Director, User Experience Design
AutoTrader.ca
May 2023 - September 2024
Automotive, e-commerce, marketplace, retail, F&I
User research
Product strategy
Data
Design transformation
Evangelism
When AutoTrader’s growth began to plateau despite record-breaking funnel metrics, it became clear that traditional optimization tactics had reached their ceiling: we had maximized conversion rates and outperformed industry benchmarks already, and needed to find a new way to hit our annual goals.
As Director of User Experience Design, I introduced the Jobs-to-Be-Done (JTBD) framework as a new way to segment and understand 26 million users: shift away from who the users are to why they use AutoTrader. Using Tony Ulwick’s Outcome-Driven Innovation (ODI) process, we combined qualitative interviews with large-scale quantitative research to map every step of the “buy” and “sell” journeys, identify unmet needs, and quantify the intent of each user segment.
The results reframed AutoTrader’s strategy. We discovered that nearly half of our users were in “research mode”; not yet ready to transact, but highly valuable if nurtured correctly. This insight led to measurable change across the business: new investments in early-stage experiences, the creation of a brand-new OEM product called Storefront (signed by Mitsubishi and Hyundai within six months), and a national marketing campaign, “The Answer is AutoTrader,” aligned around user intent. It also unlocked new revenue opportunities in financial products like pre-qualification, trade-in, and loan tools, driving a complete overhaul of dealer-facing packages.
What began as a search for incremental growth became a fundamental transformation. Jobs-to-Be-Done gave AutoTrader a quantifiable way to connect customer intent to commercial opportunity, turning UX insight into enterprise strategy and fueling the next chapter of the company’s growth.
AutoTrader operated under private-equity ownership, which means a constant pressure to sustain revenue while hitting aggressive annual growth targets. Revenue came primarily from dealerships (paying “customers”), whose main value metric was leads. As a result, the UX was deliberately optimized for the linear, lead-generating journey: search > product listing view > contact dealer.
For years we refined this flow through rigorous CRO and A/B testing. By 2022 it felt like we had squeezed out all the proverbial juice: the funnel was performing at historic highs and well above industry benchmarks, yet we were still struggling to deliver a meaningful lift.
In executive conversations, we were repeatedly debating “what else” we could optimize when the traditional knobs (traffic and conversion) were already near the ceiling. Meanwhile, GA data exposed an interesting pattern: nearly 50% of users searched without ever opening product pages, and many who did view product pages never reached out to the dealer.
My emerging hypothesis:
Almost 50% of our users were not ready to buy and are using AutoTrader for an entirely different job.
After years of user research, we already had a clear picture of the car shopping journey. It often spans four to fourteen months, and - as the second-largest purchase most people make - it naturally involves extensive research, comparison, and confidence building.
Our working hypothesis became that this “entirely different job” wasn’t about transaction details: it was about researching the market, exploring for fun (car enthusiasts), monitoring trends, getting the most out of your car ownership.
Inside a lead-driven business model, these “researcher”, “car lover”, “car owner” personas had little perceived value. If they’re weren’t submitting a lead, they’re not contributing to revenue. I argued the opposite: today’s researcher is tomorrow’s active buyer. If we meet their needs early, we increase the likelihood they’ll return and convert when ready.
Our sister marketplaces reinforced this perspective. In regular insight-sharing sessions with CarSales (AU) and AutoTrader UK, their data showed that consumers who conducted research on their marketplaces went on to submit 160% - 240% more leads within 4-6 months compared to those who didn’t research on their platform.
To me, this insight felt like a holy grail: an opportunity to solve multiple challenges at once. If we serve the needs of these users, we not only increase lead volume but also attract new audience segments, drive organic top-of-funnel growth, and unlock secondary benefits: higher NPS, lower acquisition costs, stronger brand equity, and ultimately, market share growth.
This reframed the problem entirely. My challenge was now to:
Traditional segmentation tools couldn’t solve this. Personas focus on who the user is, not why they use a product—and with 26 million users, demographic segmentation simply breaks down.
For example, a 36-year-old mother with two kids whose car just broke down will behave almost identically to an 18-year-old who just got their first job and needs a car to commute. What unites them is not who they are, but the problem they have (“I don’t have a way to commute”) and the outcome they seek (“I need a car now”).
Jobs to Be Done (JTBD) offered the most logical, structured, and scalable framework for this challenge. Originally popularized by Clay Christensen in the early 2000s through the now-famous milkshake case study, JTBD reframes customer behavior around causal motivation. Christensen argued that demographics rarely cause a purchase, but a problem to be solved, a “job to be done,” almost always does.
Building on that foundation, Tony Ulwick developed Outcome-Driven Innovation (ODI), turning JTBD into a quantitative, repeatable process for large organizations to identify unmet needs and prioritize opportunities systematically.
The most significant advantage, however, is that JTBD transcends UX. Unlike personas or funnel analytics, Jobs to Be Done connects product, marketing, and sales around a common understanding of customer intent.
