The ads are doing their job. The mobile landing pages throw the click away before it reaches the button.
The campaign bought 22.7M impressions and 644,000 landing-page views for ~$301K at a strong click-through rate — cheap, high-volume, overwhelmingly mobile traffic. Then it stalls: 99% of visitors are on mobile, median time on page is ~4.7 seconds, and the "Start My LLC" CTA ranks #3–#7 on mobile (0–1% of on-page clicks). What gets clicked instead is the cookie banner, the nav header, and the logo — the exits. Users leave before they engage the offer.
The mechanism, in one line
You're paying for reach and getting it — but the mobile lander surrounds a buried CTA with escape hatches (nav, logo, cookie prompt) and asks nothing clearly in the ~5 seconds a paid clicker gives it. The leak is on the page, on mobile, and almost entirely fixable on Bizee's side of the click.
What to do about it — in priority order
Strip the escape hatches off the mobile lander.
Remove the full nav header, unlink the Bizee logo, delay the cookie banner — the current top-three mobile interactions are all exits or dismissals, not the offer. click-map competitorMake "Start My LLC" impossible to miss on mobile — and keep it on screen.
Promote it above the fold and add a sticky mobile CTA bar so the primary action follows the user instead of sitting at rank #6. click-map benchmarkBuild genuinely mobile-first landers.
99% of traffic is mobile yet mobile converts 5–7× worse — the experience is built for the 1%. Design for the thumb, the 5-second read, the single decision. click-mapFix Trades' load performance and the rendering bug.
Trades' 2.52s median signals a load problem; the click data also flags anexcludedUrlsrender fault — both bleed the most-funded audience. click-map
The media overdelivers: cheap views at scale, evenly across audiences
Straight from the ad report. Every audience got ~$56K and returned hundreds of thousands of views at well under $1 each. There is no traffic problem — Side-Hustler alone bought 164K views at $0.34.
| Audience ad set | Impressions | Landing-page views | Spend | Cost / view |
|---|---|---|---|---|
| Side Hustler | 3,340,504 | 164,394 | $56,294 | $0.34 |
| Trades | 5,167,468 | 128,829 | $56,483 | $0.44 |
| Creator | 4,049,027 | 105,941 | $56,420 | $0.53 |
| Freelancer | 4,656,210 | 92,435 | $56,447 | $0.61 |
| Retargeting (S1–S4) | 2,986,760 | 87,853 | $45,331 | $0.52 |
| Brand / Control | 2,504,634 | 64,866 | $30,076 | $0.46 |
| Total | 22,704,603 | 644,318 | $301,049 | $0.47 |
Reporting window 2026-05-28 → 06-26. Campaign optimized for Landing Page Views — Meta is told to buy cheap views, not likely converters. ad report
The ad half overdelivers. The on-page half falls off a cliff.
Bar widths show funnel shape; the labelled figures are measured. The ad→view stages are healthy; view→action is where ~$301K of reach stops converting.
On mobile, the page's own furniture out-clicks the offer
A representative mobile lander's click distribution. The pattern repeats: the cookie banner, the nav header, and the logo — every one an exit — beat the "Start My LLC" button, which sits at rank #6.
Representative mobile lander (Controlled / Trades pattern). Trades is worse — the nav header alone takes 26% of clicks (923 users navigating away). click-map
Per-lander mobile performance
| Lander (mobile) | Pageviews | Median time | Click rate | Signal |
|---|---|---|---|---|
| Freelancer | 67,500 | 6.28s | 6.1% | Best of a weak set; strongest scent |
| Creator | 125,800 | 5.94s | 5.1% | Best CTA position; no cookie banner in top clicks |
| Controlled | 47,700 | 4.46s | 3.8% | Generic copy; cookie banner is the #1 click |
| Side Hustle | 164,800 | 5.02s | 3.2% | Most traffic, near-worst engagement |
| Trades | 108,000 | 2.52s | 3.0% | Worst across all 10 data points — load/perf suspect |
The number that reframes everything: 99% mobile, ~4.7s
Desktop looks fine (20–37% click rates) but is <1% of traffic and too small to be significant. This is a mobile campaign, and on mobile users give the page ~5 seconds. Any fix that isn't mobile-first is optimizing the 1%.
Why it leaks — the root causes, from the data
1. Escape hatches beat the offer click-map
Nav header, linked logo, and cookie banner are the top mobile interactions — each a way out. Before a user weighs the CTA, the page hands them three exits.
2. The CTA is buried on mobile click-map
"Start My LLC" ranks #3–#7 and takes 0–1% of clicks. The most important action isn't where the thumb is, and there's no sticky CTA to recover it.
3. Built for desktop, used on mobile click-map
99% mobile, ~4.7s dwell, 5–7× worse than desktop — these read as desktop pages shrunk to a phone, not landers built for a 5-second mobile decision.
