The conversion funnel is your store seen as a series of doors. A visitor lands, looks at a product, adds it to the cart, starts the checkout, pays. At every door some people turn back. The funnel counts how many go through each one, so you can see exactly where you lose them.

It concerns anyone who sends traffic to a page and wants more of it to become orders. A creator with a link-in-bio shop has a short funnel of two or three steps. A brand with a full storefront has five or six. Either way, the overall conversion rate tells you how many make it to the end; the funnel tells you why the rest did not.

What is a conversion funnel?

A conversion funnel is an ordered list of the actions a visitor must take to buy, with a count of people at each step and a rate between consecutive steps. The standard e-commerce funnel is:

  1. Session. Someone arrives on any page.
  2. Product view. They open a product page.
  3. Add to cart. They put something in the cart.
  4. Checkout started. They open the checkout page.
  5. Order. They pay.

Each step-to-step rate has a name: product view rate, add-to-cart rate, checkout rate, checkout completion rate. Multiply them all and you get the end-to-end conversion rate.

It is not the same as the sales funnel. The sales funnel is the marketing view: how a stranger becomes aware of you, gets interested, considers buying and finally purchases, over days or weeks and across posts, emails and ads. The conversion funnel is the on-site view: what happens between arriving on your store and paying, usually within one session. The sales funnel ends where the conversion funnel begins. A seller needs both, but they are measured with different tools and fixed with different work.

Related terms: drop-off is the share of people lost at a step. *Bounce rate is the drop-off at the very first step, people who leave without a second page. *Cart abandonment is the drop-off between add to cart and order.

Why it matters

A conversion rate of 1.5% is a single number with no instructions attached. A funnel turns it into a diagnosis.

Take two stores, each with 4,000 sessions and 60 orders a month, both at 1.5%. Store A: 2,400 product views (60%), 320 add to carts (13% of views), 200 checkouts (63% of carts), 60 orders (30% of checkouts). Store B: 1,200 product views (30%), 300 add to carts (25% of views), 140 checkouts (47%), 60 orders (43%).

Same rate, opposite problems. Store A has a healthy product page and a checkout that loses seven out of ten people, probably a surprise shipping cost or a forced account. Store B has a fine checkout but 70% of visitors never open a product, which points at the landing page or at traffic sent to the wrong place. Fixing A's checkout to a 45% completion would bring 90 orders; fixing B's landing to a 45% product view rate would bring about 90 too. Working on the wrong step in either store would change nothing.

How it works

  • Define the steps. Use the five standard ones for a storefront. For a link-in-bio shop, three are enough: page view, product tap, order.
  • Count people at each step for the same period. Use unique sessions, not page views, or the numbers will not chain. A month is a good period.
  • Compute the rate between consecutive steps. Divide each step by the previous one. Example: 4,000 sessions, 2,000 product views (50%), 400 add to carts (20%), 180 checkouts (45%), 72 orders (40%). End-to-end: 72 / 4,000 = 1.8%.
  • Find the weakest step. Compare each rate to the ranges below. The step furthest below its range is where the money is.
  • Segment. Run the same funnel for mobile and desktop, for each traffic source, for new and returning visitors. The overall funnel averages away the interesting differences.
  • Fix one step, re-measure, repeat. This is the loop that conversion rate optimization runs.

A worked example: a tea brand sees 3,200 sessions, 1,900 product views (59%), 250 add to carts (13%), 170 checkouts (68%), 55 orders (32%). The add-to-cart rate is at the low end and checkout completion is low. It looks at the product page on a phone: the price sits below three paragraphs, and shipping is only shown at checkout. After moving the price up and adding "Ships in 24h, $4.90" under it, add to cart moves to 18% and checkout completion to 42%. Same traffic, 96 orders.

Benchmarks and examples

Typical step rates for small stores with mixed traffic:

  • Session to product view: 40% to 70%. Below 40%, traffic is landing on the wrong page or the home page is not doing its job.
  • Product view to add to cart: 8% to 15% for cold traffic, 20% to 35% for a warm audience arriving from a post about that product.
  • Add to cart to checkout started: 50% to 70%.
  • Checkout started to order: 35% to 55%. Below 30% points at the checkout itself.
  • End to end: 1% to 3% for a storefront, 3% to 8% for a link-in-bio shop after a post.

Example for a creator: 900 taps on a link in bio after a post about a new preset pack. 620 open the product (69%), 140 add it (23%), 95 start checkout (68%), 58 pay (61%). End to end, 6.4%. A warm funnel like this one is short and steep, and the checkout step is often the only one worth working on.

Common mistakes

  • Chaining numbers from different periods or tools. Sessions from one week and orders from another make a funnel that cannot be read.
  • Using page views instead of unique sessions. One visitor who looks at four products inflates the second step.
  • Reading a funnel built on 40 sessions. Wait for a few hundred sessions and at least 20 orders.
  • Blaming the checkout for every loss. In most small stores, the largest absolute drop is between product view and add to cart.
  • Ignoring the device split. A funnel that looks fine overall can hide a mobile checkout that completes at 20%.
  • Treating the funnel as fixed. A new traffic source changes every rate; re-measure after each change in marketing.

Best practices

  • Measure before you change anything. One month of clean funnel numbers is the baseline every later decision depends on.
  • Land people on the product they came for. The session-to-product-view rate jumps when a post links to the product, not to the catalog.
  • Put price, delivery and variants above the fold on mobile. Most of the product-to-cart loss is people who could not find the answer to one question.
  • Show the total early. Shipping and taxes on the product page or in the cart, never for the first time at checkout.
  • Make the checkout one page with guest checkout. Fewer fields, autofill, the payment methods your buyers use.
  • Follow up on abandoned carts. An email a few hours later recovers a share of the checkout drop.
  • Watch one funnel per channel. The link-in-bio funnel and the storefront funnel have different shapes and different fixes.

In Roctify

Audience analytics track the steps of the funnel for each channel: visitors, product views, add to carts and orders on the link-in-bio page and on the storefront. You can read the rate between steps without exporting anything, and reports on the Pro plan let you compare periods to see whether a fix moved the step you targeted.

The steps most often at fault are already built to leak less: product pages show variants, stock and shipping, the checkout is one page with guest checkout, and Stripe, PayPal and cash on delivery are available at 0% transaction fees on every plan.

FAQ

What is the difference between the conversion funnel and the sales funnel?

The sales funnel covers the journey from stranger to customer across all your marketing: posts, emails, ads, landing pages. The conversion funnel covers what happens on the store, from arriving to paying. The sales funnel brings people to the top of the conversion funnel; the conversion funnel turns them into orders. Both are worth measuring, with different tools.

How many steps should my funnel have?

As many as there are real actions a visitor must take, and no more. A storefront usually needs five. A link-in-bio shop with a built-in checkout needs three. Adding steps that are not actions (scrolling, time on page) makes the funnel harder to read without making it more useful.

Which step should I fix first?

The one with the largest absolute loss, weighted by how easy it is to fix. In most small stores that is product view to add to cart, followed by checkout completion. Fix the obvious causes (hidden shipping, unclear price, forced account) before testing anything subtle.