the business so it can be more easily understood, tapping into the performance and transformation power of DB2 BLU acceleration.
- The sexy part: Visualize the abstractions to make sense for the customer, using Cognos, Watson Analytics and Tableau. Give the end use the freedom to choose what works from them, and empower them to self-serve.
The Citigroup analytic platform currently ingests 292 million ServiceNow records, 310 million ResolveIT records, and consumes and processes data generated by over 120,000 real time calls daily.
When a call was answered, they didn’t just solve the problem. They took the data and reapplied it to change the underlying root causes. They monitored everything over time – not just call completion times and categories, but also tapped into the unstructured data in the call records in order to aggregate all data, understand patterns and trends, decide a course of action, implement change and monitor over time.
And the results of the transformation are impressive.
Over time, fewer calls were coming into the call center because they were fixing root causes.
Chuck Limoges, Senior Vice President of Infrastructure at Citigroup, stated that in the past nine months, they have reduced the time a ticket is open by a staggering 50%! Partly because half of all incidents providing educational or navigational assistance are now routed to employee self-service. Self-service that had been injected a new lease of life with fresh new content, IT expos, marketing brochures and campus educational campaigns. In parallel with the data transformation, a big focus was put on changing the culture so self-service becomes the norm.
A positive side effect from the transformational journey was that 1,800 primary applications were identified and removed as part of the initial data quality clean up – providing added benefits to expense management and IT complexity.
This particular phase of Citigroup’s journey, started with the business, and ended with transformation. Kenski estimates they have saved millions of dollars in 2015 alone, with measurable reductions in duplicate and withdrawn tickets, and realized benefit in level two task elimination and shift left. With a drastically remodeled, re-branded self-help experience, Citigroup achieved both ticket reduction and a better client experience.
Of importance here is the fact that Citigroup never made it about metrics to measure productivity. What they were able to achieve wasn’t just insight, but an empowered workforce and optimized price per cost to serve that fundamentally changes the way the business operates moving forward.
When asked what advice she would give to other customers looking to embark on a data transformation journey, Rudden advises:
- Day-to-day operations start with data. It starts with business, and it ends with insight. Make it as accessible as possible to as many people as possible – democratization of analytics is key. Shift left!
- Long term, remember that big change happens with many small instrumental changes. Small steps can have a bigger impact than people realize.
- Trust is the right type of partnership. Rudden attributes their success to the strength of the partnership between IBM and Citigroup, because the project team “were able to understand the business – then render all measurements against business goals so customers have context as to what they are seeing.” Citigroup trusted in IBM to meet the business needs without prescribing how to solve the problem, and that was key to success.
Limoges attributes their results to building for long term success. Focus on creating the right customer experience, achieve target cost outcomes, and reshape the workplace to create the right culture and engagement. “We’ve made it our business, to use data to run our business. Our story could be your story.“