4. Trades has a perf problem click-map
2.52s median (users leaving before render) + a flagged excludedUrls bug — on the audience with the most impressions (5.2M).
5. Generic scent on weak landers competitor
"Controlled" ("Do Business. Officially.") and Side Hustle underperform; testimonials get zero engagement on generic pages but resonate for Trades — segment-specific proof matters.
6. Optimized for views, not outcomes ad report
The campaign result type is Landing Page Views — Meta buys cheap views, not likely converters. Cheap traffic + a leaky page compounds the loss.
The incumbents point the same click at a funnel, not a page full of exits
Full teardowns in the shared folder. The relevant contrast: ZenBusiness and LegalZoom's paid landers are funnel-first with minimal chrome — a stripped header, an above-the-fold offer, a lead captured before price. The opposite of a lander whose top clicks are the nav and the logo.
/free-llcBizee's own core page already puts "Start My Business" + Trustpilot above the fold — the bar the audience landers should meet. audited
Bizee's wedge: both incumbents bury auto-renewing upsells behind a "$0" headline and breed billing-surprise resentment — a transparent, no-surprise Bizee lander both converts better and anchors the future comparison page.
Prioritized recommendations (ICE)
Scored Impact × Confidence ÷ Effort, 1–10 each (higher = do first). Confidence is high — these come from the client's own click data, corroborated by competitor practice and benchmarks. First block is the client analysis's own "Now/Soon" list, sharpened; second is the independent design-panel's cross-cutting quick wins.
| Fix | Where | I | C | E | Pri | Evidence |
|---|---|---|---|---|---|---|
| Unlink the Bizee logo so it stops being an exit | all landers | 5 | 9 | 1 | 45 | click-map |
| Remove the full nav header from paid landers (kill the #1–2 mobile exit) | all landers | 8 | 9 | 2 | 36 | click-map |
| Surface "Start My LLC" above the fold + sticky mobile CTA bar | all landers | 9 | 8 | 3 | 24 | click-map benchmark |
| Delay the OneTrust cookie banner (stop it intercepting the first tap) | all landers | 6 | 8 | 2 | 24 | click-map |
Fix Trades load + the excludedUrls render bug | Trades (+ all) | 7 | 7 | 3 | 16 | click-map |
| Build mobile-first lander variants (single column, thumb-zone CTA, 5-sec read) | all landers | 9 | 8 | 6 | 12 | click-map benchmark |
| Optimize campaigns for a conversion event, not Landing Page Views (once the page converts) | Meta setup | 7 | 6 | 4 | 11 | ad report benchmark |
| Segment-specific proof + scent (kill generic "Controlled" copy) | Controlled, Side Hustle | 6 | 6 | 4 | 9 | click-map |
| Independent design-panel quick wins (Bizee's public conversion pages) | ||||||
| Tame the floating chat bubble + fix its sub-44px, low-contrast close "X" | site-wide | 5 | 8 | 2 | 20 | panel |
| Raise secondary-text contrast to WCAG AA (metadata, brand subtext, Trustpilot badge) | site-wide | 4 | 8 | 2 | 16 | panel |
Show the "$0" price + resolve competing CTAs on /free-llc + homepage | core funnel | 6 | 7 | 3 | 14 | panel audited |
Unlink-the-logo tops pure ICE (near-zero effort, real exit reduction), but the compounding win is the pair up top: remove the exits and surface the CTA — together they change what the page is for on mobile.
Independent design panel — a scoping note
The paid landers are unlisted (noindex), so a three-lens design panel (Linear designer · accessibility · data-density, run as vision models) reviewed Bizee's public conversion pages as the closest proxy. Two things carry to the campaign landers (same site chrome): the floating chat bubble obscuring the CTA, and sub-AA contrast on secondary text. Its verdict is a proxy, not the primary evidence — the click-map above is authoritative; the panel reinforces it.
What the leak is worth
The media is efficient; the waste is on the page. Of ~$301K that bought 644K views at $0.47 each, "Start My LLC" is reached by a fraction of a percent of mobile visitors. Every point of on-page lift compounds against the full 644K at no added media spend. A dollarized funnel (wasted-spend + $-left-on-table per audience) needs one input the reporting lacks — signup counts per lander — then the $0.47/view economics turn the drop-off into hard recovered revenue.
Measured (and used above)
- Ad delivery: impressions, CTR, CPC, spend, landing-page views per audience. ad report
- On-page: mobile share, time, click rate + ranking, top interactions. click-map
The gap to close to verify
- Conversion counts (signups/leads per lander) — turns the CTA leak into a dollar figure.
- Session replays of non-converters + real-user LCP (esp. Trades) to confirm the load hypothesis.
Methodology & sources
The diagnosis reads the paid funnel as measurable gates, ties the ad report to the click-map, distinguishes a page problem from a traffic problem, and prioritizes by ICE. Full framework + benchmark set in the methodology doc (shared folder). Sources: