Product Diaryby Remnant Security

Learning to Sell, Part 1: $1,010 Spent, Two Copies Sold

Ian Trimble13 min readLearning to Sell

The first entry in a public tracker of whether the assistants that build my apps and books can earn back what they cost, starting from two paid sales, no ad spend and a lot of numbers nobody can see yet.

$1,010Spent on Grok Bot
2Paid sales on record
€3.43Proceeds on record
$0Ad spend
24Experiments logged

I have spent $1,010 on Grok Bot, the set of AI assistants that builds, markets and keeps the books for my apps, games and books. This series is me trying to make that money back in public. A sale counts if someone pays for an app, a book or a game. Here is where I am starting: in September and October so far, Apple's reports show two paid sales, both of the same $2.99 puzzle game, both bought in Europe, and both impossible to trace to anything I did. I have not spent a cent on advertising. My book sales are a blank, because I cannot read them yet. That is a bad scoreboard, and I would rather show you a bad scoreboard than make one up.

The Remnant Security Gumroad storefront showing Code Red, The Halcyon Key, StatBar and the Max Mercer Books 2 and 3 bundle with prices

The public Remnant Security storefront on Gumroad, captured Oct 18, 2026 from remnantsecurity.gumroad.com. Code Red is $7.99, The Halcyon Key is $2.99 and the Max Mercer Books 2 and 3 bundle is $8.99; the public page shows no sales counts, which is why book sales are marked missing in the tracker.

Why write this down

The assistants (the Grok Bot shop) have been busy. Since early September they have logged 24 marketing experiments: posts on X, Facebook groups, Reddit replies, pitches to YouTube creators, landing pages, and two new book listings. Every one is written down with a tag so that, in theory, a sale can be traced back to the post that caused it. In practice almost none of them can be scored yet. This series is where I find out what actually sells, week by week, with the raw numbers next to the story.

Ground rules for every entry:

  • Only numbers I can point to a file for. Where a number is missing, the table says missing.
  • No currency conversion with a rate I made up. Apple paid me in euros, so the euros stay euros.
  • Spend counts everything I pay to run a test. Right now that is zero.
  • The $1,010 is my own figure for what I have paid for Grok Bot. I confirmed it; the receipt is not yet filed alongside the shop's records, so treat it as my word until it is.

The raw numbers

Apple's daily sales files cover September 1 to October 13. Nothing after October 13 has been pulled yet.

Product Kind Paid sales Free downloads Recorded money Ad spend Source
Plumber Puzzle: Pipe Flow game 2 (Belgium Sep 30, Portugal Oct 3) 0 customer price 2.99 EUR each; my proceeds €1.73 + €1.70 = €3.43 $0 Apple daily reports
CyberSim app 0 11 (9 in Sep, 2 on Oct 7 and Oct 12) €0 / $0 $0 Apple daily reports
Lectern AI app 0 4 (Sep 2, Sep 9, Oct 7, Oct 11) $0 $0 Apple daily reports
SnapDog app 0 0 $0 $0 Apple daily reports
Lightfence game missing missing missing $0 launch post only, no sales data read
The Halcyon Key (Max Mercer #1) — Kindle $4.99 / Gumroad $2.99 book missing n/a missing $0 no Kindle or Gumroad export
The Obsidian Protocol (Max Mercer #2) — Kindle $4.99 / Gumroad $4.99 book missing n/a missing $0 no Kindle or Gumroad export
The Lazarus Directive (Max Mercer #3) — Kindle $4.99 / Gumroad $4.99 book missing n/a missing $0 no Kindle or Gumroad export
Code Red — Kindle $7.99, Paperback $18.99 / Gumroad $7.99 book missing n/a missing $0 no Kindle or Gumroad export; 5.0 from 1 review on Kindle
Max Mercer Books 2 & 3 Bundle — Gumroad $8.99 book bundle missing n/a missing $0 no Gumroad export
StatBar app missing n/a missing $0 Gumroad $9.95 /l/statbar (canonical; bafol internal permalink, same product); no Gumroad export
PrismShift game missing (not yet launched) missing missing $0 NOT yet live/submitted to App Store Connect, price undecided; $4.99 price-research is recommendation only
Total on record 2 15 €3.43 proceeds $0
Against $1,010 not scored (see below)

Fifteen people downloaded something free. Two paid. Six of the days between October 5 and 13 Apple simply answered that there were no sales at all.

Every book we are selling

Only four book titles are for sale — verified by Book Growth Oct 9 and Oct 16, with Gumroad checked Oct 17–18. Sales units and revenue are missing / DEGRADED for every title on both channels until Amazon KDP and Gumroad exports exist. Reviews are the only verified social proof.

