DocuSign internal tooling

I initiated and led the design of internal tools that let DocuSign account executives scan their entire book of business at a glance, and match the right features to each customer using the company’s own usage data.

Year:

2023-24

Role:

Initiated a wide-scale independent research to understand the current landscape of internal dashboards and discrepancies with current Jobs.

Drove hypothesis-driven design explorations with PMs and end users.

Context.

A reorg at DocuSign consolidated data engineering and analysis, previously split across sales, marketing, finance, operations, and product, into a single team responsible for how the company saw its own data. The team built dashboards, and the system that the rest of the company used to build more. My initial role was to design and maintain that system, along with the ecosystem of reporting tools.

That work put me close to how each team consumed the company's data, and I used it to independently propose research on how account executives in the mid-market and SMB segments manage their books day to day. Interviewing them was not routine. Their time was the company's selling time, and unlike external participants, they could not simply be recruited, so beginning the work meant making the case to the director of product management for pulling working reps away from selling. The research produced tools shaped by how those reps actually worked, addressing the behind-the-scenes manual work and the confusion around product discovery.

A SMALL COLLECTION OF CURRENT TOOLS, MANY OF THEM BEING SINGLE-PURPOSE TOOLS.

Problems.

#1 - Account executives in these segments each carry upward of 200 accounts and work them in pairs with a business development rep, using a shared spreadsheet. The specific way each pair narrowed that set down to the accounts worth pursuing lived only in their routine and was captured nowhere.

What counted as worth pursuing was also personal. One rep treated accounts just over their consumption allowance as not worth the effort, while another treated anything past ninety percent as his most valuable.

#1 - AN ACCOUNT EXECUTIVE WITH THEIR B.D.R FINDING OPPORTUNITIES AMONG THEIR ACCOUNTS USING THEIR FRAMEWORK. 

#2 - Features were organized as a flat list, sorted at best by the plan they belonged to. Talking a customer through what might be worth adopting meant a rep holding in their head the connections between features, plans, and that customer’s situation.

Digging through Alation, I found Snowflake tables that, when joined with a way to group the features themselves, could identify which features actually fit a given customer.

#2 - TWO SOURCES AN AE HAS TO SPECULATE ON UPSELL OPPORTUNITIES AND CUSTOMER-FEATURE ALIGNMENT. 

Solutions.

For Problem #1, I designed the Book of Business Analysis. I mapped the way reps narrow their accounts to the tool's structure. Patterns that repeated across the interviews became flows, and those flows were reduced to a set of filter controls, some required and some optional, that let a rep reconstruct their own method of refining the book.

A PROPOSED MODEL FOR HOW AEs SEGMENT THEIR CUSTOMERS.

INTERPRETATIONS OF RESULTING DATA COMPOSITIONS.

The tool plots each account as a single anonymized line across six measures: envelope consumption, active users, feature utilization, add-on consumption, support tickets, and average time to completion. The rep sets their own MRR threshold, and the lines recolor around it.

Nothing is labeled until the rep selects a line, at which point an account reveals which customer it belongs to. An AE works the tool by looking for the line that stands out from the rest, like the low-revenue account that exceeded its consumption allowance within weeks of starting.

INTERPRETATIONS OF RESULTING DATA COMPOSITIONS.

To address gaps in the feature-customer fit (Problem #2), I designed the Feature Usage Baselines to show whether a customer is getting the most out of DocuSign relative to customers like them.

An AE selects a customer, and by default, the customer is matched with other customers based on 8 dimensions that the user can customize. For each feature in that group, the tool shows the customer's usage relative to the range for similar customers, whether their current plan entitles them to it, and the tier that would unlock any features they cannot yet access.

DIMENSIONS CONTRIBUTING TO COMPARISONS.

The set of similar customers is drawn from the account's own attributes, region, segment, industry, sub-industry, department, and use case, which the rep can accept or adjust until the comparison is one they can defend in front of the customer.

Each feature also carries its own view of impact over time, the customer's line against the median and range of similar customers who use it, with the measure matched to the feature: completion time for SMS delivery, transaction volume for collaboration features, envelope volume for sending. Account executives take these comparisons directly into the decks they bring to customers, since customers want to see what their peers are doing.

HOW ACCOUNT EXECUTIVES USE FEATURE USAGE BASELINE TO SELL BETTER OUTCOMES TO CUSTOMERS.

While the feature was initially designed for a customer-feature match based on prior data of similar customers, early focus groups revealed users wanting to use it as a 'proof of value' with internal measures showing how a customer's peers are benefiting as compared to the customers themselves with a certain feature. 

The interface did not require a redesign, but it shifted how I pitched the tool to leadership and new users. 

FEATURE USAGE BASELINES TOOL. 

The two designs were a byproduct of the research I initiated, however since it was an open-ended research, it spawned a collection of questions and observations later used by different teams. 

[NEEDS TO BE REWRITTEN] 

FEATURE USAGE BASELINES TOOL. 

Takeaways.

1. Prove value first, then ask for anything you want. Know when to throw your playbook, and when to pull it out.

The first few months, as much as i wanted to pause, breakdown, make sense of data, it was important to move fast. As designers, once we define our role as ICs that bring structure to ambiguities, remove hurdles for development, and help create something from nothing, it gets easier to push for authenticity. 

In the age of AI however, something from nothing has become much easier. The next frontier for designers will be something to 'something meaningful'

2. Discomfort → Curiosity → Research → Hypothesis → Design → Test → Validate → Build.

It's easy to be dismissive of discomfort; it takes months of struggle to get used to a new kind of team. Closed walls and ambiguities are usually a sign there's potential for true impact, if you can stick around long enough. 

3. Users can redefine the purpose a tool was originally designed for, and that's ok. 

4. More metrics can be helpful, but elementary metrics, when composed in the right way, can communicate an important message.