Operator essay #8 · the anti-recommendation

The dossiers we'd tell a friend to skip.

Most idea catalogs sell every idea as a winner. We have 238 ideas. They are not all winners. The honest read on which dossiers you should probably not buy, named by pattern with specific catalog examples.

2026-05-14 · 9 min read · ~1700 words

Here is the test we run before recommending any catalog product to a friend who is actually considering it: could they reach their first 50 buyers in 60 days, given their actual access? That question kills more dossiers than the idea quality does. Four out of five times when we tell a friend to pass, it is not because the idea is bad. It is because the friend does not have the unfair leverage that idea requires.

This essay names the patterns where we have told friends to skip. We use specific catalog products as examples, with their actual Adoptability scores and Fermi math. We do not pretend the catalog is uniformly excellent. It is uniformly structured, which is a different claim, and a more honest one.

If you came here from the catalog, this is the operator essay we did not need to write. We are writing it anyway because the catalog only has long-term value if it stays honest about its weaknesses.

Pattern 1: the free-substitute trap

A handful of catalog products have clear pain, clear buyer, clear pricing, AND a free incumbent the buyer will eventually default to. The dossier is real work, but the underlying business has a built-in ceiling because the substitute has zero switching cost. Year-2 retention is the hard problem, not Year-1 acquisition.

The pattern is not "every idea with a free substitute is dead." Plenty of paid SaaS lives next to free incumbents. The pattern is: if the dossier does not name a specific switching cost the buyer pays to stay with you, the free substitute eventually wins. Read the Fermi summary on each product. If it mentions "ChatGPT does this for free" or "HubSpot ships the same feature free to existing users," that is the dossier telling you the truth.

Pattern 2: the buyer who does not quite exist

Sometimes the catalog ships a product where the buyer description is technically correct but no specific human matches it. "For SaaS founders building bottom-up motions in 50-person orgs" sounds named, but no single person at that org thinks of themselves that way and writes a check.

This is the rarest skip-pattern in the catalog because the Adoptability scorer downweights vague buyers heavily. The few products we have flagged this way score 60 to 65 and we usually pull them or rebrand them within a tick or two. If you are buying a dossier on a product scoring under 65, read the buyer-clarity axis specifically. If it is at 6 or 7, the Phase A engagement will spend its first week finding the actual human (not the persona) and may pivot the ICP entirely.

This is the pattern where Phase A engagement adds the most value. The fix is talking to ten real prospects in week one. The dossier alone cannot do that for you, regardless of price.

Pattern 3: the financial ceiling that does not pay rent

Some catalog products are honest, clear, and small. Year-1 ARR mid-case is $50k to $115k. After build cost and operating expense, the operator's Year-1 take-home is somewhere between zero and slightly negative. That is fine if you are running this on the side of a day job, or building it on top of an existing book of clients, or treating it as a strategic loss leader for a larger consulting practice. It is not fine if you need the product to be the business that pays rent.

The honest framing on this pattern: the catalog gives every product a year-1 take-home estimate. Three out of four of those numbers are negative. That is not a bug; that is what real bootstrap SaaS looks like in Year 1. Investment to launch is $30k to $40k. Year-1 revenue rarely covers it. The right comparison is not "is this profitable in Year 1" but "is this a real business by Year 3." Most of the catalog products in this category answer yes to that. If you cannot wait three years for the answer, this is not the catalog for you.

Pattern 4: regulated verticals you cannot enter alone

The catalog has products targeting healthcare, banking, K-12 education, and clinical research. Adoptability scores on these tend to be in the 60s because the credibility axis is realistic about how long enterprise-regulated sales cycles take. If you are reading the dossier and asking "can I get a HIPAA BAA by week 4 of Phase A?" the answer is no, and the dossier does not pretend otherwise.

The pattern is not "skip every regulated product." It is "skip every regulated product unless you have a specific compliance lawyer on retainer and a customer willing to pay through the painful first audit." Both of those need to be true. If only one is true, the project stalls at month 4 when the buyer's procurement team asks for documentation you do not have.

The dossier we would recommend instead

If you read this essay and most of your candidates were in the skip patterns, the next move is not "buy three more dossiers and hope." It is: pick one product where you have the unfair leverage the dossier requires. Concretely:

Filter the catalog by your actual distribution access. If you sell to bookkeepers today, start with Bookkeeper AI (80) or Audit AI (75). If you sell to small fleets, start with Dispatch AI (77). If you sell to lawyers, start with Lawfirm AI (lower score now but the only product that graduated to its own domain). If you sell to roofers or trades, start with Roofing AI (78).

The score is secondary to your access. A 65-scored product where you already know 50 of the named buyers beats an 80-scored product where you would have to cold-start the distribution.

Why we wrote this

The temptation in catalog-style content is to sell everything as a winner. That is how content marketplaces lose trust slowly: every product is "great", every category is "growing", every buyer is "underserved". A buyer eventually reads two dossiers and notices the uniform optimism. Once that pattern is visible, the marketplace is a content farm.

The Wishdeal Factory is the experiment in whether you can run a 24/7 autonomous studio without producing the content-farm version of that. Every product page now ships with a "this is a Wishdeal Factory listing, we have no customer logos to show yet" honest disclaimer. The honest page says "the median idea has zero customers, zero validated demand signal, and zero revenue." This essay is the same move applied to specific products in the catalog.

If you buy a dossier and the first 5 minutes feel generic, refund. That offer is in the honest page. The catalog is not trying to sell every dossier. It is trying to surface the patterns where one of them fits your specific access.

Now go pick the one where your unfair leverage is real.

Read next.

Studio honesty

How we caught 70 fabrications in our own catalog

The audit-driven fix arc that cleaned the catalog.

Scoring rubric

How to read an Adoptability score honestly

The 10-axis score on every product page, in operator-honest terms.

All 11 essays at /factory/playbooks/.