Data-Driven Personas at Salesforce
At Salesforce, we face some special challenges around understanding our users. First, many (if not most) of our users have customized their Salesforce instance: every user sees a different version of our solution. We also have a broad, complex, and disaggregated user base: our users leverage Salesforce in a myriad of industries, regions, and market segments.
So how can we draw meaningful conclusions about our users, at scale? How can we surface insights that are both accurate (based in real data) and actionable (actually useful for lots of teams). How can we better understand your users without spending so much time and effort that our findings are unusable and outdated?
We’ve landed on data-driven personas as an efficient, research-led method for better understanding our users. User personas based on data provide our designers, product managers, doc writers, engineers, and executives with a common shorthand for our users that’s both easy to use and is based in current data about our user base. Here’s how we do it.
Creating Our User Personas
We create our personas based on data from large-scale surveys of our users. These surveys collect information about who our users are, their responsibilities in their jobs, the software tools they use (including Salesforce, of course), and their pain points and incentives. In particular, we focus on learning about our users’ job behaviors and responsibilities that are tool agnostic: for example, we first ask how often our users need to do a particular task in their job, and then we ask them which tool they use to do it.
We then take our survey data and use clustering analysis and factor analysis to reduce a broad range of behavioral variables into clear, mutually-exclusive categories: our user personas. We craft personas such that a given user is based into their group if they perform the tasks of that group at a higher rate than those of any other group. Users can fall into two personas if they perform both sets of tasks at the same rate, but that’s generally uncommon.
Surveys vs. Usage Data
We believe that an understanding of our users’ offline behaviors is key to understanding what they need from their tools, and surveys are the most efficient way to access this data at scale. Often, our survey-based persona analysis reveals that groups perform tasks at rates we didn’t expect, but don’t have the tools to serve those needs. In identifying new needs, we can better plan and build for our real users. Usage data analysis is an essential complement to our personas: once we have the groupings, we can better understand how each group behaves within the tool.
Personas vs. Job Roles
Often, we use job roles as shorthand when talking about sales behaviors. Sales Development Reps build pipeline, for example, or Account Executives close deals. While this shorthand works sometimes, our survey data indicated that job roles are not consistently correlated with behavior across our sales user base. For example, users with the title “Business Development Rep” and “Account Executive”, were found in all five persona groups, indicating that job title is not always a clear proxy for sales behavior. We encourage our stakeholders to look to behavior, rather than job role or title, to best understand a user’s needs.
Distribution & Evangelism
Once we have our persona groupings, we share them widely within our organization. We build out slide decks on each persona set by product, detailing our overall findings, breakouts on the individual personas, and giving deep-dives into interest areas that run across all personas (such as accessibility and mobile usage).
Once collateral is prepared, we distribute our findings throughout the company and focus on helping teams adopt the research. We provide Brown Bag sessions where we present the findings, hold meetings with individual teams to workshop ways for them to use the personas, and solicit feedback from teams on how personas can support their efforts.
We are also integrating our personas into our agile software development process, so that when we write user stories we have background on exactly which users we’re envisioning will use our products. We also shared our personas with our users and customers at Dreamforce, and discussed how personas can be used to enhance the development process for our customers as well.
How Can I Bring Personas to My Organization?
We’re excited to help our customers integrate personas into their work too! A great place to start is our Personas trail, UX Personas for Salesforce. This trail walks users through the personas we have created so far (Sales Cloud, Service Cloud, and Marketing Cloud), and provides guidance on how to use personas in your business. For more information about Service Cloud users specifically, check out a post from my colleague Kathy Baxter, “Moving from Contact Centers to Customer Engagement Centers.” We’re excited to continue building on our persona initiatives as our products evolve and as our user base grows!
Interested in providing feedback to the Salesforce User Research team? Join our Customer Research Program!