01 / Choose
Compare three niches
Use a 0–2 score for evidence, access to an audience, and product fit. Scores organize thinking; they do not forecast profit.
The affiliate experiment lab / with Ashley Kays
Affiliate marketing tools.
Your next useful move.
Choose a useful offer, test it honestly, and turn your evidence into content. Put AI to work at each step while keeping the decisions yours.
01 / Work smarter, keep your judgment
AI has a job.
So do you.
Follow the steps in order or open the one you need. Prompts are templates to copy into your chosen AI assistant; this page does not call an AI service.
01Find a person and a problem
- You provide
- Five real audience questions, their source URLs, and your own experience.
- AI can help
- Cluster the questions and suggest narrow audiences.
- You verify
- Are these real questions? Can you reach and help these people? AI is not demand evidence.
- You finish with
- One audience statement and one testable question.
02Compare products and program terms
- You provide
- Official product and affiliate terms, with dates and URLs.
- AI can help
- Extract comparable facts and flag missing eligibility or payout details.
- You verify
- Reopen official terms. Confirm region, allowed channels, qualifying actions, commission period, and reversals.
- You finish with
- A three-product shortlist with reasons to recommend or reject each.
03Test the product
- You provide
- Your use case, budget, and available trial or product access.
- AI can help
- Draft a repeatable test plan and evidence checklist.
- You verify
- Perform the tests yourself. Record limitations and separate vendor claims from observations.
- You finish with
- Original notes, screenshots, and a clear list of what was not tested.
04Build a useful comparison
- You provide
- Your testing notes and verified sources.
- AI can help
- Organize a brief, outline, comparison table, and draft.
- You verify
- Check every claim, disclose relationships, include alternatives and who should not buy.
- You finish with
- An evidence-led article or demonstration.
05Make visuals and repurpose
- You provide
- The approved article, real screenshots, and licensed assets.
- AI can help
- Suggest a storyboard, captions, and adapted email or video scripts.
- You verify
- Use real screenshots as product evidence. Label illustrative mockups and verify asset permissions.
- You finish with
- One complete guide plus two useful adaptations.
06Publish and distribute
- You provide
- Reviewed content, approved links, audience location, and channel rules.
- AI can help
- Create a publishing checklist and a realistic distribution plan.
- You verify
- Test links. Follow community rules. Review content before posting; no automatic group promotion.
- You finish with
- One published piece and a record of where you shared it.
07Review and improve
- You provide
- Comparable reporting periods: visits, affiliate clicks, orders, approved commissions, costs, and hours.
- AI can help
- Summarize observations, propose hypotheses, and suggest one next test.
- You verify
- Confirm reporting delays, denominators, attribution, reversals, and sample size. Correlation is not proof.
- You finish with
- A continue/change/pause decision with a next review date.
02 / From reading to doing
Your practical workbench.
No account required. Inputs remain on this page unless you choose to save them in this browser or download them. No cloud sync. Avoid sensitive information on a shared device.
02 / Choose
Compare affiliate offers
Put the terms beside audience fit. Unknown terms mean more research, not an automatic rejection.
03 / Diagnose
Clicks but no sales?
Use one reporting period and comparable definitions. Pending orders and approved earnings can have different reporting delays.
04 / Model
What would this test need?
A hypothetical one-time commission scenario. This does not model recurring revenue, retention, or delayed payments.
05 / Test
Product evidence workbook
06 / Create
Comparison brief builder
A structured brief and a copyable AI prompt—not an automatically researched article.
07 / Publish
Campaign link planner
Keep the merchant’s tracking link unchanged. Record network-supported sub-IDs separately and create them in that network’s dashboard.
08 / Learn
Personal experiment log
A private browser worksheet for your own test. It does not publish results or connect to Ashley’s public progress tracker.
03 / A tool earns its place
Build a small stack.
Use an existing tool when it does the job. A spreadsheet, one publishing channel, and your affiliate dashboard are enough to start recording a test.
The jobs your stack needs to do
- Research: original product tests and official sources.
- Draft: your preferred AI assistant, working from your evidence.
- Publish: one site, newsletter, or video channel.
- Measure: the affiliate dashboard plus your experiment record.
- Automate later: add Make or n8n only when a repeatable task merits it.
Choose tools on task fit, cost, and current terms. An affiliate program is not proof of product quality. Email providers may restrict affiliate-led businesses even when they allow some affiliate links.
04 / Research before you apply
Programs with a purpose.
These are candidates to investigate, not claims of approval or personal testing. Links are ordinary informational links; no referral IDs have been added. Check current terms on the official pages.
Automation
Make
A candidate for practical automation tutorials. Show an actual working workflow and its running costs.
Automation
n8n
Consider for more technical workflow education. Distinguish paid Cloud referrals from self-hosting.
Email and creators
Kit
Consider for creator email, welcome sequences, and digital product tutorials. Verify earning duration and tier conditions.
Email marketing
MailerLite
Consider for email-list setup tutorials. Before choosing an email provider for affiliate-led content, check its acceptable-use rules for your specific model.
Newsletters
beehiiv
Consider for newsletter publishing experiments. Check current pricing and tier-specific partner terms.
Ecommerce email
Omnisend
Consider for store lifecycle demonstrations. Content affiliates and agency client partnerships have different routes.
CRM and marketing
HubSpot
Consider for CRM education and comparisons. Its content affiliate program differs from its Solutions Partner client-referral program.
Ecommerce
Shopify
Consider for first-store tutorials. Verify eligible full-price plans, merchant locations, and audience requirements.
