The Test Setup: Non-Inferiority as a Strategic Frame
There is a persistent assumption in conversion teams: if you make calling easier, fewer people will complete the digital flow. Phone and enrollment form are in competition.
A landing page redesign at a large energy provider tested that assumption. The redesign bundled several changes: a clearer value proposition, a larger hero image, an improved CTA, and a prominent phone number given significantly more visual weight.
The team used a non-inferiority framework: the primary metric (downstream plan comparison page views) was expected to hold flat. The hypothesis was that the redesigned experience could make calling dramatically easier without degrading the digital path.
The primary metric came in essentially flat. The non-inferiority threshold passed. But the secondary metrics told the real story.
What Happened When Calling Became Easy
Phone engagement in the variant did not tick up modestly. It increased by an order of magnitude. Time-to-first phone click on desktop dropped from roughly twenty seconds to under three seconds. Call volume roughly doubled across the test period.
Here is the paradox: digital enrollment starts and confirmations moved modestly upward, not down. Adding a dramatically more visible phone channel did not pull users away from the digital path. It appeared to recruit an entirely different behavioral segment.
This is the channel expansion paradox. The intuition that two conversion paths compete assumes a fixed pool of convertible users being divided across channels. The data suggests a more accurate model: different users arrive with different conversion modalities, and an interface optimized for only one modality leaves the other segment stranded.
Choice Architecture and the Multi-Path Interface
A landing page that visually de-emphasizes the phone number is not just providing less information. It is architecturally removing an option from the consideration set for any user whose natural preference is to call. Those users do not switch to digital enrollment. They leave.
When the redesign made the phone number prominent, it was not creating new demand. It was removing the friction that was suppressing an existing demand signal.
The choice architecture principle at work is option visibility: the set of choices a user acts on is determined not just by their preferences but by which choices they can actually see and reach within normal attention bandwidth.
The Attention Economics of Phone Prominence
The concern that a prominent phone CTA steals attention from digital elements rests on an attention budget model — the idea that the visual field is zero-sum.
This model is partly correct and mostly incomplete. Attention is not uniformly valuable. A digital enrollment CTA seen by someone who was always going to use digital is not meaningfully impacted by the phone number. But attention given to a phone CTA that converts a user who would otherwise have bounced has dramatically positive marginal value.
In a high-consideration category like energy plan selection, a substantial segment of users want to talk to someone before committing. For this segment, a digital-only CTA architecture is not neutral. It says: complete this process without talking to anyone, or leave.
Making calling prominent does not distract digital users. It rescues call-preferred users.
The Business Economics: Total Customer Acquisition
The correct economic frame is total customer acquisition cost across all channels. The test data surfaced one complexity directly: call center conversion rate dropped when volume roughly doubled. More calls meant a higher proportion of exploratory callers mixed into the volume.
This is not a failure of the design. It is a call center capacity and qualification problem. The design successfully recruited more callers. The call center infrastructure then determined what percentage converted.
A channel design decision should not be evaluated solely on the conversion rate of the channel it opens. It should be evaluated on net new customers acquired, at what cost, at what lifetime value.
The Framework: When Channel Expansion Works
The conditions that made this test work are identifiable:
High consideration, high consequence decisions. Energy plans, insurance, financial products — categories where a meaningful user segment has genuine questions that a digital flow does not answer well.
A currently suppressed call signal. If phone engagement in the control is near zero with a de-emphasized number, there is latent demand to unlock.
Call center capacity to absorb volume growth. Operational readiness is part of the test design.
Non-inferiority framing for the primary digital metric. If your organization requires the test to show digital lift to ship, you will almost certainly kill channel expansion tests before they ship.
What the Data Actually Proved
This test did not prove that phone prominence helps digital conversion. It proved something more foundational: the two channels are not in competition for the same users.
An interface optimized only for digital enrollment is not a neutral baseline. It is an interface that has made a design decision to deprioritize one segment of its potential customers.
Making calling easy did not hurt digital. It served the customers digital was never going to convert.