Page one is the summary: eight working systems, priced at what a freelancer or small agency would charge to build them today. Everything after it is the homework — fifty years of agency pricing, software costs, and inflation math, with every figure tied to a named source so the number can be checked, not just believed.
Seventy-one project folders were opened in the workspace during the period; the eight systems below are the ones that carry real replacement value. Each is priced against a current market band documented in the appendix — the cross-check column says which one, so any line can be audited against §4.
| System | What it is | Status | Low | High | Priced against |
|---|---|---|---|---|---|
| Family Points Board | Live web app on Fly.io — passphrase login, real database, kid points for the household. | Live | $8,000 | $20,000 | Simple-to-medium web app band, §4 |
| milyunn.com Hub + Portfolio | Hub site, 14-page portfolio, Figma wireframe — the front door for the hosted apps. | Built | $5,000 | $12,000 | Professional website + design rates, §4 |
| Empire + ATS Blackboard | Household money system, plus the Accurate Tech Service blackboard and the pipeline that captures how the boss runs jobs. | Running | $15,000 | $30,000 | Medium web app + data-pipeline bands, §4 |
| The Court + Hyperbolic Time Chamber | A multi-model AI council with a ratified charter, admission tests, and corruption audits — a small AI lab. | Running | $20,000 | $50,000 | AI consulting project band, §4 |
| Data-Lake Distillation | Bronze → silver → gold → platinum pipeline: raw captures transcribed, indexed, and distilled into a searchable personal ledger. | Running | $15,000 | $35,000 | Data-pipeline project band, §4 |
| Vesper IP & Brand Kit | Character system for a black-resin mascot: emblem, marketing set, brand kit, LoRA training sets (positive + annotated negative), product-line draft. | In progress | $15,000 | $35,000 | Brand identity system + merch design, §4 |
| Household OS | Life operating system: task board, daily calendar reasoning loop, kid logistics, health-telemetry reports, school-records push. | Running | $10,000 | $25,000 | Automation consulting + app bands, §4 |
| Connector Army + Sweeps | A bench of free-tier AI/automation connectors, daily intel and security sweeps, inbox and password audits, client job systems. | Running | $10,000 | $25,000 | AI automation rates, §4 |
| Total replacement cost | $98,000 | $232,000 | |||
Totals are the sum of the eight rows shown — nothing else is folded in. Raw workspace file counts (32,905 files) were deliberately not used as a value metric; they include caches and tool output.
Each bar runs from the conservative freelance price (left) to the small-agency price (right), in thousands of dollars.
Source: page-one tally above. Ranges, not point prices — see the method note and §4 for the market bands behind them.
Each system is priced at what it would cost to hire the work out — not at the value of the hours spent chatting with an AI. That is the number a client or an employer would actually face.
No authoritative standard price exists for any of these services in any era — real rates vary 10–50× by scope, seniority, and geography (§7). Every figure is a low–high band.
Overlapping systems are priced once, at the scope that was actually delivered. Drafts and in-progress work (the Vesper product line) are labelled as such, not priced as launched products.
In the 1980s, a company that wanted this stack — brand, software, research, content, operations — hired an agency on a 15%-of-media commission, a dev shop by the month, and consultants by the hour. In 2026, one person in Mt. Carmel did it from his phone for the price of a streaming subscription. The rest of this document is the proof, priced era by era.
For roughly a century, advertising agencies were not paid a project fee at all. They took a 15% commission on media billings — the client paid for the ads, the media outlet kept 85%, and the agency kept 15% — plus a production markup of about 17.65%, the figure that makes a markup equal a 15% commission. That model held from the 1880s into the 1980s, then unbundled: after Omnicom's formation in 1986, fee-based research, PR, and creative/media boutiques split the full-service agency apart, and by the ANA's 2004 compensation survey fewer than 10% of advertisers still used a pure media commission.[1][2][3]
Why it matters for this ledger: in the 1970s, a “million-dollar campaign” did not mean a million-dollar agency fee — the agency's take was roughly 15% of the media placed, plus production markup. Any historical total has to say whether it is media, fee, or all-in. The famous prices below are the documented exceptions that prove how wide the range runs.