Search and marketing
Semrush
Consider for search and competitor research tutorials. Check audience eligibility and toolkit-specific payouts.
Design and websites
Framer
Consider for original templates and portfolio tutorials. Check referral attribution and advertising restrictions.
Design and websites
Webflow
Consider for website education. Review whether the affiliate or service-partner track fits your work.
AI audio
ElevenLabs
Consider for original voiceover tutorials. Use voices with appropriate permission and verify eligible subscription plans.
Creative tools
Adobe
Consider for design and photography instruction. Commission basis varies by product and billing arrangement.
Education
Coursera
Consider for thoughtfully curated learning paths. Check eligible products; do not promise jobs or recurring subscription commissions.
Freelance services
Fiverr
Consider for guides to hiring specialists. Referring a buyer purchase differs from referring a freelancer seeking work.
Print on demand
Printful
Consider for tested product and store experiments. Calculate costs and distinguish product revenue from commission.
Crafting
Cricut
Consider for original craft demonstrations and material comparisons. Verify purchase eligibility and the short attribution window.
Creative assets
Creative Fabrica
Consider for craft and design workflows. Check the license for each asset and the commission terms offered to you.
AI building
Lovable
Consider for documented app prototypes. Confirm your offered terms and show what the app actually does.
Network
PartnerStack
Discover business-software partnerships. Network acceptance is separate from acceptance into individual programs.
Network
Awin
Explore retail and other advertiser programs. Confirm geographic availability and each advertiser's requirements.
Network
Impact
Explore brand partnerships and individual program applications. Advertiser approval and terms vary.
05 / Find your people
Learn together.
Ask better questions.
Forums, groups, and organizer links checked October 6, 2026. Recheck dates and membership terms before making plans. Community posts are opinions to evaluate, not verified earnings evidence.
Forum · public browsing
AffiliateFix
Affiliate questions, guides, and follow-along discussions. Start by reading the rules and checking the date and evidence behind advice.
Forum · premium membership + public areas
affLIFT
Affiliate campaign discussion and learning community. Read the public areas first and check current membership terms before paying.
Groups · business building
Indie Hackers groups
Browse active groups about building independent businesses. Useful for sharing an experiment and asking specific questions; not every group focuses on affiliates.
Community · automation
Make Community
Ask workflow questions, learn from examples, and explore community sessions. Remove credentials and customer information from examples you share.
Conference · Bangkok · December 9–10, 2026
Affiliate World Asia
Performance and affiliate marketing event. Compare the agenda, travel costs, and specific people you hope to meet before buying a ticket.
Conference · Las Vegas · January 26–28, 2027
Affiliate Summit West
Affiliate and partnership event at Caesars Forum. Check the official organizer for current agenda, eligibility, ticket terms, and dates.
06 / Small tests, a little fun
Make the next move playful.
Pick a tiny experiment
Would you recommend it?
Fictional scenario: Tool A pays a large commission, but you have not tried it. Tool B pays less; your tests show it solves your audience’s problem, with one limitation.
Fix this comparison
Before: “This is the best tool for everyone. It will double your income.”
Reveal a more useful approach
After template: “For [specific task], my [dated test] found [documented observation]. The limitation is [evidence]. Consider [alternative] if [constraint].”
Fill the brackets only with real evidence. Do not publish the example as your experience.
Your seven-day mini mission
- Choose one audience question.
- Compare three products.
- Test one workflow.
- Capture evidence and limitations.
- Publish one useful comparison.
- Share it where welcome.
- Review what happened and choose one change.
Measure what you learn. A week without sales is still a week with evidence.
Quick answers / affiliate marketing with AI
Start with the question.
AI can organize research, turn your verified notes into a draft, and suggest what to check next. You still choose the product, run the test, verify the claims, and approve what gets published.
Where should I use AI in an affiliate marketing workflow?
Use AI to group audience questions, build a product-testing checklist, outline a comparison from your evidence, repurpose an approved article, and summarize your weekly measurements. Review the output at each step. The seven-step workflow shows the input, AI task, human check, and finished output.
Which affiliate program should I join first?
Choose a program for a product that solves your audience’s problem and that you can evaluate credibly. Check eligibility, allowed promotion channels, commission basis, attribution, reversals, and payout rules on the official page. Compare audience fit and evidence before comparing payouts.
Why do affiliate clicks sometimes produce no sales?
Possible causes include a mismatch between the reader and offer, an unclear recommendation, a broken tracking link, reporting delays, or checkout friction. A small sample can also produce zero sales. The clicks-to-sales checklist helps choose the next thing to check; it cannot establish the cause from totals alone.
What is affiliate earnings per click?
For this worksheet, approved earnings per click equals approved commissions divided by affiliate clicks for the same reporting period. For example, $30 in approved commissions divided by 100 clicks is $0.30 per click. This hypothetical measure excludes costs; check how each network defines its own EPC.
Can AI write a review of a product I have never tested?
It can help summarize sourced product information, but that is not a first-hand review. Label vendor claims and untested points clearly. Never ask AI to invent personal use, screenshots, performance results, or customer stories. Use the evidence workbook to record what was actually tested.
Does this lab send my worksheets to AI or save them online?
No. These worksheets use browser calculations and templates. Copying a prompt does not send it to an AI service. Saving is optional and local to this browser; exporting downloads a JSON file. The personal log is separate from public experiment tracking.
See the work behind the guidance
Meet Ashley through her work.
Explore Ashley Kays’ product, design, marketing, and customer-experience projects.