A well-known example is Intercom, whose product team adopted JTBD to re-segment their users and even restructure their product lineup. By shifting from generic personas to customer “jobs” (e.g., acquire new customers vs. support existing customers), they repositioned a single all-in-one product into four focused offerings, each built and marketed for a specific job. This shift unlocked explosive growth and validated the business value of the JTBD approach.
AutoTrader’s brand promise is to be “the most trusted place for Canadians to buy and sell cars”. To live up to that, we needed to deeply understand the two primary jobs we were solving for: “purchase a car” and “get rid of a car.”
What specific steps make up each process? What are the desired outcomes of those steps? And what related emotional or social jobs might influence user decisions?
STEP 1
QUALITATIVE DISCOVERY: IDENTIFY CORE JOBS
To operationalize Jobs to Be Done at scale, we used Tony Ulwick’s the Outcome-Driven Innovation (ODI) methodology, starting with extensive consumer interviews to capture desired outcome statements. Data from the interviews helped us crystallize the steps that go into a process of two primary jobs AutoTrader wants to solve: buying a car, and selling a car.
Below is a simplified view of the “Purchase a Car” job and its related steps.
Each step carried its own set of needs and desired outcomes, which we codified as job statements.
Because this journey isn’t linear (for some users, defining a budget might happen before research, while for others it happens later), we treated each job step as its own mini-job. This approach allowed us to evaluate how well we currently supported that specific intent and where new opportunities existed.
STEP 2
MAP JOBS TO EXISTING EXPERIENCES
Next, we mapped each identified job step against the existing AutoTrader experience to:
1) identify which corresponding site pages or tools currently served that need;
2) expose gaps where the experience was incomplete or misaligned.
This exercise revealed a key insight: we were effectively supporting only about 25% of all the total job steps in the end-to-end car buying journey.
That meant ¾ of the real consumer needs - especially early-stage research and decision-support jobs - were underserved or unsupported. To determine where to invest, we now needed to understand the size and the value of each job segment and its potential ROI.
STEP 3
QUANTIFY SEGMENTS BASED ON JOBS
Partnering with our data analytics team, we launched a site-wide intent survey designed to understand “What are you here to do today?”.
Each response corresponded to one of the identified job categories with subsequent questions drilling down into a more nuanced understanding of where they’re at in their journey (e.g. “I don’t know what I’m looking for nor do I know my budget” vs “I know my budget but don’t know what I’m looking fo”).
It was a comprehensive 10-min survey with conditional visibility and decision trees that ran for just over two months, collecting over 5200 responses. This data allowed us to:
STEP 4
BUILD SEGMENTATION MODEL
The survey results were then extrapolated to create a job-based segmentation model for 26 million users around job categories:
For the first time, we had a clear, quantifiable view of why people used AutoTrader, framed by the progress they sought to make rather than who they were or which stage of the funnel they were in.
STEP 5
MAKE IT FAMOUS
The JTBD survey quickly became a breakthrough moment internally, bridging UX, product, marketing, and analytics under a single shared model of user intent.
To embed this mindset across the organization, one of my key mandates became to “make it famous.”
Over the next six months, we presented the JTBD framework and findings in:
It soon became part of our shared language. Product roadmap discussions began with:
“What problem are we trying to solve? What is the job to be done here? How big is this user segment?”
The Jobs-to-Be-Done segmentation effectively became AutoTrader’s north star for customer understanding, uniting business strategy, user experience, and growth initiatives under one framework.
The insight that nearly half of our audience was in “research mode” and not yet ready to transact triggered measurable changes across three major areas: product experience, revenue strategy, and brand positioning.
We invested heavily in the “I’m here to research” experience, enhancing key surfaces like the Home Page, Global Navigation, My Account, Search Results (SRP), and Vehicle Details Pages (VDP) to better support exploration, comparison, and confidence-building. Instead of forcing every user into a make-and-model search, we introduced features built around help-me-choose intent. We also repositioned the Editorial section from a pure SEO play to a trusted, unbiased content hub for users not yet ready to buy.
One of the most significant outcomes was the launch of AutoTrader Storefront, a new OEM product built from JTBD insights to capture undecided shoppers earlier in their journey. I envisioned, pitched, and secured executive buy-in for the concept despite early skepticism. Within six months, we signed Mitsubishi and Hyundai as pilot partners. Storefront allowed OEMs to showcase model lineups, brand stories, and research content on-platform, turning what was once a “non-revenue audience” into a monetizable one.
JTBD also became the foundation for our 2024 national campaign, “The Answer is AutoTrader.” The campaign spoke directly to the real problems users were trying to solve (“Cost confusion?”, “Selling slowly?”), helping reposition AutoTrader from a transactional marketplace to the go-to platform for any auto-related need.
Finally, survey data validated our expansion into financial and ownership-related products - including pre-qualification tools, loan pre-approval, and trade-in estimates -triggering a full overhaul of dealer-facing product packages to support them.
What began as a search for more leads and traffic evolved into a broader transformation: a new, quantifiable way to connect customer intent to commercial opportunity, proving yet again that UX insights can drive incredible business growth.
