The store’s weekly report looks comfortable. Products are getting views, carts are being started, and orders are coming in. Then a marketer checks the visitors who arrived from a paid campaign on their phones. Many of them add a product to the cart, but few seem to reach the payment page. The storewide numbers did not show that narrow break in the journey.<\/p>\n
There are several possible explanations. The cart might be awkward on a small screen. Shipping costs might appear sooner than visitors expect. The payment page might have a tracking gap. Some people might simply be comparing products. A chart alone cannot choose among those stories. It can, however, show where to look first.<\/p>\n
That is the useful job of a conversion Funnel. Put meaningful visitor signals in the order they should occur, see how far the recorded journey progresses, then open the people and pages behind a weak stage. This guide shows how to build that view in Hitsteps using Segments, and how to turn a drop-off into a careful next step.<\/p>\n
Pageview totals count attention to pages. A single conversion rate compresses an entire journey into an entry and an outcome. Both can be useful, but neither tells you whether the trouble is before cart, between cart and payment, or after payment. A busy payment page can also look encouraging even if the campaign visitors you care about rarely reach it. Other visitors may be supplying most of that traffic.<\/p>\n
Imagine the store above has a stable overall order rate while a new mobile campaign is running. If most existing customers complete checkout and campaign visitors stall earlier, the blended rate can stay steady. Looking only at the overall number could lead the team to buy more of the same traffic or redesign the wrong page. Looking only at cart pageviews could create a second mistake: a page can be viewed repeatedly by a smaller group, and a pageview total does not show how many tracked journeys continued to payment.<\/p>\n
The first useful question is narrower: Among the campaign visitors the site records, which milestone is the first one they often do not reach?<\/strong> A Funnel gives that question an order. A Segment gives it an audience. Together they make a specific investigation possible, such as checking the mobile cart experience for paid visitors instead of treating the whole checkout as broken.<\/p>\n Treat that finding as a location for investigation, not as a diagnosis. If people match the cart stage and do not match the payment stage, you know the expected next signal was not recorded in the measured path. You do not yet know whether a visitor left, changed devices, skipped a step, encountered an error, or completed an action your tracking missed.<\/p>\n A Segment<\/a> answers a grouping question: which tracked visitors match a defined set of conditions? You might define visitors from a campaign, people who performed an Add to cart Action, or visitors who reached a particular thank-you page. The conditions can use supported signals such as source, campaign, device, page activity, and Actions. They are not limited to URLs.<\/p>\n A Hitsteps Funnel<\/a> places those Segment matches in an ordered chain. It measures entries, progression through the stages, drop-offs, and completion of the chain. A Funnel needs at least two stages. Its first stage can be Everyone<\/strong> when you want a sitewide entry point, or a Segment when a specific audience is the real starting group. Everyone is the broad entry option, not a stage to insert halfway through a journey.<\/p>\n Think of each stage as a question about evidence. Did a tracked visitor enter the relevant audience? Was an Add to cart Action recorded? Did that visitor then reach the payment page? Did the final thank-you signal arrive? The Funnel compares recorded progress through those questions. It does not inspect someone’s thoughts or establish that a page caused the next action.<\/p>\n The quality of the Funnel depends on the quality of its stages. A Segment called “Qualified lead” is useful only if its conditions represent a signal your site actually captures, such as a completed form with the qualification detail you need. A name alone does not make a lead qualified. Before arranging stages, check the page paths, campaign tags, and Action values that each one relies on. If a payment step uses several URLs, decide which recorded signal reliably marks arrival there. If a thank-you page can be reached without completing the intended action, do not silently call it a sale.<\/p>\n Use exact order<\/strong> for a process where each recorded milestone must appear in sequence. A simple checkout question might be Add to cart, then Payment page, then Thank you. A path that reaches Add to cart and Thank you without a recorded Payment page match will not satisfy that exact chain. That is appropriate if payment-page arrival is a required part of the question and the tracking is reliable.<\/p>\n Exact order is also a useful check on your setup. If people clearly place orders but the Funnel shows almost no progress past Payment page, test the payment-page tracking before changing cart copy. Some checkouts move across domains or use a hosted flow. The Funnel can only use signals that reach it. Its strictness makes a missing signal visible, but it cannot tell whether the missing signal reflects behavior or instrumentation.