Sprout Social | 2023
B2B SaaS
Product Design
Systems Thinking
Project at a glance
ROLE
Product Designer
TEAM
Product, Eng, Data Science
TIMELINE
6 months
Overview
I led the design of Network Benchmarking, giving social media managers a way to compare their performance against broader social network trends without having to leave Sprout to find answers. The project was handed off to me with a few months of prior exploration already behind it, but no clear direction. I had to reset the approach before any design work could begin.
Through the work, I collaborated with data science, introduced a brand new metric to Sprout’s reporting offering, defined project scope, designed the benchmarking experience end to end, and laid the groundwork for expanding to new networks and metrics.
Impact
Grew from 2,500 monthly report views shortly after launch to over 4,500 sustained monthly views two years later
Drove a 204% increase in usage of the related Engagement Rate widget within custom reports over 18 months
Eliminated support ticket spikes during network anomalies by giving customers a self-serve way to check broader trends
Customer demand to expand beyond Facebook MVP validated the opportunity, leading to Instagram and LinkedIn support shortly after launch
Opportunity
Give social media managers a way to see how they stack up
Data is core to a social media marketer’s job. They’re constantly proving the ROI of their work to stakeholders, coworkers, and leadership. But there was a clear gap in that reporting story: customers had no way to see how their performance compared to broader network trends or industry standards.
The signals were hard to miss. Customer Success was fielding regular requests for industry averages and benchmarking data, with users saying outright that they had to leave Sprout and turn to Google or marketing blogs to find it. And whenever a social network had an anomaly, or a customer’s own performance dipped, support tickets followed: My performance is down. Is this just me, or is something bigger going on?
The goal was to build benchmarking directly into Sprout’s reporting framework, giving customers the answers they were currently looking elsewhere to find.
Problem
Inheriting a started effort and a metric problem nobody had solved
The project had been handed off to me with a few months of prior exploration behind it, but no clear direction. As I sat in on project syncs, reviewed the existing work, and listened to customer research calls, one thing became clear: after months of effort, nobody had narrowed in on an approach. There were more open questions than confident answers.
I couldn’t confidently design anything until we defined the data underneath the feature. So I stepped back and made a plan:
Define the benchmarking methodology
Lock down MVP data scope (network vs. industry benchmarks)
Then design the experience
Underneath the misalignment was an unanswered data question. Sprout’s metrics were all volume-based: impressions, engagements, audience growth. But volume alone can’t support a fair comparison: a small business with a couple thousand followers looks nothing like a brand like McDonald’s when you’re just counting raw numbers. Even standard medians couldn’t fully solve for that scale problem. The real question wasn’t can we show customers a benchmark..it was how do we create a comparison that’s actually meaningful across businesses with wildly different social footprints?
Approach, part 1
Finding the right metric backed by real data
To define the methodology, I needed to see real data plotted out as customers would see it. I partnered with data science to pull actual customer data across the metrics product had been discussing, plus a few I’d been considering myself.
One initial direction was to plot a user’s absolute percentile ranking over time. But once I put real data into the chart, the direction fell apart. Daily fluctuations were so significant the chart offered no stable, meaningful insight. It was visual noise, with no fixed point of comparison for a customer to work toward.
Engagement rate was another promising option. Engagement rate is a percentage-based metric dividing engagements by impressions. This naturally levels the playing field between small and large accounts. There are outliers where low impressions inflate the rate, but the risk was small given average impressions across our customer base.
I believed it was the right foundational metric, especially for an MVP aiming to provide value regardless of account size. I shared my explorations and findings with the team. Seeing real data in context did the work. We aligned quickly, and engagement rate became our foundational metric.
Interlude
Designing the Engagement Rate Widget
With the back end underway, I designed a new Engagement Rate widget in parallel, following our existing reporting patterns and simply defining the new metric. This provided customers immediate value with new ways to understand their performance across social networks while the harder benchmarking work continued behind the scenes.
Approach, part 2
Defining the MVP scope: industry vs network
That harder work was scope: would benchmarking compare customers against a network overall, or their specific industry? Product and customers wanted industry-specific benchmarks, but I wasn’t convinced we had the data to support them meaningfully.
I asked our PM to dig into our customer metadata, and we found a real roadblock: industry data was manually entered by account managers in Salesforce at sign-up, often left blank or under-specified. On top of that, Sprout’s many advertising agency customers—each managing clients across a range of industries—were all globally bucketed as “Agency,” erasing the industries they actually represented.
Once that gap was clear, we narrowed our MVP scope to Network Benchmarking, where our data confidence was solid.
Solution
Designing for how social media managers actually think
With the metric and scope defined, I turned to the experience itself.
First, I looked at how we might layer benchmarking onto the brand new Engagement Rate widget by overlaying the network median directly on a customer’s existing data. But that raised two problems.
Discoverability would suffer, buried far down a page inside an existing report
Placement didn’t match user mental models. Social media managers look at their standard reports for quick, tactical, day-to-day updates. Benchmarking is a highly strategic, periodic gut-check, especially when something feels off
So I proposed a dedicated Benchmarking tab instead. It kept the two use cases distinct, guaranteed high discoverability for the launch, and gave us dedicated space to grow the feature over time. That’s where we landed for the MVP.
For the chart itself, I initially tested plotting six percentile lines, but the visual noise was overwhelming. I scaled it back to three — the 25th, 50th, and 75th percentiles — giving customers a clear sense of low, average, and high performance, and tabled the rest as a potential addition to our premium analytics offering down the line. I defined the customer’s own performance as a bold black line, so it read with immediate contrast against the dashed, colored benchmark lines, and worked with our design systems team to accommodate that deviation from standard chart patterns.
Below the chart, I designed a summary performance table for a quick read on aggregate performance across the reporting period. It looked simple, but it required real care — the copy structure and ordinal numbers needed to hold up across international translations without breaking.
To reduce support burden at launch, I also designed onboarding walkthroughs to introduce the feature, and wrote in-app documentation explaining exactly how our benchmarking calculations worked so customers could self-serve answers instead of filing a ticket.
Result
A self-serve tool that scaled with real customer demand
Once the feature shipped, I mapped out a drill-down framework so customers could click into any day on their chart and see exactly which posts drove a spike or dip in performance — turning a high-level trend into an actionable answer.
You can see it in action in Sprout’s own benchmarking data. On Facebook, Sprout sits around the 6th percentile — which sounds alarming until you know the context: Sprout intentionally doesn’t invest heavily in Facebook, because their target audience isn’t there. On Instagram, they sit stably around the 39th percentile. Zooming out to 3- and 12-month views surfaces macro trends too, like predictable weekend drop-offs — exactly the kind of pattern a customer could use the tool to catch on their own.
The business impact matched the customer signal that inspired the project in the first place. Benchmarking saw consistent, healthy adoption — growing from 2,500 monthly report views shortly after launch to over 4,500 sustained monthly views two years later. The Engagement Rate widget I designed alongside it drove a 204% increase in usage within custom reports over the following 18 months, adding real value for premium analytics customers.
Support spikes during network anomalies disappeared, too — customers now had a self-serve way to check whether a dip was isolated to them or part of a broader trend. And the demand didn’t stop at Facebook: customer requests to expand benchmarking to other networks confirmed we’d found a real, unaddressed need. Instagram and LinkedIn support followed shortly after launch.