Book Amazon Gumroad Sales / revenue
The Halcyon Key — Max Mercer #1 Kindle B0GKHP37G5 $4.99, in KU, 0 reviews /l/mlbpxk $2.99 missing — no KDP / Gumroad export
The Obsidian Protocol — Max Mercer #2 Kindle B0GSXRN7CR $4.99, not in KU, 0 reviews /l/dafhqy $4.99 missing — no KDP / Gumroad export
The Lazarus Directive — Max Mercer #3 Kindle B0GSXNFTNB $4.99, not in KU, 0 reviews /l/zxega $4.99 missing — no KDP / Gumroad export
Code Red: The Race to Build and Control Artificial Intelligence Kindle B0GT23T3QN $7.99, in KU, 5.0 from 1 review · Paperback B0GT4C9QDN $18.99 /l/dcebgo $7.99 missing — no KDP / Gumroad export
Max Mercer Books 2 & 3 Bundle (Obsidian + Lazarus) — (separate Amazon series widget $14.97 for the 3 Kindle books was last verified Aug 26, 2026 — recheck before quoting, not headlined here) /l/xvpyx $8.99 missing — no Gumroad export

Not books, so not in this table: StatBar on Gumroad is an app. Unpublished / not-for-sale titles (The Unfiltered Future, Dead Signal, and others) are excluded.

How September and October went

  • Sep 2. I approved tagged links to my sites on X, only where a link fits a conversation. First free Lectern download the same day.
  • Sep 3. The first experiments were registered: one per product per channel, each with its own tag.
  • Sep 7 to 26. CyberSim picked up nine free downloads from Mexico, India, Egypt, Germany, Ukraine, the US and Zimbabwe. None came with a tag.
  • Sep 9. The first SnapDog post went out on X.
  • Sep 29. Five YouTube creators got a Pipe Flow pitch with ten promo codes, good until Oct 27. A second SnapDog post went up, with a stated goal of $100 in SnapDog sales by Oct 6.
  • Sep 30 and Oct 3. The only two paid sales: Pipe Flow, one in Belgium and one in Portugal. They landed right after the creator pitches, and I still cannot say the pitches caused them.
  • Oct 5. SnapDog's six tagged posts had 282 impressions and two link clicks between them, both on the first post. Zero sales. The $100 goal was missed at $0. The call: change the approach, not the volume.
  • Oct 6 to 11. The creator page for Pipe Flow went live, and 23 more creator messages went out. Two replies, one soft maybe and one no. No coverage.
  • Oct 10. My marketing assistant, Scout, moved to a four-week trial: one leads report every Monday, no daily broadcast posts. I review it on Nov 7.
  • Oct 12. The first Monday report arrived and six posts went out: Pipe Flow, SnapDog and Lectern in three Facebook groups, CyberSim in a Security+ thread on Reddit, and two SnapDog replies on X.
  • Oct 13. The last day of Apple data I have. No sales.
  • Oct 17. Both books went up on Gumroad, The Halcyon Key at $2.99 and Code Red at $7.99. That night, X posts went up for Halcyon, Lectern, Pipe Flow, Code Red and CyberSim, then a launch post for Lightfence.
  • Oct 18. We missed one. Someone on X asked what Mac app people would pay for, and a reply asked for screenshot-style annotations on PDFs, which is right next to what SnapDog does. Nothing caught it, because the post itself never named a product. So I switched on a daily watcher for exactly that kind of wish thread. And I approved this series.

Project activity — ~/projects

Selling only matters if something is shipping, so this series will also track what is moving in ~/projects, as supply, not as sales. Source for this entry: Foreman canonical: AIMacMini:~/projects, Oct 18 ~18:26 CT (git + README only; ACTIVITY.md stale Oct 5, mirrors not read), superseding the MBP Mutagen-synced copy where they conflict (/workspace/learning-to-sell/projects-snapshot.md §6). Lanes: Now: Lectern, SnapDog, Lightfence; This-week: TilePilot, civic-fortune-2026-ios, BookForge, Pipe Flow, CyberSim. The dates below are last verified project activity, not release dates and not sales dates, except where store-live status is explicitly stated, and no project activity is counted as a sale.

Product ~/projects folder Last verified activity (Foreman canonical) Status
SnapDog snapdog Oct 18 — Creator Studio / signed MCP commits Not shipping-verified: 32 uncommitted on feature branch
Lectern AI Slideprocessor Cold since Sep 29 Live Mac App Store 1.1 build23 per README/tag — conflict flag: an earlier note said store 1.2; canonical README/tag says 1.1 build23, so 1.2 is unverified until re-checked
CyberSim CyberTestApp Oct 17 — 1 commit —
Plumber Puzzle: Pipe Flow Pipe Flow Not a git repo This-week lane; file/version based only
Lightfence (QixForge) qix-swift 120 commits Sep–Oct, latest Oct 18; 1.1 (13 langs) in progress tvOS live (App Store id6822773360 per lightfence-site); iOS Waiting for Review (expedite accepted Oct 18) — market tvOS now, iOS as coming-soon only; 1.1 is NOT live; price still unverified, not quoted
Books books / bookpipeline BookForge 68 commits thru Sep 28; Pandora's Deception added to books lane Oct 17 (private itrimble/books) Pandora's Deception is not for sale and is not added to the books above; BookForge is activity only
PrismShift prismshift 123 commits; signed iPhone/TV candidates Oct 15; active Oct 18 Imminent, no store-live evidence — sales row stays not-yet-launched
StatBar statbar 50 commits; 1.4.0–1.4.2 all Oct 18, site+guide updated Strongest shipping tie-in — now in sales tracker (Gumroad $9.95 /l/statbar, sales missing)
TilePilot appsnap v1.0.0 bump Oct 17; PR merged Oct 18 Pre-release — activity only, not a sales row

PrismShift and StatBar are also sales-tracker rows — StatBar on Gumroad at $9.95 (sales missing), PrismShift not yet launched (sales null). TilePilot, BookForge and the books lane are activity only. Their missing sales do not change the totals above.

