Income Nobody Can Switch Off

Listicles and roundups earned almost 22% of all AI citations in a survey of 1,000 marketers, because they are built from clean, liftable pieces.

Human-written content held readers for 2:31 on average, versus 1:54 for fully AI-written content, across 3,000+ pieces from 100 companies.

AI-written content with a human edit barely improved (1:59), so the human has to shape the substance, not just polish it.

AI engines judge whether your site has any business covering a topic. Depth in one lane beats range across many topics.

Customer reviews ranked as the #1 outside trust signal, community recommendations (Reddit, forums) came second, and industry awards ranked last.

Brand mentions and review sentiment now work like ranking factors and should be tracked with the same discipline as keyword rankings.

Generic queries get answered without citations. Expert-level, situation-specific questions are the ones that require a source.

Use AI for research, outlines and gap-finding. Have a person write or heavily rewrite the final version with real opinions and client examples.

Reformat top pages: one point per header, one fact per bullet, tables for numbers, and add schema markup.

Rewrite your AI-drafted articles so a human supplies the voice, opinions and real client examples.

Audit your last 20 posts and flag any that are outside your core topic lane.

Ask happy customers for reviews, reply to negative ones, and check what Reddit and forums say about you.

List the hardest questions your best clients asked this year and turn each into a detailed, industry-specific article.

Low effort cost about $0.80 per run, versus $3.26 at max effort, which is roughly 4x the cost for the same job.

Start on medium effort and move up only if the answer isn’t landing, rather than starting at max.

Fast mode doubles the rate you pay, so turn it off unless you truly need the speed.

Route work by tier: cheapest model for scoping and cleanup, mid-tier for planning and checklists, Astra only for the final judgment call.

Disconnected or unused tools still load their instruction manuals into every message. One example went from a 55,000-token load to 3,000.

Switching models or effort mid-task breaks your cache and reprocesses everything at full price, so pick once and stay.

Past about 272,000 input tokens, the whole request is repriced: input at 2x and output at 1.5x.

Shorter prompts barely matter. Compacting costs a full extra send. Screenshots cost more than pasted text. PDFs are charged twice per page (text plus image).

Run the usage-diagnosis prompt first to see what’s actually draining your limit.

Set reasoning effort to medium and switch off fast mode.

Turn off every plugin and tool you haven’t used in over a month.

Start a fresh conversation for each new workflow, and keep your working context well under 272k tokens.

Set up sub-agents so cheaper models handle scoping and planning while Astra only reviews.

Convert PDFs to plain text, and paste text instead of screenshotting it.

The Call of Duty campaign shown paid $1.75 per 1,000 views on TikTok and YouTube and $1.50 on Instagram, with a $2,500 cap per post.

Join campaigns while the remaining budget is strong, because a nearly exhausted budget can run out before your clips go live.

Study the top-performing clips first. In this example, the winning pattern was a famous-weapon hook paired with gameplay.

Using only the brand’s official footage folder avoids copyright and approval rejections.

Claude scanned 110 TikToks and 142 Shorts and found the edit patterns: burned-in text from second 0, two cuts for two platforms, and creator audio over library tracks.

Common rejection reasons: footage from outside the official folder, background music, hidden likes, a missing exact caption phrase, and taking a post down within 30 days.

CapCut’s desktop AI clipper turns long footage into short clips, and the browser version can’t do this.

Submissions are due within 30 minutes of creating the video, and the creator recommends at least 10 clips a day.

Join Content Rewards on Whop and shortlist campaigns with healthy remaining budgets and strong CPMs.

Connect Whop to Claude and ask it to rank current campaigns by payout.

Screenshot the top-performing clips and have Claude extract hook patterns and Reddit sentiment.

Paste the brand’s requirements into Claude and ask for a plain-English editing and caption checklist.

Edit only official footage in CapCut desktop, 9:16 vertical, with approved hooks and exact captions.

Submit each clip within the required window and track views against a payout calculator.

