Conversion funnel analysis on website

Conversion Funnel Analysis on Website

Funnel analysis identifies at what stage users leave without completing goal. For e-commerce: view → cart → checkout → payment. For SaaS: signup → onboarding → activation → payment.

GA4 Funnel Setup

// Event markup for each funnel step

// Step 1: Product page view
gtag('event', 'view_item', {
  items: [{ item_id: product.id, item_name: product.name, price: product.price }]
});

// Step 2: Add to cart
gtag('event', 'add_to_cart', {
  currency: 'RUB',
  value: product.price,
  items: [{ item_id: product.id, quantity: 1 }]
});

// Step 3: Checkout start
gtag('event', 'begin_checkout', {
  currency: 'RUB',
  value: cartTotal,
  items: cartItems
});

// Step 4: Payment info
gtag('event', 'add_payment_info', {
  payment_type: 'card',
  value: cartTotal
});

// Step 5: Purchase
gtag('event', 'purchase', {
  transaction_id: order.id,
  value: order.total,
  currency: 'RUB'
});

Create funnel in GA4: Explore → New exploration → Funnel exploration. Add steps, include breakdown by device/segment.

Analysis via BigQuery

-- Conversion at each funnel step
WITH funnel AS (
  SELECT
    user_pseudo_id,
    MAX(CASE WHEN event_name = 'view_item' THEN 1 ELSE 0 END) AS viewed,
    MAX(CASE WHEN event_name = 'add_to_cart' THEN 1 ELSE 0 END) AS added,
    MAX(CASE WHEN event_name = 'begin_checkout' THEN 1 ELSE 0 END) AS checkout,
    MAX(CASE WHEN event_name = 'purchase' THEN 1 ELSE 0 END) AS purchased
  FROM `project.analytics.events_*`
  WHERE _TABLE_SUFFIX BETWEEN '20240301' AND '20240331'
  GROUP BY user_pseudo_id
)
SELECT
  COUNT(*) AS total_users,
  SUM(viewed) AS viewed,
  SUM(added) AS added_to_cart,
  SUM(checkout) AS started_checkout,
  SUM(purchased) AS purchased,
  ROUND(SUM(added) * 100.0 / SUM(viewed), 1) AS view_to_cart_rate,
  ROUND(SUM(checkout) * 100.0 / SUM(added), 1) AS cart_to_checkout_rate,
  ROUND(SUM(purchased) * 100.0 / SUM(checkout), 1) AS checkout_to_purchase_rate,
  ROUND(SUM(purchased) * 100.0 / SUM(viewed), 2) AS overall_cvr
FROM funnel;

Funnel Segmentation

-- Funnel by device
SELECT
  device_category,
  COUNT(DISTINCT CASE WHEN step >= 1 THEN user_id END) AS step1_users,
  COUNT(DISTINCT CASE WHEN step >= 2 THEN user_id END) AS step2_users,
  COUNT(DISTINCT CASE WHEN step >= 3 THEN user_id END) AS step3_users,
  ROUND(COUNT(DISTINCT CASE WHEN step >= 3 THEN user_id END) * 100.0 /
        NULLIF(COUNT(DISTINCT CASE WHEN step >= 1 THEN user_id END), 0), 1) AS cvr
FROM funnel_data
GROUP BY device_category;

Analyzing Drop-off Reasons

After identifying "bottleneck"—investigate why:

Tools:

• Hotjar/Clarity: session recordings of users stuck on step
• Heatmaps: where they click on problem page
• Form Analytics: which fields they abandon

// Track abandonment on checkout form
document.querySelectorAll('#checkout-form input').forEach(field => {
  field.addEventListener('blur', () => {
    gtag('event', 'checkout_field_blur', {
      field_name: field.name,
      has_value: field.value.length > 0
    });
  });
});

// Track page exit without submission
window.addEventListener('beforeunload', () => {
  if (document.querySelector('#checkout-form') && !formSubmitted) {
    gtag('event', 'checkout_abandonment', {
      last_field: lastFocusedField
    });
  }
});

Funnel Timing Analysis

Time between steps is also informative:

SELECT
  user_id,
  MIN(CASE WHEN event = 'view_item' THEN timestamp END) AS view_time,
  MIN(CASE WHEN event = 'add_to_cart' THEN timestamp END) AS cart_time,
  MIN(CASE WHEN event = 'purchase' THEN timestamp END) AS purchase_time,
  TIMESTAMP_DIFF(
    MIN(CASE WHEN event = 'purchase' THEN timestamp END),
    MIN(CASE WHEN event = 'view_item' THEN timestamp END),
    MINUTE
  ) AS time_to_purchase_minutes
FROM events
GROUP BY user_id
HAVING purchase_time IS NOT NULL;

If average view→purchase time = 3 days, need retargeting reminder within these days.

Timeline

Event markup + BigQuery analysis + segmentation + report: 3-5 business days.