Build a landing page
A page brief built from your tested angles, with your real offer and call to action.
- Hero headline: the winning headline, word for word
- Problem section opens with the winning pain line
- One call to action, repeated, linking to your booking page or form
- Sections: hero, problem, offer, how it works, call to action
Show the full prompt
# Build a landing page: 1Signal · Agency test Signal Prompt Pack sp_644003038f06 for 1Signal, prepared by 1Signal. Results as of 2026-10-18. Partial: 3 of 5 tests finished. ## The business * Business: 1Signal * Offer: 1Signal · Agency test * What it sells: A Meta testing system agencies run with each client: an offer report from the client's intake, then five tests. Each test changes one thing (pain, headline, offer, image, ad text). Every ad pauses at 400 impressions. Each winner carries into the next test. The agency and client review each result together; the client approves each test before it spends. It ends with a handover: the top ads and the evidence behind them. Design partner offer: 3 slots; we run the first client program with the agency, and the client pays only its own ad spend (about $190 for all five tests). 5 of 9 free first tests became paying clients, all within five weeks. We never pitch, sell to or solicit the agency's clients. * Type of business: Software and service for marketing agencies (B2B) * Who it is for: For agencies running Meta ads (from undefined) * What the buyer wants: Clients who stay longer and trust the next decision. Proof they saw themselves before they recommend anything: "As long as I can say, hey, we saw these results ourselves." New clients won with evidence instead of guarantees. * Country: United States (undefined) * Website: https://1signal.io * Call to action leads to: https://1signal.io/?utm_source=meta&utm_medium=paid_social&utm_campaign=1signal-agency-discovery * Brand: take colors, fonts and logo from the website https://1signal.io * Language: English ## What the tests showed Every line here was tested as a Meta ad (Facebook and Instagram feed), optimized for link clicks. New landing pages, emails or LinkedIn messages made from these lines still need their own test. Status: Partial: 3 of 5 tests finished. ### Step 1: Pain (finished) What it tested: The pain line in the ad text. Each ad states one problem. Find which problem makes the buyer stop and click. Held fixed in the headline slot on every ad: "For agencies running Meta ads". Winning line (chosen by a person), use as is (Pain line 8 (P-10-CLI-8fa Client)): ```text True or False? My clients keep asking which part of their results is my job and which is theirs. ``` Numbers: 404 views, 15 link clicks, 3.71% click rate, 4.62 USD spent. How sure: Tied: several ads are statistically level; the numbers cannot pick one. The winner's click rate is 36% above the median of the other 11 compared ads (2.73%). Data: 12 of 12 ads reached the stop rule. 7 ads tied at the top, so the numbers cannot separate them. Treat these 3 as equally strong alternatives: Tied line (Pain line 2 (P-12-AGE-2e2 Agency owner)): ```text True or False? My prospects have been burned so many times that I need case study after case study. ``` Numbers: 423 views, 17 link clicks, 4.02% click rate, 4.88 USD spent. Tied line (Pain line 2 (P-04-AGE-c2a Agency owner)): ```text True or False? I'm on a hamster wheel of losing clients, even ones I've had for years. ``` Numbers: 405 views, 15 link clicks, 3.7% click rate, 5.16 USD spent. Tied line (Pain line 8 (P-02-CLI-6b8 Client)): ```text True or False? A screenshot of Ads Manager no longer convinces my clients that anything is working. ``` Numbers: 439 views, 16 link clicks, 3.64% click rate, 5.31 USD spent. Caution: 7 ads tie at the top on these views, so the numbers cannot pick one of them. ### Step 2: Headline (finished) What it tested: The headline. Find which headline carries the winning pain best. Built on the Step 1: Pain winner (Pain line 8 (P-10-CLI-8fa Client)), held fixed in this test. Winning line (chosen by a person), use as is (Headline 8): ```text Every good month, your client asks what you did ``` Numbers: 407 views, 19 link clicks, 4.67% click rate, 3.65 USD spent. How sure: Tied: several ads are statistically level; the numbers cannot pick one. The winner's click rate is 57% above the median of the other 13 compared ads (2.98%). Against the baseline ad: +91%. Data: 14 of 15 ads reached the stop rule. 8 ads tied at the top, so the numbers cannot separate them. Treat these 3 as equally strong alternatives: Tied line (Headline 2 (HEADLINE-224 Headlines V02)): ```text Your client cannot tell your work from their own luck ``` Numbers: 408 views, 19 link clicks, 4.66% click rate, 4.1 USD spent. Tied line (Headline 4): ```text Proving your value gets harder after every slow month ``` Numbers: 407 views, 14 link clicks, 3.44% click rate, 3.97 USD spent. Tied line (Headline 14): ```text Make your share of the result impossible to miss ``` Numbers: 405 views, 13 link clicks, 3.21% click rate, 4.04 USD spent. Comparison ad (a baseline, it could not win) (Headline 2 (B02-WEBS-84e B02 Website Control)): ```text Seven answers shape five focused ad tests ``` Numbers: 408 views, 10 link clicks, 2.45% click rate, 3.56 USD spent. Caution: 8 ads tie at the top on these views, so the numbers cannot pick one of them. Meta barely showed Agency or