Target small local businesses with a lead-quality pain point: dentists, chiropractors, real estate, massage therapists.

Typical pricing is $300 to $500 per month per client, and the model works because it is monthly recurring revenue.

Lead scoring separates real prospects from bogus form entries, saving staff from chasing low-quality leads.

The agent shown researches the company website and public sources, runs 10 observable SEO and conversion checks, scores fit against an ideal customer profile, and creates follow-up tasks for hot leads.

White labeling lets you keep 100% of payments, and the cost doesn’t increase with more clients, so margins stay high.

Triggers connect the agent to a website form or tag, so scoring runs automatically on every new lead.

Position the agent as taking a task off the receptionist’s plate, not replacing them.

Further agents (review requests, social posting, competitor watch) can be stacked to raise the value of each account.

Pick one niche and one city, then list 20 businesses from a Google “more places” search.

Choose one narrow agent to sell first, such as lead scoring.

Build and test it on a sample contact so you can show real output.

Quantify the value in your pitch: staff hours per week times hourly cost.

Connect the agent to the client’s form with a “new lead” trigger.

Offer a second agent to existing clients after 30 days.

Emails under 100 words got an 11.9% reply rate, versus 1.9% for 100 to 200 words and 0.3% beyond 200. The sweet spot was 60 to 90 words.

Subject lines under 30 characters opened at 83.7%, versus 69.5% for longer ones, and including a number performed best.

One follow-up lifted replies from 4.1% to 6.6%, and two follow-ups topped out around 6.9%. More than that irritates people.

Directors and managers were emailed about 3x more often than CEOs, and at firms over roughly 50 people the team lead is usually the decision-maker.

Tuesday and Wednesday outperformed Thursday and Friday.

One commenter ranked what matters most as offer, then recipient, then timing, then copy. “Can I write a guest post?” is a weak ask.

A correction, a missing source or an original statistic gives an editor a reason to act now.

A template-sounding first line ends the read, and personalization is what stands out.

Cut every outreach email to 60 to 90 words with one clear ask.

Rewrite subject lines to under 30 characters, and test one with a number.

Group your last 50 emails by ask type and count positive replies, not total replies.

Send one follow-up a few days later, and stop after two.

Target team leads and managers at larger companies, and founders only at small ones.

Track opens to spot deliverability problems before rewriting copy.

Effort means evidence of useful work on the page. Examples include original data, tested products, unique images and authentic reviews.

Mass-generated template pages (5,000 location pages with only city names swapped) are low effort and doorway-page risks.

“10 best protein powders” with no testing is low effort, while “10 best protein powders we’ve tested” with methods and photos is high effort.

Recycled or stock images count as low-effort media, and original photography and video signal effort.

Put the answer first. A 600-word egg-history intro before “how to boil an egg” is filler.

Forum and social pages gain effort from the depth and number of contributors.

A Google presentation from 2016 said “we do not understand documents” and that it learns from user reactions, so quick back-clicks are a negative signal.

Google’s John Mueller has said its systems can “lose faith” in websites.

Score your top pages on Shepard’s five questions: edited, unique, evidence of work, beyond common facts, deep discussion.

Move the direct answer or verdict to the top of every page.

Replace stock images with original screenshots, photos or custom explainers.

Add first-party data, testing methods or real examples to each key article.

Remove or consolidate near-duplicate template pages.

Add a clear, custom hero image to commercial pages so the purpose is clear at a glance.

Start at one Short a day and watch the 48-hour views in YouTube Studio.

Move to three a day only once the 48-hour mark is above 50,000, then to five.

Jumping straight to high volume can look bot-like and trigger suppression.

If volume rises but views don’t, quality may be lacking, or the channel may be suppressed.

He recommends turning off “publish to subscription feed” so Shorts are pushed to a broader audience.

Some mass-market formats can handle up to 24 videos a day, while narrow niches run out of content and can see diminishing returns.

Some channels cap at one a day. If views fall as you upload more, go back down.

