Funnel comparison: every affiliate's funnel side by side

A single funnel tells you how your program converts. A hundred funnels lined up next to each other tell you something far more actionable: which partners actually carry clicks all the way through, and which ones stall. Two affiliates can send identical click volumes and look equal on a top-line report, while one converts at every stage and the other leaks visitors at the first step. Averaged together, they hide each other. Funnel comparison pulls them apart. This guide explains how a funnel matrix works, why the metric you view it in changes what you learn, and how timing adds a dimension raw counts can't.

Why one funnel isn't enough

Most reporting shows you the funnel for a whole program, or for one entity at a time. That's useful for understanding overall shape — how many clicks become leads, how many leads become sales — but it averages away the thing you most need to see. Your program-wide funnel is the blend of your best and worst partners, and the blend is nobody's actual behavior. To act, you need to compare: to see that affiliate A converts 8% of clicks to sales while affiliate B, on the same offer, converts 1% — and to see where B loses them.

Comparison is what turns a funnel from a description into a decision. Once you can see partners side by side, questions answer themselves: who deserves a higher payout, who needs their traffic examined, which offer converts a given channel best, where in the journey your money is leaking. It's the analytical move behind most real optimization, and it builds directly on the attribution chain described in the attribution guide — every stage in the funnel is an attributed event.

Reading the matrix

A funnel comparison lays out as a matrix. Each row is one entity — an affiliate, an offer, a campaign, or a channel — and each column is a stage of the journey, in the order the funnel actually flows. The first data column is clicks: the width at the top of the funnel, the starting point every later stage is measured against. Then comes one column per stage — a lead, a registration, a sale — each marked as a conversion or an event depending on how it's counted (a distinction the reports guide explains in full).

A small bar under each value compares that cell against the largest in its column, so strength jumps out visually: scan down a stage's column and the strongest performers are obvious without reading a single number. That visual layer is what makes a matrix scannable rather than just a spreadsheet — you see the pattern before you read the figures.

Choosing what to compare — and against what

Two controls define any comparison. The first is the dimension: are you comparing affiliates, offers, campaigns, or channels? The second is the scope: within what? You might compare every affiliate within a single offer — to see which partners convert that specific product best — or every offer within a single affiliate — to see which of your products a given partner is best at selling. The dimension and the scope together frame the exact question you're asking.

By default a comparison shows the top rows by click volume, which surfaces the entities that matter most. But you can also hand-pick exactly the rows you want lined up — and a hand-picked entity always appears even if it sent zero clicks, which is quietly powerful: it lets you prove a negative. When a partner claims they've been driving traffic and the numbers say otherwise, putting them explicitly in the matrix with a row of zeros ends the debate.

The metric is the message

Here's the part that separates a casual look from real analysis: the same matrix tells you completely different things depending on which metric fills the cells, and switching between them is the core skill.

Count shows how many reached each stage — the raw volume. It's the honest picture of scale, but it favors big partners simply because they're big.

Percent of clicks shows each stage as a share of that row's own clicks. This normalizes for size, so a small partner and a large one compare fairly. It's the view that answers "who converts best," independent of who sends the most.

Step percent is the sharpest diagnostic. It shows each stage as a share of the previous stage — the conversion from one step to the next. This is what pinpoints exactly where a funnel drops off. A partner might look fine on clicks-to-sale overall but, in step percent, reveal that they lose 90% of visitors between the landing page and the lead form — a specific, fixable problem you'd never see in the aggregate. When you want to know not just that a funnel underperforms but where, step percent is the answer.

There are also revenue and payout views, which fill the cells with money recorded at each stage — useful when the question is economic rather than behavioral.

Adding time to the picture

Counts and shares tell you how many convert; they say nothing about how fast. Turning on timing adds each stage's median and p90 time from click to that event — the typical delay, and the delay for the slower tail. This matters because speed is signal. Two partners might convert the same share of clicks, but one does it in an hour and the other takes two weeks — which tells you something real about their traffic quality and their audience's intent. Timing is often the tiebreaker between partners who look identical on conversion rate alone, and it's the kind of nuance that separating traffic by sub-ID can sharpen even further.

From comparison to action

A comparison is only useful if it leads somewhere, so any row drills through: click an entity and you drop into its full single funnel and every other report panel, scoped to that entity and date range. The matrix tells you which partner to investigate; the drill-through tells you why. And the whole matrix exports, so a comparison can become the backbone of a partner review, a payout-tier decision, or a monthly optimization ritual.

Every number in a comparison follows the same rules as the rest of your reporting — attributed clicks only, your own event catalog, your account's timezone — so what you see here reconciles with what you see everywhere else. There's no separate, contradictory version of the truth.

A worked example: two partners who look identical

Picture two affiliates on the same offer, both sending 10,000 clicks in a month, both producing 100 sales. On any top-line report they're twins — same volume, same 1% click-to-sale rate, same revenue. A payout decision based on that report treats them as equals. Now line their funnels up in step percent, and the story splits. Partner A converts 40% of clicks to a lead, then 25% of leads to a sale — a leaky top but a strong close. Partner B converts 8% of clicks to a lead, then 12.5% of leads to a sale — they're barely getting anyone into the funnel, but the few who enter convert hard.

That single view changes what you'd do with each. Partner A's problem is the landing step, which might be a creative or a targeting fix you can coach. Partner B's problem is traffic volume at the top, or a mismatch between their audience and the offer — a different conversation entirely. Add timing and it sharpens again: if A's leads convert in an hour and B's take two weeks, A's traffic is hotter even where the rates match. None of this is visible in the aggregate; all of it is obvious the moment the funnels sit side by side. This is why experienced operators reach for comparison before they touch a payout — and why pairing it with sub-ID analysis, to see which of a partner's sources drives the difference, is where the deepest optimization happens.

Why side-by-side wins

The deepest reason to compare funnels rather than read them one at a time is that optimization is inherently relative. "Good" only means anything against "less good." A partner converting 3% is excellent if the field converts 1% and poor if it converts 8% — and you can't know which until you line them up. Funnel comparison makes that relativity visible, at every stage, in whichever metric fits the question.

LimeliJourney treats the funnel as a first-class comparison surface, part of the broader reporting feature set, precisely because side-by-side is where the decisions live. If you want to see your own partners lined up — and watch step-percent reveal exactly where each one leaks — a demo will run the comparison on your real traffic with you.