Spot funnel leaks and call split tests in ClickFunnels

Build me an app I open every Monday to find where my ClickFunnels funnels leak money and to decide which split tests to call. It has two main views plus a weekly snapshot that gives me history ClickFunnels does not keep.

View one is the funnel leak view. Use List Funnels to fill a funnel picker down the left, and use Fetch Funnel for the selected funnel's name and details. For the selected funnel call Fetch Funnel Stats and request the per-step breakdown, which gives views, opt-ins, sales and earnings for each step. Draw a step by step drop-off chart in funnel order: each step is a bar sized by views, with the drop between consecutive steps called out as both a count and a percentage. Under the chart put a table with one row per step showing views, opt-ins, sales, earnings and the view-to-conversion ratio.

Build the step ordering from the Fetch Funnel Stats per-step breakdown combined with List Pages, which includes funnel membership, so you can match each step to its page and title. Important feasibility note: do NOT use Fetch Funnel Structure. It is marked Closed Beta and fails unless the workspace has been granted access. Derive ordering from the stats breakdown and page membership instead.

Highlight the single worst step by view-to-conversion ratio, calling it out visually in both the chart and the table. Next to it, estimate the revenue I would recover by fixing just that step: take the step's current conversion ratio, compare it against a healthier benchmark from the same funnel such as the median step ratio, and show how much additional earnings would flow through the rest of the funnel if that one step converted at the benchmark rate. Show the arithmetic in plain language underneath the number so I can sanity check it, and be honest that it is an estimate rather than a promise.

View two is a split test scoreboard for the selected funnel. Call List Split Test Steps to find every split test in the funnel, then Fetch Split Test Step for each one to get its two variants and their traffic weights. For each variant page call Fetch Page Stats to get that variant's own views, opt-ins, sales and earnings. Render each test as a card with the two variants side by side, showing the traffic weight and the individual numbers for each, plus conversion rate and earnings per visitor.

The scoreboard must not declare a winner until both variants have cleared a minimum visitor threshold that I set in the app. Store my threshold and default it to something sensible like 300 visitors per variant. Below the threshold, show the test as still collecting and state how many more visitors each variant needs. Once both variants clear it, name the leading variant, show the size of the gap, and note if the two are close enough to be effectively tied. The whole point is that I stop calling tests early on a lucky lead.

The Google Sheets snapshot is load-bearing, not decoration. ClickFunnels stats endpoints return point-in-time totals with no history, so trends are impossible without persisting my own. Once a week, and also on demand from a Save this week's snapshot button, use Append Values to write one row per funnel step into a Google Sheet: snapshot date, funnel name and id, step name, views, opt-ins, sales, earnings. On load, use Get Values to read the sheet back and show week over week trend arrows next to every step and every headline metric, with the change since the last snapshot on hover. Do not write a second snapshot for a week that already has one, and if the sheet is empty explain that this visit is establishing the baseline and arrows appear next week.

Let me act without leaving the page: from the highlighted worst step, a button launches a new split test on it using Create Split Test Step, wrapping that step's existing page. Confirm which page is about to be wrapped before creating anything, then refresh the scoreboard so the new test appears as still collecting.

Keep this a diagnosis and decision tool rather than a revenue summary. Handle funnels with little or no traffic gracefully by saying there is not enough data yet instead of showing misleading ratios, format money and percentages consistently, and make the whole thing readable at a glance on a Monday morning.