He compares consistency to a credit score: not missing days makes a channel more reliable and more likely to get pushed.

Post one Short daily and log your 48-hour views weekly.

Step up to three a day only after you exceed 50,000 views at the 48-hour mark.

Turn off “publish to subscription feed” on new Shorts and compare reach.

Test higher volume for two weeks and revert if views decline.

Improve quality (hook, pacing, analytics) before adding more volume.

Hire and train editors once the channel earns roughly $1,000 to $3,000 a month.

Email 1 (deliver the promise) had a 57.09% open rate and 19.28% click-through rate, and led to 223 buyers and nearly $55,000 in revenue.

Email 1 included a low-ticket offer in the P.S., a 20% discount for 24 hours, which serves fast buyers.

Email 2 went out nine hours later as a workshop reminder and last call on the discount (49.5% open, 6.89% click-through).

Email 3 bridges to the bigger problem. The $27 course opens the door to a flagship offer that costs over seven times more.

Email 4 is a “receipt” of student results, Email 5 answers objections as an FAQ with testimonials, and Email 6 adds real urgency.

Open rates drift down across the sequence (from 57% to 36%), which she says is normal.

The final email alone generated over $7,000, but only because the deadline was real.

Each email sells something different: engagement, a quick solution, belief, proof, certainty, scarcity.

Make email 1 deliver exactly what the signup promised, with a clear link to consume it.

Add a relevant low-ticket offer in the P.S. of email 1 with a time-limited discount.

Write one bridging email that deepens the problem before introducing your core offer.

Build a proof email (results) and an FAQ email that answers unspoken objections.

Close with a real-deadline email, and don’t fake scarcity.

Track open, click and revenue per email to find which one actually earns.

There are 36.2 million US small businesses, 57% say reaching customers is their biggest problem, and only 19.8% use AI at all.

Pick niches by ticket size, urgency, lead value versus lead cost, weak online competition, and owner-run crews of 2 to 20 people.

Go “one door down” from crowded plumbing and HVAC, to things like garage floor coatings, hardscapes, waterproofing and hood cleaning.

Sell the result, not the tool. For example, “every missed call gets a reply in under a minute.”

Pages written at a fifth to seventh grade reading level converted at 11.1% versus 5.3% for professional-sounding copy.

Suggested pricing: a $2,500 to $5,000 install plus a retainer around $1,500 a month, priced on what a job is worth, not your hours.

His audit of 11 Charlotte sites found 9 had no email capture, 8 had no price-first step, 9 had phone issues, and none had traffic problems.

If leads rise but booked jobs don’t, fix follow-up first. Phone handling, not the funnel, may be the issue.

Pick one big-ticket niche and one metro, then list 100 businesses.

Scrape their websites and have an agent audit conversion gaps (email capture, price-first step, phone, follow-up).

Build one pitch funnel in the prospect’s brand and record a short personal video.

Send a short first email: one observed gap, one rebuilt page, one 15-minute ask.

Do the math with the owner: leads per month, close rate, job value.

Run two or three free case-study builds before you set your pricing.

For five or more locations, use a hub-and-spoke structure: a locations index, state-level hubs, and city pages with a “nearby locations” module.

Link each city page to its six closest neighbors using drive time, not straight-line distance. One control test showed a 7% traffic increase from proximity linking alone.

Service-area pages paired with localized content and internal links increased traffic by 107% in one Sterling Sky study.

Content links carry roughly three to five times the weight of footer links.

Health targets: orphan pages under 5% of the index, important pages within three to four clicks, a 70-20-10 anchor mix (branded/natural, partial, exact), and a crawl-to-index ratio of 90%+.

Phase 1 (weeks 1 to 2) needs no new content: fix broken links and connect orphans. He says audits typically find 20 to 40 orphan pages and 15 to 30 broken links per site.

Compare internal link counts between peer city pages, since disparity often explains why one ranks and another doesn’t.

Revenue math: 200 extra visits at 4% conversion and $150 per job adds $1,200 per month per location.

