Funnel Analysis: Visualizing and Diagnosing Conversion Problems

Potential customers pass through a series of steps from the first touchpoint to the target action. At each step, a portion of prospects abandon the process (loss of intent or motivation, technical issues, etc.). Understanding where and why these losses occur is critically important for directing resources effectively. Funnel analysis visualizes these drop-off points, but diagnosis and resolution cannot be derived from the chart alone.

Funnel Chart

A funnel chart visualizes the breaks and density changes between a starting point and a target¹. It resembles a 100% stacked bar/column chart². The starting point (head/base) is the widest section, and the point where the target is completed (neck) is the narrowest.

Stages are typically represented top to bottom. However, left-to-right layouts are common in dashboards for space efficiency.

Design Rules

A funnel chart must contain at least 4 stages, including the top and bottom sections³. With fewer stages, identifying problem areas becomes impossible. More stages enable more detailed examination of the process⁴.

A perfectly triangular funnel chart hides the decreases between stages. The visualization should use a design where drop-offs at each step are clearly distinguishable⁵.

If the differences between stages are too small, bar or column charts are more effective than funnel charts^{6 7}.

Funnel Models

Sales Funnel and AIDA

The sales funnel has been used since the early 1900s to map the process from awareness to purchase⁸. The AIDA model, developed by E. St. Elmo Lewis⁹, consists of four stages: Awareness, Interest, Desire, and Action.

AIDA’s limitation is its linearity. Customers can move back and forth between stages and exit the funnel by changing their research behavior¹⁰.

Purchase Funnel

A model that expresses consumer-based purchase stages in 7 phases: Awareness, Interest, Evaluation, Decision, Purchase, Reevaluation, Repurchase¹¹. Also referred to as customer funnel, marketing funnel, and conversion funnel.

Conversion Funnel

A model covering search, performance advertising, and lead generation processes, referencing the purchase funnel. Used to illustrate the path a consumer follows through online advertising or search in ecommerce. Each stage can be independently optimized and conversion rates tracked¹¹.

AARRR (Pirate Metrics)

A product-focused growth framework defined by Dave McClure (500 Startups, 2007): Acquisition, Activation, Retention, Referral, Revenue¹². More applicable than AIDA in SaaS and mobile app contexts because it covers retention and referral stages.

Practical Application

An example funnel using the last 7 days of data from an ecommerce site:

This funnel shows the drop-offs between the cart and payment steps. Each drop-off can be addressed with separate communication strategies. Users who hold products in the cart but do not proceed to payment can be retargeted. A drop-off at the payment step may indicate technical problems (payment method errors, slow loading) or UX issues (complex forms, unexpected additional charges).

GA4 Funnel Exploration

GA4’s Funnel Exploration report provides funnel analysis directly within the platform¹³:

• Open vs closed funnel: A closed funnel requires users to enter at the first step; an open funnel allows entry at any step
• Up to 10 steps can be defined
• Segment comparison: Up to 4 segments can be analyzed side by side
• Elapsed time between steps: Shows how long users take to move from one step to the next
• Next action analysis: Shows what drop-off users did next
• Trend view: Tracks changes in funnel performance over time

Drop-off Diagnosis

A funnel chart shows where losses occur, not why. Drop-offs can be examined in 5 categories¹⁴:

1. Attention loss: Weak visual hierarchy, buried CTAs
2. Clarity loss: Unclear messaging, pricing complexity
3. Trust loss: Missing trust signals, unclear return policy
4. Friction: Long forms, mandatory account creation, slow page loading
5. Motivation loss: Value proposition not clear enough at the decision point

High-volume exit points are not always the most critical problem. An upstream intent mismatch can cause greater revenue loss than a visible drop-off at checkout¹⁵.

Accurate diagnosis requires combining quantitative data (drop-off rates, transition times) with qualitative data (session recordings, user surveys).