<\/p>\n Use forward order<\/strong> when a visitor may reasonably skip an optional stage and still reach the outcome. For a content journey, consider Campaign visitor, then Product guide, then Pricing, then Signup complete. Some visitors may go straight from the campaign page to pricing and signup. Forward order lets that journey keep moving toward the final stage even when the guide stage is skipped. It still respects the direction of the chain; an earlier stage recorded later does not turn a backward path into forward progress.<\/p>\n Choose the rule to match the business question, not to make the completion figure look larger. If the product guide is genuinely required for a regulated or assisted process, exact order is the better definition. If it is optional reading, forward order gives a truer picture of successful paths. Write the choice down so a later comparison uses the same rule.<\/p>\n These recipes are starting points, not promises that every site records the same events. Give each stage a clear source of evidence, verify that a real visit can match it, and name the decision you would make after reading the result.<\/p>\n Chain:<\/strong> Everyone, then Add to cart Segment, then Payment page Segment, then Thank you Segment.<\/p>\n Start with Everyone if you want the broad shape of the site’s purchase journey. Define Add to cart from the cart Action or other recorded signal your store actually sends. Define Payment page from the tracked checkout step that visitors should reach before completing payment. Define Thank you from a completion signal that reliably corresponds to the intended outcome. Use exact order when those are required milestones in the flow you are testing.<\/p>\n Why it matters:<\/strong> The space between adjacent stages can suggest a very different next task. If visitors rarely add anything, examine product pages, availability, and the path to cart. If carts appear but payment-page matches thin out, inspect the cart and checkout handoff. If payment-page matches are healthy but the completion stage is weak, check the payment experience and the final signal. Treat a thank-you page as an observed completion marker only when it truly represents a completed order in your setup.<\/p>\n What to investigate:<\/strong> Open the Add to cart stage, look at available visitor paths and device context, and test the cart-to-payment route yourself on a phone. Check whether the payment page is tracked across any redirect. A click heatmap on a weak cart page can show aggregate click patterns around the checkout control, while the Funnel shows where recorded journeys thin out. Keep those reports in their proper roles: a general page heatmap does not automatically isolate your Funnel audience.<\/p>\n Everyone is useful for a broad view, but a storewide entry mixes shoppers with people who came to read or browse. For the opening paid-mobile question, make a second, focused Funnel with a paid mobile campaign Segment as the entry when that audience is the decision. Keep the same downstream milestones. Compare like with like across the same period before deciding that a campaign needs a new landing page or the cart needs work.<\/p>\n Chain:<\/strong> Referral or campaign Segment, then key Action Segment, then qualified lead or thank-you Segment.<\/p>\n Suppose a service business sends partner referrals to a page explaining a consultation. The entry Segment identifies visitors from that referral source or a campaign tag the site records. The middle stage represents a meaningful Action, such as opening the booking form or selecting a service, if that Action is actually tracked. The final stage should represent the completed enquiry or a recorded qualification signal, such as a submitted form that carries the information needed for follow-up.<\/p>\n Why it matters:<\/strong> A source report can say that referrals arrived, and a form report can say that enquiries arrived. The Funnel asks whether the tracked referral journeys passed through the intended action and reached the final outcome. If the key Action receives little activity, inspect the landing page promise and the call to action. If the Action is common but completed enquiries are scarce, test the form itself, including its mobile layout, required fields, error messages, and success signal.<\/p>\n Be precise about “qualified.” A generic thank-you page proves only that its tracked condition was met. If qualification is decided later by a person or a separate system and is not sent to Hitsteps, call the final stage a submitted enquiry instead. If the booking form can be opened through several routes, forward order may be sensible for optional middle steps. If the key Action is a required gate, use exact order and check the instrumentation first.<\/p>\n Chain:<\/strong> Campaign visitors, then product engagement Segment, then qualified Segment, then conversion Segment.