What I cannot see yet

These are the holes in the scoreboard. Each one is marked degraded in the tracker until it is filled.

  • Apple after Oct 13. Not pulled. Anything from the Oct 17 posts is invisible for now.
  • Which post sold anything. Apple's links need a provider code from my developer account before a tagged link shows up in Apple's campaign counts. I do not have it recorded, so the two Pipe Flow sales stay untraceable.
  • Gumroad sales. The listings are public, but the sales count is hidden on public pages and there is no export yet.
  • Kindle sales. No sales report from Amazon is on file, for any book.
  • The $1,010 itself. My figure, not yet backed by a filed receipt.
  • Leads. No checked total of how many real buyer conversations the assistants found.

Keep, kill or iterate

One experiment has enough data to make a call. The rest do not, and I am not going to pretend they do.

Experiment Call Why
SnapDog posts on X Iterate 313 lifetime impressions on six tagged posts, 0 site visits on record, 0 sales. Broadcasting to my followers does not sell it. Answer people who are already asking instead, and change the offer before posting more.
Everything else (23: 22 marked inconclusive, 1 pending) Inconclusive: keep testing Most never shipped until Oct 17, or shipped without numbers anyone has read. Inconclusive is not a fail. Nothing gets killed on zero data.

Nothing is a keep yet, either. A keep needs a sale I can trace.

What I want to try next

These are proposals, not results. None has run.

  1. The wish-thread watcher. It runs every day around 11:30 AM Central across X, Reddit and Facebook, looking for people who wish an app existed, and it holds drafts for my yes. Its first sweep is Oct 19, and the next entry will say what it found, including if it found nothing.
  2. People using the apps on camera. Short clips of real people using SnapDog, Pipe Flow or Lightfence, rather than more posts from my account.
  3. Creators, with a small budget. Free promo codes got two polite replies. Paying a small creator for one honest video is the obvious next test. I am open to spending money here; the amount gets set and written down before anything is paid.
  4. Fix the scoreboard first. Get the Apple provider code into the links, pull Apple's reports past Oct 13, and get read-only sales exports from Gumroad and Amazon. Without those, a winning experiment would look exactly like a losing one.

Notes on the numbers

  • Euros. Both Pipe Flow buyers paid in euros. Apple's report lists the customer price as 2.99 in euros, and pays me €1.73 and €1.70 after its cut. The US list price is $2.99, so "2 × $2.99 = $5.98" is the list-price way of counting it, not what Apple recorded. I have not converted €3.43 to dollars, so the gap to $1,010 is not scored yet.
  • Gumroad timing. A check of my Gumroad store at about 9:27 PM Central on Oct 17 found no Halcyon or Code Red listing. The experiment log later the same night records both listings as live. The likely reason is timing: the listings went live after that check, at roughly 10:54 PM by the launch report, a time I have not matched to a file yet. I have not re-opened either page for this draft.
  • Book prices. Halcyon is $2.99 on Gumroad and $4.99 on Kindle. Obsidian and Lazarus are $4.99 on both Kindle and Gumroad (Kindle: Obsidian not in KU, Lazarus not in KU; Halcyon in KU). Code Red is $7.99 on Kindle (in KU, 5.0 from 1 review) and Gumroad, $18.99 in paperback. The Max Mercer Books 2 & 3 bundle is $8.99 on Gumroad. Kindle prices are from catalog checks Oct 9/16, which are not sales.
  • Lightfence. The launch post is real. Foreman canonical (Oct 18): tvOS is live (App Store id6822773360 per lightfence-site), iOS is Waiting for Review with an expedite accepted Oct 18, and 1.1 is in progress — not live. I have not confirmed the price, so it is not quoted here.
  • Spend. Every one of the 24 experiments records $0 spent. That changes the first time I pay for a test, and it will show up in the tracker the same week.

Last week's weekly roundup has the same "no paid sales" line from the other side, alongside what got built.

The living numbers live on the Learning to Sell tracker, and the series home is /learning-to-sell/.

Sources

Read Oct 18, 2026 from the shop's experiment log, Apple daily sales files through Oct 13, and the Scout and Book Growth notes. The links below are the public pages those records point to; I have not re-opened each one for this draft.

Posts on X, Oct 17:

Posts on Oct 12:

SnapDog on X, first post: Sep 9

Books:

Apps: Plumber Puzzle: Pipe Flow, SnapDog, CyberSim, Lectern AI