Grometrics
Log in
Back to Guides

Funnels

8 min read

Diagnose where visitors drop off before buying

Use funnel steps to find whether the problem is traffic quality, page intent, checkout friction, or offer clarity.

The question

Visitors are arriving, but purchases are not. The problem could be traffic quality, message match, pricing clarity, checkout friction, or the offer itself.

A funnel turns that vague worry into a specific step. Instead of guessing why revenue is stuck, you can see where qualified people disappear.

The concept

A funnel is a sequence of events that should happen before a customer buys. In Grometrics, that might be landing page, pricing page, signup, checkout started, and revenue.

The point is not to create a perfect model of every path. The point is to find the biggest leak you can actually fix.

Build the path

Choose steps that represent real buying progress. A good funnel should move from interest to evaluation to purchase, not from random pageview to random pageview.

For a SaaS product, that might be landing page, pricing page, signup, onboarding step, checkout, paid. For a course, it might be sales page, curriculum section, checkout, paid.

1. Start with the money event

Revenue is the final step. Work backward from it.

2. Use intent steps

Pricing views, signup starts, checkout starts, and goal events are stronger than generic pageviews.

3. Keep it short

Four to six steps is enough for most decisions. Long funnels create noise.

4. Save one core funnel

Use a stable funnel so week-over-week changes are meaningful.

The walkthrough

Open Conversions and choose the funnel that matches the buying path. Start with the full funnel, then segment by source if the top-level view is hiding a channel-quality problem.

You are looking for the step where enough qualified people disappear that fixing it would move revenue.

Mock Conversions view showing where visitors leave before becoming customers.

1. Find the largest drop

Start with the step where the most qualified people disappear.

2. Segment by source

Check whether one channel is causing the drop-off.

3. Inspect the page

Open the page or event before the drop and decide what to change.

4. Compare to revenue

Prioritize leaks that sit closest to payment or affect the highest-value source.

Interpret the leak

A drop before pricing usually points to message match. The visitor arrived, but the page did not connect their problem to your offer.

A drop after pricing usually points to packaging, trust, urgency, or price clarity. A drop at checkout usually points to friction, payment trust, unexpected cost, or technical failure.

1. Landing to pricing

Clarify the promise, audience, proof, and call to action.

2. Pricing to checkout

Improve plan clarity, guarantees, proof, and objection handling.

3. Checkout to paid

Remove friction, verify payment methods, and check errors or abandoned sessions.

4. Paid to retained

If refunds are high, fix expectation-setting before purchase.

Common mistakes

Do not fix the biggest percentage drop blindly. A 90% drop from 20 visitors may matter less than a 25% drop from 5,000 visitors near checkout.

Do not combine every source into one story when channels behave differently. Paid visitors, referral visitors, and direct returners may need different fixes.

1. Using vanity steps

Scroll depth and time-on-page can be useful context, but they are weak funnel steps.

2. Ignoring source quality

A funnel may look broken because one channel sends low-intent traffic.

3. Changing too much

Fix one step, measure, then move. Multiple changes make the result hard to trust.

Decision framework

If visitors drop before pricing, improve message match. If they drop after pricing, improve offer clarity. If they drop at checkout, reduce friction or fix payment trust.

If one source has a much worse funnel than the others, fix the source or landing path before changing the whole site.

Rule of thumb: fix the largest drop-off closest to revenue first.

Try it in your data

Open Conversions to inspect where visitors drop between interest and purchase.

Start tracking for free →

Related guides

Attribution

8 min read

Where did that sale actually come from?

Trace a sale back to the source, page, and session that created it without guessing from payment data alone.

Channels

8 min read

Find your best channel in 5 minutes

Use revenue per visitor, conversion rate, and source quality to separate channels that pay from channels that look busy.