<\/p>\n Imagine a campaign that promotes a particular product line. Its entry Segment uses the campaign identifier, not all paid visitors. Product engagement might be a recorded product interaction or a visit to a relevant detail page. Qualification needs its own observable definition, perhaps an explicit plan-selection Action or a request for the product category that makes a lead useful to the team. The final conversion is the specific signup, enquiry, or completed action the campaign is meant to produce.<\/p>\n Why it matters:<\/strong> A campaign can deliver many arrivals but few visitors who show interest in the promoted product. That points you toward the ad promise, audience, or landing page. Another campaign may bring fewer arrivals but a larger share reaching the qualification stage. That suggests a different decision from optimizing a generic visit count. The Funnel makes each transition visible so the team can discuss traffic quality in terms of recorded behavior rather than calling every visitor “high intent.”<\/p>\n Use forward order if product engagement is a helpful but optional sign of interest. A visitor who skips that page yet takes the qualification Action can still count as moving toward conversion. If qualification requires the engagement event by definition, keep the stages and rule consistent with that requirement. Inspect actual paths from both campaigns before changing spend. A small difference in stage counts can also reflect tagging mistakes or different landing routes, so confirm the inputs before treating it as a campaign verdict.<\/p>\n Start with the stage counts, but read more than the narrowest shape. Look at entries, each stage’s progression from the previous stage, and the final completion. A large gap between two stages tells you where a recorded path thinned out. The distance from entry to completion tells you how demanding the full chain is. For a broad Everyone entry, do not expect every visitor to want the final outcome. For a campaign entry, ask whether the campaign audience and date range match the decision you are making.<\/p>\n Next, open the stage just before the gap. In Hitsteps, selecting a Segment stage opens the visitor report for that Segment; Everyone opens the unfiltered visitor report. Review a few available paths and their source, device, and page context. Look for patterns shared by several visitors, including paths that did continue. The stage link takes you to matching visitor context, so do not assume every row in that report completed the entire Funnel. Use the ordered stage result and the visitor details together.<\/p>\n Now check the signal on the far side of the gap. If cart visitors seem not to reach payment, make a real test visit. Does the payment page load under the tracked site? Is the event sent when the page appears? Does the visitor move through a step you did not include? A missing or badly named Segment can produce a dramatic drop-off. Fix measurement before writing a story about customer reluctance.<\/p>\n When the tracking holds up, choose a page to examine. Compare the promise that brought the audience in with the page’s main message and call to action. Use a click heatmap<\/a> to see aggregate interaction on a weak page, and check live visitor context<\/a> when you need to understand what is happening now. Test the page on the devices represented in the visitor report. A heatmap or a live path adds evidence; neither proves the visitor’s motive.<\/p>\n Finish with one written hypothesis and one small change. For example: “Paid mobile visitors reach cart but often do not reach payment; we will test a clearer checkout control and verify the payment page signal.” Note the date, the Funnel definition, and the campaign conditions. When new data arrives, compare a similar audience and period. If you change the offer, traffic source, and checkout at once, you will have a harder time explaining any movement in the Funnel.<\/p>\n The Funnel preview helps you check a draft chain using up to two days of genuine live Segment match history. It is a recent sample, so a thin preview does not necessarily mean the journey never happens. Confirm the selected stages and their recorded signals before saving, especially if a Segment was created only recently.<\/p>\n After you save, Hitsteps prepares eligible recent Funnel history in the background and shows progress. That saved preparation can use recorded Segment matches that already existed, subject to available and retained data. A new Segment begins recording live matches only after it is created. Its analytical report may later reconstruct some earlier matches from retained data, but those reconstructed matches never advance a Funnel or fill its past stages. Do not treat an analytical Segment rebuild as a retroactive customer journey.<\/p>\n The saved view may therefore cover less history than a pageview report or a Segment’s analytical chart. A site cannot recover an Action it never sent, a page that was never tracked, or data that was not retained. Wait for preparation to finish before comparing the saved Funnel with its early preview, and record when you created or materially changed its stages.