Fleet Optimizer — Nayara Marques
Loadsmart · Chicago · 2021—2023

Capacity matching for private truck fleets

Fleet Optimizer found paying loads for companies that run their own trucks, to fill trucks that would otherwise drive back empty. The offers went out as a plain match email. A design sprint produced a web platform to replace it, but when people from fleet teams compared the two, the tactical staff, who handle the loads day to day, chose the email; only their managers preferred the platform. We redesigned the email instead, and the response rate rose 250%.

Fleet teams received paying loads by email. A sprint designed a platform to replace it, but the people who handle the loads chose the email. We redesigned the email, and the response rate rose 250%.

Role
Senior Product Designer, sole designer. I ran the research and facilitated the sprint.
Team
Me, 2 PMs, and colleagues from product, business and engineering in four countries. Sponsored by The Home Depot.
Context
A new, experimental product line at Loadsmart, a logistics technology company.
Contribution
qual + quant research · sprint facilitation · concept design · comparison testing · product strategy

01 — Context & my role

Loadsmart automates freight pricing and booking. Fleet Optimizer targeted private and dedicated fleets: companies that own their trucks or serve one shipper.

The team had an idea for the product, matching fleets' spare trucks with shippers' loads, but no evidence that fleets wanted it. My job was to find out what private and dedicated fleets had in common, and to take the findings to the PM and VP of Product, even when the findings argued against the team's plan.

View Matches. The web platform the design sprint produced, where a fleet would upload its spare capacity and receive matching loads. The comparison test showed that fleet staff preferred the email, so it was put on hold.

02 — The problem

The team assumed fleets would want to fill their empty return trips. Interviews with six people from fleet teams showed why most did not.

Four findings from the interviews mattered:

Managers spend most of their time in spreadsheets, chasing fleet metrics. Dedicated fleets are paid for the round trip. Around 70% of return trips run with no load, and the customer already pays for it. Dedicated service always comes first. A return load (a backhaul) that risks the committed route isn't worth it. Loading time decides whether a return load is worth taking. A load the driver has to wait for, strap down and cover with tarps can add four hours to the driver's day.

So backhauls were not the priority the team had assumed. And whether a fleet has a truck free for one depends on how much its main shipper sends, the driver's legal hours of service and the kind of load.

03 — Constraints

The limits the work ran under, starting with a design sprint that no fleet manager could join.

Every idea the sprint produced was a guess until fleet staff could test it. Online, with participants in the US, Peru, Brazil and Argentina. Investment depended on evidence. The company could change strategy, and later did.

04 — Process & key decisions

Four stages: interviews, the sprint I facilitated, a comparison test, and the decisions the test produced.

I interviewed six people from fleet teams for 30 minutes each. Their answers showed that return loads mattered less than the team had assumed.I ran it online with the team in four countries. It produced View Matches, the platform shown under Context & my role.Six participants saw the platform and the current email, one after the other, and picked one. The staff who handle loads day to day chose the email, and their managers chose the platform, a split the sprint had not planned for.

Key decisions

The sprint's winner was View Matches, a site for uploading capacity and receiving load matches. I tested it against the existing match email, which sends fleet staff loads that fit their spare trucks, instead of testing it on its own. We rebuilt the email so it opens with the lane, the route from origin to destination, and kept View Matches on the roadmap for managers.

05 — The solution

The redesigned match email is built for fleet staff who decide on a load from their phone, often out in the yard. The figures show the email before and after, then the parts that changed.

Before

The old email opened with the fleet's internal number. On a phone, the price and the button to accept were below the first screen, so staff had to scroll before they could decide.

After

The new email opens with the lane, then the price and the accept button, so staff can decide without scrolling. Feedback options come last.

The subject and first line name the origin and destination, so staff don't have to open the fleet's record to see where the load goes. Fleet staff decide whether to accept a load mainly on its price, so the price and the accept button come first. The email says whether the load is a one-off or repeats every week on the same lane, so staff know whether they are committing a truck once or every week. New sections let staff reply with feedback on a load, or tell Loadsmart about more trucks they have free, instead of only saying yes. The feedback also showed why loads were rejected: mostly loading time.

06 — Impact

Two results: the comparison test decided where fleet staff would receive matches, and the redesigned email changed how many of them replied.

The platform

For the people who handle loads, the platform was the wrong place to receive matches.

Recommending the email meant putting on hold a platform the team had already designed.

What shipped Considered · only managers chose itView Matches platform ChosenRedesigned match email

The email channel

The old email hid the price and the accept button below the first screen on a phone, so fleet staff replied less than they could have.

The team had read four replies in six months as a sign that email didn't work.

Results

  • +250%response rate after the redesign
  • 4replies in the six months before: the baseline

07 — What I learned

Four lessons, two about the product and two about the practice of running research and a design sprint.

Asked whether a design is good, people tend to agree. Asked to choose between two, they decide. An email is a less impressive thing to show than a new platform, but it was the better product decision. Use the channel users already check. The sprint got the team in four countries to agree in a week, but without users its output was only a guess. Now I plan how a sprint's output will be tested before the sprint starts. A few days of interviews changed the plan while it was still cheap to change. The amount of research should match the size of the decision.

08 — What this case doesn't cover

The vertical did not last, so some outcomes can't be shown. These are the gaps, and how to read the numbers above.

The company later stopped working with private fleets, so there is no long-term revenue figure. The 250% is a rise in email replies from a small starting number, and says nothing about revenue. Only the matches view is shown; the rest was never tested or built. Six interviews and six test sessions, dedicated fleets only. Enough to redirect the plan, not to generalise. Findings about drivers came from managers. No driver was interviewed.