Crawl the site with Screaming Frog and export internal link counts from Search Console.

Find orphan pages by comparing your XML sitemap with crawl output, then link to each.

Fix broken internal links and check mobile link errors.

Add a nearby-locations module and breadcrumbs to all city page templates.

Check anchor text against the 70-20-10 ratio and reduce exact-match overuse.

Report gains using the traffic × conversion × job value formula.

Design the above-the-fold area to signal the page’s purpose and key attributes. In his case, a drag-and-drop upload element and a “no sign up” label.

Keep text minimal to avoid diluting tokens, and put key entities and n-grams in the first 2% of tokens.

He reports growth from about 50 to 8,000 clicks a day with zero content or visual change, just time and click data, and says rankings can dip 50% before recovering.

Cleaning AI “signatures” from HTML, CSS class names and div hierarchies may reduce footprint, since AI site builders reuse the same patterns.

He cites BrowserStack losing 70% of traffic in 8 months after publishing 2,000 AI documents in one day, despite strong authority.

Mature a new subfolder for about 3 months before scaling, then publish in randomized daily volumes across a 7 to 8 month rollout.

Cover roughly 20% more topics than your main competitors by scraping eight competitor sitemaps.

Free tools can replace clicks lost to AI answers (for example, a case-compensation calculator for a law firm).

Link building plays three roles: PageRank, mentions, and claims (attributes like “biggest firm in Texas”) that shape AI answers.

Redesign your top landing page so its core attribute (free, no sign up, instant) is visible in the first screen.

Cut on-page text to what supports relevance, and place key entities early.

Audit AI-generated pages for repeated class names, HTML patterns and boilerplate.

Plan any large content rollout over months with randomized daily publishing.

Scrape competitor sitemaps and map the topic gaps you can fill.

Add one free interactive tool to a page that currently relies on informational traffic.

The keyword playbook takes a seed topic and search market and returns a prioritized list based on volume, CPC, rankability and the lowest domain rating on the page.

Each site is built by a chain of agents, so if something breaks only a small piece needs repair.

Generated pages include a short AI-answer block, tables, internal links, review videos, FAQs and a table of contents.

Content can be set to refresh every 30, 60 or 90 days.

Output can be published to Netlify in one click, with WordPress and other options available.

Playbooks chain the steps: keyword research, minisite, calculator, then backlinking and promotion.

Backlinking is semi-automated, with the human handling captchas, positioned as a safer alternative to older mass-link tools.

The demo site was built from a one-line request (“build me a niche site about cold plunges”), and more detail gives better output.

Run the keyword playbook on Aha Studio on your target niche and shortlist the winnable terms.

Build a 10-page minisite from the best keyword and add your affiliate link.

Generate images and infographics before publishing so the site looks finished.

Publish to Netlify (or WordPress) and set a 30-day content refresh.

Add a calculator to the site, then run the backlinking playbook.

Try the free demo tools linked on the sales page to see the output firsthand.

A niche scanner shows which directories already rank and which have the best opportunity, and can pull easy keywords from them.

The build runs as a chain of agents, so any failure is isolated and repairable.

The demo started with 42 candidate pages and filtered down to 25 listings, with weak-data listings excluded.

Photo options: Wikimedia Commons (best for places like parks), business-site photos, or generated images with your own key.

A directory can be expanded with a second campaign (another state or city) under the same domain, and he mentions sites of 10,000+ listings.

Your own data can be added by CSV upload.

Monetization options include ad placements once traffic is decent, sponsorships, and rank-and-rent.

Output can be a zip download, WordPress, Netlify or other hosts.

Scan a niche and location to find the directory with the best opportunity.

Start with a small 25-listing build to test quality before scaling to 1,000.

Choose a photo source (business-site photos or Wikimedia) and generate any missing images.

Publish to Netlify or WordPress and check filters, listing pages and internal links.

Add a second city or state to the same campaign to grow the directory.

Plan monetization up front: sponsorships first, ad networks once traffic builds.