Conversion Catalysts preference study

Overview

FedEx wanted to know which value propositions actually moved prospective shippers toward registration, and which ones were carrying the discount-led acquisition stack on inertia. THEFT Studio ran a global Max Diff preference study across 508 prospects in five regions, balanced across personal and business segments, testing 30 USP buckets and applying TURF analysis to find the smallest message set that reached the broadest audience. The readout gave acquisition and product marketing a preference-backed basis to reframe the registration funnel away from discount-led copy and toward reliability, data security, and practical fit.

The problem

Acquisition copy at FedEx had been discount-led for years. Free account, dollars-off, percentage-off — the standard B2B incentive stack. Conversion held, but no one had stress-tested whether discounts were actually doing the work, or whether they were just the easiest thing to ship. The team needed evidence to either keep the strategy or change it, not a creative recommendation that traded one untested message for another.

The right shape for that question was a forced-choice preference study at scale: show prospects the same set of value propositions, force them to rank, see what holds across regions and segments. Discounts could compete on the same terms as reliability, security, and brand promise — and prospects would tell us, by their choices, what actually moved registration.

Method

Max Diff preference study across 508 prospects in five global regions: US/CA, LATAM, Europe, MEISA, and APAC. Sample split 254 personal shippers / 254 business shippers, balanced for regional coverage. Thirty USP buckets tested, drawn from existing FedEx acquisition messaging plus competitor scan plus first-principles value-prop hypotheses.

Max Diff isolates which USPs move prospects toward a behavior — registration in this case — by forcing repeated best/worst choices across rotating subsets. The output is three views: feature count (how often a USP wins), average feature utility (how strongly it wins), and preference share (the population-level pull). On top of that, TURF analysis on message combinations to find the smallest message set that reaches the broadest audience without redundancy.

Outcome

Acquisition and product marketing got a preference-backed basis for rewriting registration-funnel messaging. Reliability and data security took their place at the top of the stack, with free account, competitive rates, and fast delivery as the practical-value core underneath. The reframe — discount-led to reliability-led — became the spine of the next acquisition push.

The study also shifted internal vocabulary. Future acquisition decisions had a preference baseline to argue against. New messaging gets tested against the existing preference rank instead of being shipped on opinion.