Info products

1---
2name: info-product-playbook
3description: Find, validate, build, and sell an info product from scratch — a paid guide, course, template pack, or system built from knowledge. Use when the user wants to create an info product, find a profitable topic, validate demand before building, price and package knowledge, sell a digital product, or turn expertise into recurring revenue. Covers demand discovery across Reddit, Quora, Facebook groups, YouTube, Amazon reviews, and course marketplaces (with scraper workflows), a validation gauntlet that runs before anything gets built, product construction, the value-first selling system, and the newsletter conversion that makes it recurring. Based on a product that ran to roughly $1k/month recurring and still generates sales from posts made two years ago.
4---
5 
6# The info product playbook
7 
8An info product is knowledge packaged so someone can buy the shortcut: a guide, a system, a template pack, a course. It's the fastest product there is to build (no inventory, no code, no fulfillment) and the easiest to sell badly, because most people build first and look for buyers second. This skill runs the order that works: find proof people already pay, validate before building, build fast, sell by giving value away, then convert buyers into a list that pays monthly.
9 
10The receipt behind the method: a PDF guide in a deeply known niche, sold by giving value and answering questions where the audience already gathered, converted into a newsletter running at roughly $1k/month, with the original posts ranking top-5 on Google for the niche term and still producing sales DMs almost two years later with zero new work. Every phase below is that arc, generalized.
11 
12## The one prerequisite
13 
14Pick a niche the user genuinely knows. This is not motivational advice; it's mechanical. Every later phase (spotting which questions signal money, writing answers that read as expert, handling the DMs, making the product actually good) runs on niche depth, and no amount of scraping substitutes for it. If the user has no such niche, the first task is picking the closest one to their real experience, not the one with the biggest numbers. A mediocre niche known deeply beats a great niche known from the outside.
15 
16## Phase 1: demand discovery — mine where the audience already talks
17 
18The product topic is never brainstormed; it's discovered. The target is a problem that appears repeatedly, that people already pay to solve, and that current solutions solve badly. Run as many of these sources as fit the niche, and log every find in one research file: the problem, the exact words used, the source link, and any evidence of money changing hands.
19 
20**Reddit.** The richest source for most niches. Find the subreddits where the audience complains, asks, and celebrates. Reddit's public RSS endpoints (append `.rss` to any subreddit or thread URL) return posts and comments with usernames, text, permalinks, and timestamps, no API key needed, at roughly one request per minute per IP, which is enough for scheduled monitoring; Apify's Reddit scrapers pull at volume when a full historical sweep is wanted. What to flag: the same question asked monthly by different people, threads where the top answer is long and gratefully received (proof the answer has value), anyone saying "I'd pay for", "is there a course on", "someone should make", and complaints about existing products in the niche.
21 
22**Quora.** Underrated because it's unfashionable, which is exactly why it works: questions there are explicit demand statements, and the metrics are public. Question follower counts and view counts show how many people want an answer; a question with thousands of followers and no great answer is a product topic wearing a sign. Apify has Quora scrapers for pulling questions, answers, and stats by topic at scale. Bonus: Quora answers rank on Google for years, so the discovery platform doubles as a sales channel later.
23 
24**Facebook groups.** Where several audiences live almost exclusively: local trades, hobbies with older demographics, parenting, many B2C niches. Honest constraint: Facebook aggressively blocks automated scraping of groups, and running scrapers on an account risks losing it, so treat groups as a manual-observation source: join the 3-5 biggest groups in the niche, read a few weeks of posts, note what gets asked over and over and which answers get hundreds of comments. Apify's Facebook tools work for public pages and the ad library (competitor ads running for months = a product that's making money), but group mining is done with human eyes.
25 
26**Amazon book reviews.** The single fastest demand-mining trick in this playbook: pull the reviews of the top books in the niche (Apify or manual). Five-star reviews reveal the transformation people wanted; three-star reviews are a list of what the existing products failed to deliver, in the buyer's own words. "I wished it covered X" appearing across several books IS the product spec, pre-written by the market.
27 
28**Course marketplaces.** Udemy, Skillshare, Gumroad discover pages, Whop: what's selling in the niche proves the category, and the reviews on the bestsellers show the gaps. A crowded shelf is a good sign, not a bad one; competition is proof of spend, and the goal is a sharper angle, not an empty market.
29 
30**YouTube.** Comment sections under the niche's popular videos (scrapable via Apify or the YouTube Data API): recurring questions, "part 2 please", confusion about steps. High views on tutorial content proves people seek this knowledge; the comments show what the videos still leave unsolved.
31 
32**Search data.** Google autocomplete and People Also Ask for the niche's key terms map what gets actively searched, and keyword tools show volume. Search demand matters double here because info products keep selling from ranked content for years; the receipt product still sells from a two-year-old top-5 ranking.
33 
34After the sweep, cluster the findings. The winning cluster has all three: repetition (many people, same problem), money (they already pay for worse solutions: books, courses, tools, services), and a gap (the paid solutions get specific, repeated complaints).
35 
36## Phase 2: validation — prove it before building it
37 
38Building first is the classic failure. The product gets built only after at least two of these signals fire:
39 
401. **The already-pays test.** Concrete evidence of spend on this problem: competing products with visible sales or review counts, services charging for it, tools with pricing pages. "People complain about X" is not validation; "people pay $40 for a worse answer to X" is.
412. **The value-post test.** Post genuinely helpful content on the problem where the audience gathers (a Reddit post giving away real answers, a Quora answer, a group comment) and measure what comes back: upvotes prove interest, but DMs and "where can I learn more" replies prove purchase intent. This is a free ad test that also builds the reputation the selling phase needs. One strong post with strangers asking for more is worth more than any keyword number.