<\/p>\n If you enable completion notifications, they relate to a recorded arrival at the Funnel’s final step. They are not alerts for every stage or a way to send notifications for old reconstructed matches. Use them when a completed enquiry or other final action merits a timely follow-up.<\/p>\n Finally, a tracked visitor is not automatically a known person. A phone visit and a laptop visit should not be joined in your interpretation unless your tracking actually links them. This is particularly important for lead journeys that begin with a referral and finish after a return visit. Describe what the recorded path supports, and keep cross-device assumptions out of the conclusion.<\/p>\n Choosing stages because their names sound right.<\/strong> “Payment page” may be an old URL, a test page, or a path that no longer appears in the live checkout. Review the Segment definition and make a known visit before relying on its count. A stage that does not match real tracking is a measurement problem first.<\/p>\n Treating a drop-off as proof of abandonment.<\/strong> A visitor who does not appear at the next stage may have stopped, skipped an optional page, switched devices, or encountered a tracking gap. Say “did not reach the recorded stage in this path” until you have stronger evidence. The Funnel points to a transition worth checking, not to a proven reason for leaving.<\/p>\n Using exact order on an optional path.<\/strong> If visitors can legitimately bypass a guide or comparison page, exact order excludes valid completions from the chain you defined. Decide whether the intermediate step is required. Choose forward order for optional steps and keep the final outcome meaningful.<\/p>\n Assuming a Segment’s reconstructed history fills the Funnel.<\/strong> Earlier analytical matches help you review eligible retained activity. They do not create past progression. If the Funnel appears sparse just after new Segment stages were created, allow future live matches to accumulate and read the saved history status accurately.<\/p>\n Comparing unrelated audiences or dates.<\/strong> A broad Everyone entry, a narrow campaign entry, and a holiday promotion answer different questions. Keep the entry definition, order rule, tracking setup, and time period visible when discussing a change. Check whether a campaign tag or checkout flow changed during the comparison.<\/p>\n Reading a visitor path as a complete identity record.<\/strong> One browser path can provide useful context, but it does not automatically include a person’s activity on every device. Look for repeated patterns in the recorded group, and be modest about what one path can explain.<\/p>\n You do not need a catalog of Funnels to begin. Pick one journey with an outcome you can verify and a decision you could make after seeing the weak step.<\/p>\n 1. Name the question.<\/strong> “Where do paid mobile shoppers stop between cart and checkout completion?” is more useful than “How is ecommerce doing?” Write down the audience and the final signal you care about.<\/p>\n 2. Create or check the Segment stages.<\/strong> Start with the smallest set of recorded milestones that answers that question. Verify the source tag, page URL, or Action behind each Segment. If a stage cannot match a real visit, fix it before arranging the chain.<\/p>\n 3. Choose the entry and order.<\/strong> Use Everyone for a broad site journey or an entry Segment for a focused audience. Choose exact order when every stage is required; choose forward order when a middle stage may be skipped. Keep the final stage tied to the outcome.<\/p>\n 4. Preview and reality-check.<\/strong> Read the recent sample, remembering that it uses up to two days of live Segment history. If a stage unexpectedly has no matches, check tracking, date, and Segment definition before concluding that visitors never go there.<\/p>\n 5. Save and let history prepare.<\/strong> Check the preparation status before treating the saved report as complete. Keep the distinction between existing live match history and reconstructed analytical Segment matches in mind.<\/p>\n 6. Open the weakest transition.<\/strong> Read adjacent stage counts, open the Segment visitors behind the earlier stage, and inspect available paths. Test the page and next signal yourself. Use a heatmap or live visitor view when it can answer a specific follow-up question.<\/p>\n 7. Make one small decision.<\/strong> Improve a confusing call to action, repair a missing tracking signal, test a form, or wait for more ordinary traffic. Record what you changed and why, then return to the same Funnel definition for the next review.<\/p>\nWhat a Hitsteps Funnel measures<\/h2>\n
Exact order when every stage is required<\/h3>\n
Forward order when the middle can vary<\/h3>\n
Three Funnels worth building<\/h2>\n
1. Ecommerce: from cart to a completed checkout<\/h3>\n
2. Lead generation: from source to a meaningful enquiry<\/h3>\n
3. Campaign quality: from arrival to evidence of fit<\/h3>\n
Read the Funnel from the weak transition outward<\/h2>\n
Know what the history can and cannot show<\/h2>\n
Common mistakes that lead to the wrong fix<\/h2>\n
Build the first useful Funnel this week<\/h2>\n
Where Hitsteps fits<\/h2>\n