| Item | Date | Cost then | What the number actually is |
|---|---|---|---|
| Nike “Swoosh” logo | 1971 | $35 | Student designer Carolyn Davidson, $2/hr × 17.5 hrs (≈ $221 in 2019 dollars). She later received 500 shares of stock. An outlier, not a market rate.[10] |
| Coca-Cola “Hilltop” commercial | 1971 | $250,000 | Production cost of “I'd Like to Buy the World a Coke” — described as the most expensive ad ever made at the time; ≈ $1.5M in today's money per the source.[12] |
| National Airlines “Fly Me” campaign | 1970s | ≈ $9M / year | Annual spend on the campaign, agency Wells Rich Greene. Media-era money, not a design fee.[11] |
| NeXT logo — Paul Rand | 1986 | $100,000 | A single logo, at a premium designer's price, for Steve Jobs.[9] |
| BP rebrand | 2000 | $211M+ | Total rebrand rollout, not a design fee — signage, stations, everything. Sources put it at $211M–$221M and differ on the year (2000 vs 2008 retellings). Do not compare it directly with a logo fee.[9][10] |
| Pepsi branding package | 2008 | $1,000,000 | Arnell Group's complete branding package — not the logo alone.[9] |
Read the spread honestly: $35 to $1,000,000 for “a logo” across four decades. Famous-logo prices prove the range exists; they are not a rate card. What a normal business pays today is in §4.
No reliable published fee schedule for 1970s–1990s copywriting projects was found in the research. The documented economics of that era are the 15% commission above and the named transactions — so this ledger does not back-fill a 1980s copywriting rate by inflating today's prices. Where a number doesn't exist, the ledger says so (§7).
In 1994, a basic corporate website could run about $10,000 — and that was for pages, not software. Through the late 1990s the price of a single website fell from $7,500–$12,000 toward $1,000–$5,000 by about 2010, as templates and builders commoditised the brochure site.[13][14] But the thing on page one of this ledger is not a brochure site. A database-backed app with user logins barely existed as a commodity product in 1995 — and today it is a $15,000–$60,000 medium build that takes a team about six months.[5][6]
Typical price range for a single website, by period (thousands of dollars, nominal — not inflation-adjusted).
Source: CoolWebTemplates' historical price series[14] — an old blog source, carried here as indicative only, exactly as the research flags it. The 1994 anchor (~$10,000 for a basic corporate site) comes from a UniForum / IDC-era account.[13] For scale today: a professional 15-page small-business site runs $2,250–$10,000 (WebFX via JLB).[15]
| Tier | Typical cost | Typical timeline | Basis |
|---|---|---|---|
| Simple web app | $5,000–$15,000 | ≈ 3 months | CodeBridge tiers; Spdload's model prices the simple build at 480 hrs × $40/hr = $19,200.[5][6] |
| Medium web app | $15,000–$60,000 | ≈ 6 months | Spdload model: 960 hrs = $38,400 at $40/hr. This is the band a solo builder's avoided cost should be framed against.[5][6] |
| Complex web app | $60,000–$250,000+ | ≈ 9 months | At U.S. rates (~$150/hr) a complex build easily exceeds $200,000. General custom software runs $50,000–$450,000 initial development.[5][6][22] |
| Average project (survey) | $66,499 | ≈ 9 months (≈ $7,139/mo) | Clutch survey figure, via CodeBridge.[5] |
The Points Board on page one — logins, a real database, live hosting — sits between the simple and medium tiers above. In 1995 there was no shelf price for it at any number, because the product category didn't exist yet. The fair “then vs now” is not brochure-site money; it's the medium-app band, and the team months that come with it (§6).
These are the bands page one was priced against. All of them are industry ranges/estimates, not fixed prices — U.S. rates sit at the top of global marketplace ranges, and the same job can legitimately quote 10× apart depending on who is holding the pencil.
| Service | Low | High | Source |
|---|---|---|---|
| Logo only | $500 | $3,000 | Dribbble, updated May 12, 2026[4] |
| Basic brand identity (logo, palette, type) | $1,000 | $5,000 | Dribbble[4] |
| Brand identity system (logo system, colours, type, messaging, guidelines) | $5,000 | $20,000+ | Dribbble[4] |
| Full branding package (strategy, identity, messaging, assets, guidelines) | $11,000 | $70,000+ | Dribbble[4] |
| Enterprise brand development / rebrand | $50,000 | $200,000+ | Dribbble; global transformations $150,000–$500,000+[4] |
| Copywriting — freelance hourly | $20/hr | $150/hr | Entry $20–30, intermediate $30–50, expert $50–150 (global marketplaces)[25] |
| Copywriting — per project | $250 | $25,000+ | Retainers $1,000–$5,000+/mo; per word $0.10–$1.00[26] |
