[object Object]

Funnels in Freshmarketer track users through a defined sequence of events. Built right, they expose where the business loses people. Built wrong, they produce a chart with a 0.5 percent end-to-end conversion that nobody can act on.

Define the funnel before building

Three constraints:

  • 4 to 7 steps. Below 4, you are not learning much. Above 7, drop-off compounds and the chart becomes uninteresting.
  • Each step represents a user decision, not a page view. “Viewed pricing” is a step; “Viewed pricing for 3 seconds” is not.
  • Steps must be sequential and ordered. If users can do them in any order, build separate funnels.

Common funnel: visited site, viewed product, added to cart, started checkout, completed purchase.

Time window

Funnels measure conversion within a time window. Default 7 days; configure based on your sales cycle.

For B2B (long consideration), 30 to 90 days. For e-commerce (impulse), 1 to 7 days. The window starts at the first step; users who complete subsequent steps outside the window do not count.

If your time window is too short, you understate conversion. Too long, you include noise (unrelated returns to your site).

Step matching

Each step is defined by an event or page URL. Be specific:

  • Page URL: /pricing matches /pricing only, not /pricing/enterprise. Use wildcards if needed.
  • Event: a tracked event (e.g., “Add to Cart”) with optional parameters.

Vague step definitions match too many users; precise ones match exactly your intended behavior.

Segment the funnel

Single-segment funnels hide useful patterns. Always overlay segments:

  • Traffic source (paid vs organic vs direct).
  • Device (mobile vs desktop).
  • New vs returning.
  • Geographic region.

If conversion rate from paid traffic is 2 percent and organic is 8 percent, your problem is not the funnel; it is the paid acquisition mismatch.

Drop-off analysis

The visual funnel shows percent dropping at each step. The biggest drop is your biggest leverage point.

Investigate before assuming. A big drop between “viewed product” and “added to cart” might be:

  • Product page weak (fix copy or images).
  • Pricing surprise (expose pricing earlier).
  • Inventory issues (out-of-stock items kill conversion).

Each requires a different intervention.

Conversion vs completion

Funnel conversion is users who complete all steps. But users may complete the same goal via a different path. A user who skips “view product” by going straight from search to “add to cart” is not in your funnel but is converting.

Build complementary funnels for alternate paths. Compare conversion rates; the path with higher conversion is your golden path.

Comparing funnels over time

Track funnel performance week over week. A degrading conversion at a specific step usually points to:

  • Recent UI change that broke something.
  • Tracking change that miscounts.
  • Traffic mix shift (more low-intent traffic).

Investigate degradation immediately; recovery is easier within days of the change.

Reverse funnels

Sometimes start from the conversion and work backwards. Of users who purchased, what was their path? Of users who churned, what was their last action?

Reverse funnels reveal hidden paths to success. Forward funnels reveal where you lose people. Use both.

Reporting cadence

Funnels are not daily metrics. Review weekly for trend, monthly for deep analysis. Daily noise drowns the signal.

Surface the top three funnels on a weekly executive report with a one-line interpretation, not just the chart.

What to do this week

Pick your most important conversion (signup, purchase, demo request). Define the funnel as 5 steps. Build it in Freshmarketer with one segment overlay (say, traffic source). Identify the largest drop-off and form one hypothesis to test.

[object Object]
Share