With new digital streams coming online every day, one of the biggest challenges facing analytics and big data projects is agility. The transition to digital and understanding a customer’s digital footprints requires ingesting data from multiple channels like web, social, email, and applications. It can prove daunting to get analytics projects off the ground. In fact, according to Alan Duncan, Research Director at Gartner, “One of the biggest challenges businesses face when implementing a digital strategy is ‘faster implementation,’” with an associated survey stating 48% of respondents find it to be a key challenge.
One of the keys is to build a team focused on agile analytics. Think of it as a team chartered with rapid delivery of analytics across the organization, with a process, cultural, and architectural best practices to work with business teams, use emerging technologies, and create quick wins with analytics — all while executing in a coordinated, integrated way.
Justin Waite, Senior Director, Analytics & Commerce Cloud at Salesforce, is tasked with just that. With more than a hundred people globally, Waite’s team’s mission is to use emerging technology to help Salesforce employees work smarter with customers, and use those insights to help Salesforce customers embrace the latest innovations with confidence. “It’s about proving out early-stage best practices, seeing what works and what doesn’t, but all the time being customer-centric,” said Waite.
With more than 80 Analytics Cloud projects under the team’s belt in 2016 alone, and hundreds delivered since starting out a few years ago, his team has some great analytics best practices to share.
One example of the kinds of projects that Justin’s team recently delivered is around customer success, aligning Salesforce employee skills with customers’ demand for projects — to maximize utilization and elevate project outcomes. The goal was to be more strategic and proactive around aligning resource management with customer satisfaction.
With employees tagging skills that they’re competent in, Waite’s team layered on Analytics Cloud dashboards to provide summary and detail insights into resources and skill levels across locations. This enabled managers to identify opportunities to improve training from the right people with the right skills and proactively align internal resources with customer needs.
As part of this employee analysis, the team added 360-degree customer analytics, providing visibility into project success, including early warning flags, customer satisfaction scores (CSAT), number of cases open, pipeline, and bookings analytics.
Before Analytics Cloud, the team could see customer scores at the account level. After, managers gained a more strategic perspective, with the ability to compare and rank customers, perform comparative analysis, and compare customers to benchmark stats across regions and geographies ― managers now get early insight into broader indicators, as well as account-level detail.
Projects like these have enabled Waite’s team to formulate seven best practices, which they use with every Analytics Cloud project:
Do not to attempt to analyze every metric. Work with a department to understand what metrics it really needs, and which ones it needs to track business impact. Identify a handful of top-level summary metrics, and some more detailed actionable operational drivers.
Successful delivery of analytics means working with the business owner, who is typically the one who’s been creating reports with spreadsheets and manual extracts. If you work exclusively with IT, you run a big risk of missing the mark on the problem you’re solving. Listen closely to the business owner — and understand who he or she is serving — to get a clear perspective on the data and metrics that matter.
Old-style waterfall methods simply don’t work with modern, fast-moving analytics projects. To avoid gridlock, get an initial cut on the data and metrics, and show it to the business, get feedback, iterate, and check back with the business stakeholder a week later. Rinse and repeat. Prototype and be agile — there is no definition of “done” with analytics.
Analytics isn’t just about finding insights. Tools like Analytics Cloud are incredibly useful for exposing issues with data, perhaps with some drill-down paths that don’t have data, because the underlying business process needs to be addressed. Looking at summary data and then drilling down often helps to identify them. Identify a data steward, slice and dice across varying levels of detail in Analytics Cloud, then go back and fix the issues with the data extraction — or communicate with the LOB teams to give them the opportunity to address the underlying business process.
For global deployments, don’t underestimate cultural differences with dashboards. The metrics and how they are displayed and communicated may work well in one country or business unit, but not with another. Use personalization to tailor them to your audience to maximize adoption.
Organizational security is “who” can see what data based on where they are in the organizational hierarchy. Visibility and flexibility matters. For example, some managers may want to see two levels beneath them and three levels above them. Be sure to get feedback on your organizational security model design to ensure controls and reduce maintenance.
While you can load in the finest grain of data detail, like individual clicks or the most detailed products, it isn’t necessarily useful for analysis. In fact, doing so may open privacy concerns or add unneeded complexity. Align the data levels with only what’s necessary to balance usability, performance, and privacy.
Justin Waite and his team have learned a lot about improving business productivity and process with analytics. The next time your business requests a dashboard or analytics app, try these best practices to create agile data-driven improvements.To learn more about Analytics Cloud, watch the demo.