Salesforce: Buyer-Led Sales
Prospecting has traditionally assumed the form of a high volume, non-personalized “spray and pray” approach. These activities are typically reserved for junior sellers as they are not strategic in nature but critical to long term pipeline health. It is becoming more critical that prospecting is not only done by junior “pipeline builder” sales roles, but also by more senior “deal closer” roles who rely on pipeline to meet revenue targets.
Advancements in technology such as AI meant that Salesforce could re-imagine the sellers role and reliably automate these tasks to provide sellers with more time back to focus on strategic sales further down the funnel.
My Role (2023)
I was the UX manager and lead strategist for this project. I designed initial north start concepts for the yearly Salesforce CKO company kick-off conference. I led a team of designers and a user researcher to strategize an MVP for release later in the year.
SErvices
UX/UI Design, Design Strategy, Design Mentorship, Lean UX, Workshop Facilitation
Salesforce “CKO” Northstar
I worked with our product leadership to develop a vision for how we might reimagine the top of funnel selling experience for both buyers and sellers. These prototypes were intentionally designed to push on what was possible but were grounded in technical feasibility. These designs were shared at CKO, the company kick off process led by Marc Benioff to determine what investments to make for the rest of the year.
Buyer Experience
A prototype envisioning AI as a personal BDR that provides a guided walk-through of a company’s website and connects the interested to buyer to a personalized sales space. I created the initial wireframes, another designer developed the final UX for this.
Seller Experience
A prototype envisioning an AI automated prospecting assistant, that identifies potential prospects, ranks them according to a buying signal, and suggests personalized outreach messages tailored to each prospect.
Developing an MVP
After CKO, this project got the green light. The prototypes were also shared during a Gartner review and they were received with excitement and positive feedback that Salesforce wanted to capitalize on. These two factors made the project a high priority and highly visible internally.
Expediting UX / lean ux
For the MVP, we had an extremely tight timeline and a big ask to re-imagine prospecting with AI. One of the risks of having such a short timeline is it forced us to narrow our thinking early on in the process without user feedback.
The process I outline next helped us make educated decisions at each stage in the process. I also created a cadence of twice weekly cross functional UX checkins to get internal feedback as quickly as possible, since we didn’t have any time to spare to be out of alignment.
Existing Research
Normally for a project envisioning a new product idea, I would advocate spending a couple weeks doing discovery user research to broaden our thinking before narrowing it. However, we didn’t have this time. Luckily, Salesforce has been in the space of helping sellers sell for a long time, so we turned to secondary and existing research to guide our thinking.
Inspiration & Competitors
Next we looked at what competitors were doing and at inspiration from UX solving “parallel” problems. For example, we considered how dating apps solved the problem of making an educated “recommendation” of people for you to consider. We also looked at recruiting software.
Design Principles & “Jobs to be done”
We developed working design principles based on the secondary research insights and takeaways from reviewing inspiration and competitive landscape.
Salesforce uses the JTBD “jobs to be done” framework to organize design thinking from a user’s perspective. We identified the top jobs for our users to help us focus on solving for those needs first.
Increase my trust in the quality of prospects
Optimize my time for the most quality prospects
Maintain a steady pipeline across my funnel
Decrease the time it takes to send quality outreach messages
Focus my time on progressing good prospects
Workshop: Brainstorming & Alignment
I like to use workshops as focused time for weaving together big, broad thinking with narrowing to make decisions. At this stage in the project, it was time to start creating UX designs. However, before jumping to design decisions, we wanted to make sure our thinking was broad and we hadn’t missed anything. There were also lots of smart people involved and opinions about how things should be solved, but we didn’t have a lot of time to swirl on feedback loops.
We facilitated a workshop with the technical and product leads to accelerate the this process, to broaden thinking, create alignment, and arrive at decisions.
One important area of focus was to be smart and thoughtful about how our AI prospect recommendations logic worked. This would impact the end UX a lot.
UX for where the puck is going
One decision our team made early on this process was to stay laser-focused on creating a tool that sellers would actually want to use. One common pain point for the adoption of Salesforce tools by end users is that it is often designed to meet the needs of management vs. the actual end user of the product.
Making this decision meant we intentionally decided not to tackle some of the systemic complexity that comes with prospecting. For example, one major complexity in this space is attribution and how organizations track this to incentivize creating pipeline as well as closing deals.
However, our vision and goal was to make a lightweight, easy to use tool that account executives would be excited to use. The MVP would help validate this hypothesis, so we deprioritized anything that fell outside this scope.
CREATING A PROSPECTING Recommendations “Engine”
I collaborated with technical leads and data science to develop how the recommendation engine should work. This was a back and forth iterative process. I believe these kind collaborations create a better end product and I love creating them.
There were two buckets of data. The first bucket, “Fit” data, was made up of characteristics about the company and the buyer that could predict that they’re a good buying candidate. The second bucket was “Intent” data, or actions the buyer was taking that signaled they were in some stage of the buying process.
We worked together to define what made a prospect a “best match” for the seller to prioritize that day. We created the spectrum of “Strong-General-Unknown.” Originally we had played with a “High-Medium-Low” type of framework, but we decided that with signals we simply couldn’t know what we didn’t know. We didn’t want to rank a prospect as a “low” fit when perhaps we were just missing information about them. The intent of our engine was to empower the seller with positive signals, not to discourage them from pursuing a prospect.
UX Concept Explorations
Next, we started exploring design concepts. These wireframes were created by two other designers I was working with on this project.
Choosing a Concept
Ideally at this stage, we would have time to collect user feedback and help guide the choice of a concept. However, we had an ambitious timeline and we felt confident enough in our collective expertise to make a choice. After an internal review, we chose to commit to Concept B: Recommended Today. We believed it had the right balance of making a strong recommendation that would save sellers time and be a value add, while still giving them the ability to dig in and find prospects based on their expertise.
Rapid Research
Even though we needed to narrow on a concept, we decided to prioritize getting feedback on that concept with some rapid user research. Doing it at this stage would allow us to keep our process moving while still providing insights that would reduce the risk of mistaken assumptions. It’s much easier to change a wireframe than code.
Going Deeper: USEr flows and design details
With a concept strategy and user feedback, our team went deeper into creating designs and flows for the seller experience. One piece of complexity we needed to think about is how to recommend a prospect which could be any kind of object in Salesforce (an account, a person), and how to show if we knew who in the organization was showing intent to surface a recommended contact within that organization.
Dreamforce Demo (Taking it HIgh fidelity)
After the wireframes were completed, our engineering team had enough UX to be unblocked. I switched focus to taking the designs higher fidelity for a Dreamforce demo. We had a couple of asks from our executive leadership. One, we wanted to integrate more closely with marketing, so we brought in marketing segment data. And two, we wanted to show how it could work with Sales Cadences, the tool that had been developed to help pipeline builders or BDRs. In this demo they can add the prospect to a sales cadence to automate the outreach.
Salesforce Announces New Product
After Dreamforce, Salesforce officially announced the new prospecting tool driven by CRM data and AI. I ended my tenure with Salesforce around that time, however I hope the new product is doing well!
