People.ai's design system underwent three major changes between 2018 and 2022; each change had different objectives and challenges.
Over the course of my time at People.ai, I had taken over the management and maintenance of our component library with the intention of building it out into a full design system to improve consistency within our product and efficiency for our team when working.
[
2018-2019
When I joined People.ai, the team had a component library that covered what was in the application that covered various basics, but also left room for interpretation otherwise:
- A limited color palette and no rules around usage meant that colors were arbitrarily added when needed and did not necessarily mesh with colors already in use.
- While icons were generally taken or adapted from Feather Icons, there was similarly no guidance around usage or sizing, which resulted in inconsistent usage and unbalanced visual weight.
Color System
Our marketing team began to work on a brand refresh, including a new logo and a new identity. Between this and another designer beginning to work on data visualizations, I began working closely with our marketing creative director on a new color system for our application.
People.ai's initial color system (left) versus the expanded color system (right).
We determined that we should focus on creating something that both sales and marketing can use to be aligned.
IDEATING AND BUILDING...
In 2020 and 2021, People.ai made acquisitions that added differing products with different design paradimgs to its product portfolio.
In addition to user interviews and internal shadowing, I also went through reports generated by our internal marketing operations and sales operations managers to get a sense for how the two different organizations slice, dice, and use our data. We also had walkthroughs of some of the reports that our customers generate and use as additional reference. The goal was to find what the commonalities of these reports are so that we had a qualitative basis for what "alignment" between the two organizatons could mean.
The product manager and I also had a second round of customer interviews where we also spoke with sales operations and SDR managers in addition to following up with some of the marketers we had spoken with previously. I had created some sketches based on the overlapping data from the different reports to help address the marketing pain points that we had identified, and we not only further looked into the dynamic between sales and marketing but also gathered some feedback on our concept during these interviews.
During this process, we were asked to work with engineering to create an early proof-of-concept that could be demonstrated at an upcoming conference. During that second round of customer interviews, we were able to divide the data pertaining to marketing campaigns into a "good to know" bucket and a "what is actionable" bucket, and we used everything we had collected and synthesized so far to create the proof-of-concept:
- Campaign memebers have a lifecycle that's based on their interactions with the corporation, so we used these as a top-level filter in the form of tabs.
- Campaign members are typically assigned to a sales rep (either an AE, SDR, or BDR depending on the company and how they split their different sales functions), so this information lets sales managers and marketers know who is lagging in reaching out to campaign members.
- Campaign members also have their campaign status that, in conjunction to where they are in the lifecycle, determines what the the expected approach is.
This proof-of-concept was shown at the SiriusDecisions 2018 conference, where we were able to collect feedback from hundreds of conference attendees.
The feedback we gathered was incredibly valuable, not only in validating the concept but also in helping us narrow our focus to individual contributors in field marketing and demand generation.
BUILDING AND REFINING...
I continued to iterate on the design for Campaign 360 by working closely with our marketing to further flesh out what the stages of a customer's lifecycle is and to help work everything into different user journeys.
- A Field Markteter or a Demand Generation Manager can view a campaign that they have created or completed, see at a glance the number of campaign members that had different levels of engagement (none, single-directional, bi-directional, etc.), which members fall into each category, and which reps need to be following up with the marketing-generated leads and contacts. When selecting a campaign member, the marketer can see in more detail what the nature of the completed and upcoming activities with that member are.
- A Sales Manager can come into the interface and look up their reps to see if their respective leads are being followed up on and which reps are lagging in their SLAs.
Our product manager, as well as our customer success team, worked with a couple of customers to conduct further validation of this offering as well as begin to enable customer roll-outs.
IN THE END...
Up until it was discontinued in mid-2020 due to a pivot towards a focus on front-line sales managers, Campaign 360 consistently had the highest engagement rate out of all the modules in People.ai's web app since its release in February 2019, as well as a 100% retention rate for the first 13 months. However, it only had 6 weekly active users by the time it reached EOL (it had seven at the onset).
Feedback from users has included one customer stating that their global demand center has begun to rely on Campaign 360 to inform their decision-making.
LESSONS LEARNED...
There were a couple of lessons learned that were applied towards streamling and improving later projects:
- Design and Engineering actually had started working on the product almost-simultaneously. The back-end architecture was considered locked and finalized while research was still being conducted, which resulted in limitations on what data could be represented.
- As evidenced by the low amount of weekly active users, most customers opted to continue to use their own dashboarding due to the lack of flexibility in data to be displayed. Later efforts in addressing marketing usescases were focused on providing marketers the data they needed through APIs rather than through a UI.
- A large chunk of the research was oriented around determining what persona to focus on. As the product manager and I were both new to the field in which People.ai operates, we might not have been able to avoid this, but we now had a sense for how questions and jobs to be done generally apply across senorities. This doesn't necessarily translate to functions and verticals, however.
- In the whirlwind of trying to research, design, and ship, I didn't take a step back to make sure we were organizing and processing our notes effectively; in later projects, I pushed for conducting debrief sessions and creating documents and affinity diagrams as much as possible so that we could have more reliable recall of findings.