| Sales page — direct-response copywriter | $5,000 | $25,000+ | Top tier $25,000–$50,000+, sometimes plus royalty[27] |
| Marketing agency retainer (2025) | $3,000/mo | $25,000/mo | Median $7,500/mo mid-market; comprehensive programmes $12,000–$35,000/mo, media spend separate. Full-service: SMB $2,500–$10,000/mo, enterprise $15,000–$50,000+/mo[28][29] |
Standard branding timeline today, for reference in §6: 4–12 weeks standard; accelerated 2–4 weeks (+10–30%); rush 1–2 weeks (+30–50%). A comprehensive identity at Chicago agencies runs 9–19 weeks; enterprise, 6+ months.[4][30]
| Service | Low | High | Source |
|---|---|---|---|
| Web development — U.S. freelance hourly | $50/hr | $150/hr | North America / W. Europe specialists. U.S. full-stack: entry $40–60, mid $75–120, senior $150–250+[5][23] |
| Web development — U.S. agency hourly | $100/hr | $200+/hr | U.S.-based agency band[31][32] |
| Web app — project price | $5,000 | $250,000+ | Simple → complex tiers, §3. SaaS MVP anchor: $25,000–$75,000+[5][33] |
| AI / data consulting — independent hourly | $75/hr | $350/hr | Junior freelance → senior independent. Boutique firms $125–$500/hr; Big 4 $250–$1,000+/hr[20] |
| Data engineer — U.S. freelance hourly | $92/hr | $145/hr | SecondTalent, Sep 2026[21] |
| AI / data project — project price | $5,000 | $150,000+ | Small (audit, single dashboard) → large transformation; mid-scale $15,000–$50,000; retainers average ≈ $10,000/mo ($3,000–$20,000)[19] |
Floor for contrast — not a U.S. rate: low-code/offshore AI automation runs $20/hr (no-code specialist) to $80+/hr (senior AI engineer), and global marketplaces list web developers at $15–$50/hr. Page one was priced against the U.S. bands above, not this floor.[37][38]
| Service | Low | High | Source |
|---|---|---|---|
| Single merch graphic | $49 | $299+ | Marketplace starter (1 graphic / 1 product) → custom illustration add-on. Tee or hoodie artwork $149+, cap/mug $99+[34] |
| Tech pack, per style | $100 | $800+ | Simple tee → technical outerwear; hoodie/trousers $200–$400[35] |
| Packaging design project | $200 | $2,000+ | Basic → premium multi-variation. Hourly: freelancer $20–45, specialist agency $100–200[39][36] |
| Product designer — U.S. | $60/hr | $120+/hr | Day rate $400 → $800+[40] |
Cross-check against page one: the Vesper line ($15,000–$35,000) covers a brand identity system ($5,000–$20,000+) plus character and merch design work — inside the bands above, not above them. The Court and the data lake ($15,000–$50,000 and $15,000–$35,000) sit inside the AI/data project band ($5,000–$150,000+, mid-scale $15,000–$50,000).
The claim going around is that a Fortune-500-scale million-dollar budget from back then is “three million in today's money.” Run through the Bureau of Labor Statistics' CPI-U, the honest version is: roughly $4 million today — $3.81M on the last complete year (2024), $4.09M by September 2026. So the claim is right on a 2024 basis and slightly understates it on the latest monthly figures.
Method: value today = amount × (later CPI-U ÷ earlier CPI-U). Annual-average CPI-U used: 1980 = 82.41 · 1990 = 130.66 · 2000 = 172.20 · 2024 = 313.69 (BLS series via OfficialData).[8] September 2026 monthly CPI-U = 336.940 (BLS via FRED, via PointsFrog).[7] A full-year 2026 average does not exist yet as of Oct 24, 2026, so every “2026 dollars” figure names its basis. CPI-U covers ~93% of the U.S. population and an average consumer basket — it is not a price index for agency labour specifically.
$1M spent in each year, converted to 2024 dollars (last full-year average) and September 2026 dollars (latest month).
Source: BLS CPI-U via OfficialData and PointsFrog.[7][8] Worked cross-check from the source itself: $100 in 1990 × 336.940 ÷ 130.7 = $257.80; $50,000 in 2000 ≈ $97,834 in September 2026 dollars.
| Amount | In 2024 dollars | In Sep 2026 dollars | Factor (Sep 2026) |
|---|---|---|---|
| $1 in 1980 | $3.81 | $4.09 | 4.09× (+308.9%) |
| $1,000,000 in 1980 | $3,806,000 | $4,089,100 | 4.09× |
| $1 in 1990 | $2.40 | $2.58 | 2.58× (+157.8%) |
| $1,000,000 in 1990 | $2,401,000 | $2,578,000 | 2.58× |
| $1 in 2000 | $1.82 | $1.96 | 1.96× (+95.7%) |
| $1,000,000 in 2000 | $1,822,000 | $1,956,700 | 1.96× |
Factors: 1980 ×3.806 / ×4.0891 · 1990 ×2.401 / ×2.578 · 2000 ×1.822 / ×1.9567 (CPI-U, BLS).[7][8]
Pepsi's $1,000,000 branding package (2008) and Paul Rand's $100,000 NeXT logo (1986) were already modern money next to the 1971 bargains — Davidson's $35 swoosh is the famous exception, not the rule. The era gap that actually matters is structural, not inflationary: in 1980 the work required an agency, a media buy, and a production house; in 2026 the same categories of work clear through one person and a phone — priced in §4, timed in §6.