client, so it has too few views to judge. ### Step 3: Offer (finished) What it tested: The offer and its call to action. Find which offer and call to action the buyer clicks. Held fixed in the headline slot on every ad: "Every good month, your client asks what you did". Built on the Step 2: Headline winner (Headline 8), held fixed in this test. Winning line (chosen by a person), use as is (Free Checklist): ```text Free Checklist: The 11-point retainer protection checklist that walks you through documenting your specific moves before the results email goes out, so clients never wonder what they are paying for when numbers climb. ``` Numbers: 433 views, 15 link clicks, 3.46% click rate, 4.15 USD spent. How sure: Tied: several ads are statistically level; the numbers cannot pick one. The winner's click rate is 55% above the median of the other 7 compared ads (2.23%). Against the baseline ad: +132%. Data: 8 of 8 ads reached the stop rule. 5 ads tied at the top, so the numbers cannot separate them. Treat these 3 as equally strong alternatives: Tied line (Free audit): ```text Free audit: send us one client's Meta ad account and we show you, in plain numbers, which results came from your decisions and which the market handed you. Bring it to your next client call. Request your free audit. ``` Numbers: 402 views, 12 link clicks, 2.99% click rate, 3.84 USD spent. Tied line (Free Mini Course): ```text Free Mini Course: Five daily emails that teach you how to frame your ad work so every good month lands as proof of your skill, not a lucky streak, and clients stop second-guessing the invoice ``` Numbers: 448 views, 12 link clicks, 2.68% click rate, 5.17 USD spent. Tied line (Free Workshop): ```text Free Workshop: A 60-minute live session where we build a results attribution story for a real campaign you are running right now, so you walk away with a client-ready slide that separates your decisions from market tailwinds ``` Numbers: 405 views, 10 link clicks, 2.47% click rate, 3.87 USD spent. Comparison ad (a baseline, it could not win) (Offer 1): ```text Run your first client campaign with us: your client funds their own Meta ads, software billing stays off, and we guide the first campaign with you. ``` Numbers: 404 views, 6 link clicks, 1.49% click rate, 3.85 USD spent. Caution: 5 ads tie at the top on these views, so the numbers cannot pick one of them. ### Step 4: Image (not finished yet) Not a result yet. Do not use its lines as evidence. ### Step 5: Ad text (not started) Nothing tested yet. Do not assume a result. ## Facts you may use * Real offer and its terms: [MISSING: the real offer, price or trial length] (the historical tested Step 3 offer is evidence above, not confirmation of current terms) Historical tested wording does not confirm current commercial terms. Keep it as evidence; use only confirmed Facts and the current page CTA for a new page. * What the business can deliver: [MISSING: what the business can actually deliver] ## Never claim * No list saved. Still: no invented results, numbers, guarantees or customer stories. ## Missing information * Brand colors and logo are not saved. The prompts point to the website for them. * The real offer and its terms (price, trial length). Saved in Setup under offer facts. * What the business can actually deliver. * Claims the business never makes. * Step 4: Image: not finished yet. * Step 5: Ad text: not started. Where a fact is missing, leave a visible placeholder like [MISSING: price]. Never fill it in. ## Your task Build one landing page with one conversion goal. Write the full page copy, section by section, ready to paste into a page builder. * Conversion goal and call to action: [MISSING: the offer and its call to action]. Use one call to action and repeat it. Every button links to https://1signal.io/?utm_source=meta&utm_medium=paid_social&utm_campaign=1signal-agency-discovery. * Look: match the colors, fonts and logo of https://1signal.io. * Hero headline: the winning Step 2 headline, word for word. * Problem section: open with the winning Step 1 pain line; tied pain lines may follow as the problems the reader recognizes, written as the page's own words, not as quotes. * Offer section: state the offer and only the terms listed under Facts. [MISSING: what the business delivers] * Picture: [MISSING: a tested picture (Step 4 not finished)]. * Proof section: [MISSING: real customer proof]. Leave it as a placeholder unless proof is listed under Facts. * Keep the page short: hero, problem, offer, how it works, call to action. Add a section only when the facts support it. Write in English, the language the ads ran in. Keep tested lines word for word. ## Rules * Use only the facts and tested lines in this brief. Do not invent customers, testimonials, results, statistics, guarantees, awards or deadlines. * Do not invent personal details about any reader or recipient. * Tested lines are ad copy written for the business. Never present them as quotes, reviews or statements from customers, buyers or other companies. * Respect the Never claim list. * Mark every tested line you reuse with [TESTED] and everything you write new with [NEW, NOT TESTED]. * No long dashes and no dash between clauses. Use a period, a colon or a comma. * The source lines were tested as Meta ads. Anything you write from them is a new version that still needs its own test. Do not call it proven or validated.