423. **The pre-sell.** A one-page description of the product with a buy or waitlist button, shared in the same channels or with the people who DMed. Actual emails or actual pre-orders are the strongest possible signal. If the audience won't give an email for it, they won't give money for it.
434. **The competitor-gap check.** Read every review of the two closest competing products and confirm the planned product answers the top three complaints. If it can't beat the incumbents on their weaknesses, the angle isn't sharp enough yet.
44 
45If validation fails, the research file already contains the next candidate cluster. Killing a topic in a day costs nothing; killing it after three weeks of building costs the whole project's momentum.
46 
47## Phase 3: build it fast and complete
48 
49**Format follows the problem, not ambition.** A PDF/Notion guide for knowledge and systems (fastest to make, easiest to update). Templates, swipe files, or spreadsheets when the buyer wants the artifact more than the explanation. Video only when the subject is genuinely visual (software walkthroughs, physical technique); video multiplies production time and most "courses" would convert and deliver better as a sharp guide plus a few screen recordings. Version one is the smallest format that fully solves the problem; a $30 guide that ships this week beats a $200 course that ships never.
50 
51**Structure by outcome, not by topic.** The buyer wants to get somewhere. Open with the result and who it's for, then the steps in the order someone would do them, each specific enough to execute today (the same completeness bar as a great Reddit post: if a reader would need to ask "but how exactly do I do X", X isn't done). Include the mistakes and the "what if it doesn't work" paths, because handling failure modes is what separates products from blog posts. End with the fastest path: "if you only do five things, do these."
52 
53**Write it with AI, verify it with expertise.** Drafting with an agent collapses the build to days, but every claim, number, and step gets checked against the user's actual knowledge, because the niche's real practitioners are the audience and one wrong specific costs the product its credibility. The user's own numbers and stories go in; they're what no competitor can copy.
54 
55**Price on value with tiers.** Anchor to what the outcome is worth and what the audience already pays (from Phase 1's research), not to page count. A simple ladder works: the core product, a higher tier with templates/tools/extras, and later a premium tier with access or updates. Underpricing reads as low quality in most niches; the honest test price is "would the people who DMed pay this without flinching," and price is the easiest thing to test by simply changing it.
56 
57**Delivery is boring on purpose:** Gumroad, Whop, Lemon Squeezy, or a Stripe link on a simple page. No custom checkout builds; the product sells or doesn't based on demand and distribution, never checkout polish.
58 
59## Phase 4: sell by giving the value away
60 
61The selling system is the same mechanic that validated the product, run at volume. The core rule: the free content gives away real answers completely, and the product is the organized, complete, done-for-you version for people who want the shortcut. Holding back "the good stuff" kills both the content's reach and the trust that converts.
62 
63- **Reddit:** the full playbook lives in the reddit-playbook skill if installed (value posts with everything given away, money in the DMs, accounts with age and karma, humans click submit). Info products are the easiest thing that playbook sells because the post can literally contain a working version of the method.
64- **Quora:** answer the exact questions Phase 1 found, completely, over weeks. Quora tolerates a relevant link at the end of a genuinely great answer better than Reddit does, and answers compound on Google. Ten excellent answers in the niche outperform a hundred thin ones.
65- **Facebook groups:** be the person who answers, consistently, with no links (most groups ban them); the profile and DMs carry the product. Slowest channel, highest trust when it lands, and for some niches it's where all the buyers are.
66- **Search long tail:** the discovery content IS the SEO play. Posts and answers written around the niche's search terms keep ranking and keep selling for years; title them for what buyers search, and let the two-year receipt be the expectation: this channel pays slowly and then forever.
67- **DM conversion everywhere:** answer the person's specific situation first, fully; then the product as the shortcut, one line, no pressure. The generosity is the pitch.
68- **Proof accumulates:** every buyer result, screenshot, and thank-you goes into a proof file and then into the sales page and posts. Real numbers only, always attributed the way they happened.
69 
70## Phase 5: the newsletter turns it recurring
71 
72One-off sales are the start; the list is the business. Convert every buyer and every warm DM to an email list (a free chapter, template, or the "updates included" promise makes the ask natural). Then the rhythm: consistent, genuinely useful emails in the niche (the same give-it-away rule), with the product and its higher tiers woven in as the shortcut. This is what turned a PDF into roughly $1k/month recurring in the receipt case: the list buys the next thing, funds the next product, and compounds while the old posts keep feeding it new readers. The second product sells to the list on day one, which is why the second product is always easier than the first.
73 
74## The failure modes, named
75 
76- Building before validating (the big one; the phases exist in this order for a reason)
77- Picking a niche from spreadsheet numbers instead of knowledge
78- Holding back value in the free content to "protect" the product
79- Making a video course because courses feel more legitimate, when the guide would ship in a tenth of the time
80- Vague products ("improve your marketing") instead of specific outcomes ("get your first 10 clients from local businesses")
81- Fake numbers anywhere; the niche's practitioners are the audience and they can tell
82- Stopping after launch week: the long tail and the newsletter are where this model actually pays
83 
84## Running this skill
85 
86When the user has no topic: start at Phase 1, run the multi-platform sweep for their niche (or help pick the niche first), and deliver the research file with clustered candidates ranked by repetition, money, and gap evidence. When the user has a topic: run Phase 2's validation gauntlet on it before any building, and be willing to fail it. When the user has a product already: start at Phase 4, audit the current distribution against the channels, and build the missing ones plus the newsletter. At every phase, deliver concrete artifacts: the research file, the validation post drafts, the product outline, the channel-by-channel content calendar, the email sequence. The user's real numbers and real niche knowledge go in everywhere; nothing generic ships.