The strongest honest claim the evidence supports: AI assistance roughly halves the time on well-scoped coding and content tasks, and can compress a first prototype from weeks to days. It does not reliably compress a whole team product — brand validation, review, debugging, rollout — from months to minutes. Both halves of that sentence are in the table, because a ledger that only prints the flattering half isn't a ledger.
| Task | Then (team) | Now (AI-assisted) | Evidence |
|---|---|---|---|
| Brand identity project | 4–12 weeks standard; comprehensive 9–19 weeks; enterprise 6+ months | AI logo generators $0–$100 and marketplace concepts in hours/days at the low end. A full human-led identity still takes weeks — “minutes” applies to a first draft, not a validated identity system | Agency timelines & price guides[4][24][30] |
| Small web app / MVP | Team build ≈ 3–9 months (simple → complex); Clutch average ≈ 9 months | Vendor claims of an MVP in 1–2 weeks vs 6–8, and prototypes “in a single day” | Vendor claims, not controlled trials[6][41] |
| One defined coding task | Control group: 2 hrs 41 min to build a JavaScript HTTP server | Copilot group: 1 hr 11 min — 55.8% faster (n = 95) | Controlled lab study (Peng et al., 2023). Caveat: a defined exercise, not production work[16] |
| Real-world coding throughput | Pre-AI baseline | +26.08% completed tasks (pull requests / commits / builds), larger gains for less experienced developers | Randomised field experiment, Microsoft / Accenture et al. (Cui et al., 2024)[16] |
| Writing a marketing article | Non-AI writers: 3 hrs 48 min average | AI users: 3 hrs 24 min — 10% less time | Orbit Media survey, n = 688 content marketers, 2025[18] |
| Content production, weekly | Manual production | 116 minutes saved per marketer per week (≈ 100 hrs/yr) | Vendor survey — Adobe GenStudio, 350+ marketers[42] |
| Research / content pipeline | Manual competitive analysis ≈ 30 min; keyword research ≈ 3.5 hrs; blog post 4.2 hrs; 50 product descriptions 8.5 hrs | Claimed: brief in 2 min; keywords in 30 min; post in 1.8 hrs; descriptions in 2.2 hrs | Weakest evidence on this page — aggregator/vendor claims, directional only[43][44] |
A 2025 randomised controlled trial (METR) found 16 experienced open-source developers working on their own repositories were 19% slower with early-2025 AI tools (95% CI: 2–39% slower).[17][45] And in Atlassian's State of Developer Experience surveys, 87% of developers say AI helps them ship faster, yet 72% report the same or increased delivery friction — and the share saving ≥ 10 hrs/week fell from 68% (2025) to 25% (2026).[46]
What that means for page one: pricing a solo builder's output at the avoided agency fee is fair for the deliverables that exist — but the review, debugging, and rework time is real, and this ledger's 28 days of evenings were not free. They were just paid in a different currency.
Marked plainly, per the rules of this ledger: a missing number is shown as missing, never filled in.
| 1970s–1990s copywriting fee schedules No reliable published schedule was found. The era's documented economics are the 15% commission (§2) and the named transactions — nothing else is claimed. |
| A 1990s price for “a database-backed app with logins” Does not exist as a like-for-like figure — the 1990s prices in §3 are brochure websites. The comparison is framed against today's medium-app band instead, and says so. |
| A single “standard” price for any of these services None exists in any era. Every figure here is a range that varies 10–50× by scope, seniority, and geography — which is why page one is a range too. |
| The BP rebrand total Sources disagree: $211M vs $221M, and the year is retold as both 2000 and 2008. Shown as “$211M+ rollout” with the disagreement attached. |
| A full-year 2026 inflation figure Not yet published as of Oct 24, 2026. §5 therefore gives both bases — the 2024 annual average and the September 2026 month — rather than one blended “2026” number. |
| How strong the rate sources are Several current-rate sources are vendor/agency blogs or marketplace compilations, not audited price surveys. The strongest primary anchors are BLS CPI-U, the ANA compensation survey, the Peng/Cui experiments, and Orbit Media's 2025 survey. Each source below is labelled by how it was read. |
Research read on Oct 24, 2026. Verified live means the page was opened and read in full; Index means the figure comes from a search-result record of the page. Figures are industry ranges/estimates unless labelled as a named, dated transaction.
Source [47] is listed for completeness — it informed the research's rate cross-checks but is not cited inline above.