How to Diagnose Landing Pages for More Conversions

TL;DR

Improve conversion performance by first identifying the universe of possible conversions, then diagnosing what customer behavior reveals and why it occurs. A cable company increased conversions by 10.6% after naming its form the “Quick Mobile Transfer,” reducing friction, and adding navigation that encouraged visitors to conclude the process would be fast and easy.

Transcript

good afternoon everyone we're back with another episode and I would like to take you immediately to a strange slide you can see a picture on the one hand you'll see a form that we recently optimized for one of the largest cable companies in the country and next you'll see a particular fish kind of an ugly deepwater fish and I want to ask you a curi... Read More

Key Insights

• The discovery triad is a diagnostic method built around how, what, and why. Teams begin with the desired improvement, ask what the data reveals about customer behavior, and then investigate why customers are behaving that way before selecting an optimization strategy.
• A name can foster a stronger customer conclusion than a direct claim. Renaming the cable company’s form the “Quick Mobile Transfer” implicitly communicated speed and simplicity, supporting the desired perception without relying only on promotional statements about the process.
• The treatment produced a 10.6% relative increase after the page had experienced two years of testing without improvements. It combined a new process name, reduced friction, and added navigation, with all changes directed toward the conclusion that the transfer would be fast and easy.
• A variable cluster is a collection of coordinated page changes intended to produce one customer conclusion. In the cable company example, naming, navigation, and friction reduction worked together, so the experiment evaluated the combined strategy rather than attributing the outcome to one isolated element.
• The universe of possible conversions is the subset of incoming traffic that could, should, or would realistically convert. Evaluating all visitors as equally convertible can make a strong page appear weak and lead analysts to address page design when traffic qualification is the real issue.
• A reported 1% conversion rate can conceal a 10% rate among realistically convertible visitors if 90% of traffic is unsuitable. The example is intentionally simplified, but it demonstrates why teams must estimate qualified conversion opportunity before judging performance or prescribing page changes.
• Unqualified visitors should be separated from the conversion funnel when their goals cannot match the intended action. For the cable company, some visitors lived outside serviceable areas, while others only wanted to cancel and needed a prominent route out of the transfer process.
• The 10.6% increase was measured across all traffic rather than only the universe of possible conversions. Because many visitors could not qualify for the intended action, the improvement among realistically convertible customers represented a more substantial effect than the aggregate result alone suggested.

Questions & Answers

Q: How do you diagnose a landing page before changing it?

Start with the discovery triad: define how performance should improve, ask what the data tells you about customer behavior, and then ask why customers behave that way. The data should be interpreted as evidence of behavior rather than treated as a metrics report alone. The resulting explanation should guide the hypothesis and determine which coordinated page changes deserve testing.

Q: What is the universe of possible conversions?

The universe of possible conversions, abbreviated as UPC, is the smaller group within total page traffic that could, should, or would realistically complete the intended action. Some visitors may be unqualified, outside the service area, or pursuing another goal. Separating those visitors from genuine prospects provides a clearer basis for interpreting conversion performance and identifying the actual problem.

Q: Why can an overall conversion rate be misleading?

An overall conversion rate treats every visitor as if conversion were possible, even when much of the traffic is unsuitable. The transcript illustrates that a page showing a 1% overall rate could effectively convert 10% of eligible visitors when 90% of traffic is wrong. Without estimating the universe of possible conversions, a team may try to fix the page instead of its traffic mix.

Q: How can naming improve landing page conversions?

Naming can influence the conclusion customers form about a process. The cable company renamed its form the “Quick Mobile Transfer,” a label intended to suggest that the task would be fast and easy. The principle is that people may treat a name as revealing the essence of something, giving it more persuasive weight than an explicit claim placed on the page.

Q: What changes increased the cable company’s conversions?

The treatment reduced friction, added navigation, and renamed the process the “Quick Mobile Transfer.” These elements formed a variable cluster designed to create one conclusion: transferring service would be fast and easy. The experiment generated a 10.6% relative increase after the company had spent two years testing the page without producing an improvement.

Q: Why should cancellation traffic be removed from a conversion funnel?

Visitors who only want to cancel are not realistic prospects for a service-transfer action, so keeping them in that funnel distorts the traffic mix and frustrates customers. In the cable company example, people could not determine how to cancel, called customer service, and complained. A more prominent cancellation link helped those visitors leave the funnel and clarified the group that could actually convert.

Q: What does the discovery triad reveal about customer data?

The discovery triad connects a desired performance improvement to customer behavior and its cause. The “what” question asks what the data tells the analyst about customer behavior, not merely what the report says. The “why” question then seeks the reason for that behavior. Together, these answers provide the insight needed to form a hypothesis for improving the page.

Q: Why was the 10.6% increase more significant than it appeared?

The 10.6% relative increase was calculated using all traffic reaching the page, including visitors who could not qualify for the intended action. Some lived outside areas the cable company could serve, while others wanted to cancel. Because the realistically convertible audience was smaller than total traffic, the measured aggregate lift understated the treatment’s impact within the universe of possible conversions.

Summary & Key Takeaways

• The discovery triad turns data into an optimization strategy by asking three connected questions: how to improve results, what the data reveals about customer behavior, and why customers behave that way. The behavioral answer to “what” and the explanatory answer to “why” should determine how a page is changed.

• A cable company needed moving customers to register service at a new address. Its redesigned mobile experience reduced friction, added navigation, and renamed the form the “Quick Mobile Transfer.” These changes were designed as a variable cluster that fostered one conclusion: completing the process would be fast and easy.

• Conversion rates can be misleading because not every visitor can, should, or will convert. The universe of possible conversions isolates qualified traffic from visitors who cannot receive service or only want to cancel. Understanding that smaller group prevents teams from diagnosing the wrong problem and clarifies the significance of measured improvements.