# The Investor Nobody Paid

*Debora den Iseger · October 2026*

*A thesis on ownership, attention and agents. What if the shopper is your shareholder?*

## Abstract

For twenty years we have said that if the product is free, you are the product. This paper argues that is wrong in one important way: **you were never the product. You were the investor nobody paid.** Every purchase, search and return visit puts value into the companies we use; almost none of it comes back.

Three things make this the moment to change it. People are saying no when they are finally asked. Agents are starting to shop on our behalf, which forces the question of who they work for. And the infrastructure for those agents is being poured now, before the concrete sets.

Using public filings, national spending data and real share prices, the paper prices one ordinary year of a shopper's life, proposes **dividend brackets** that scale with what a company keeps, maps who inside each company could say yes, and ends with steps shoppers and enterprises can take today. None of it is radical. It is **common sense innovation**: ideas that should never have needed inventing, rebuilt for a world that forgot them.

**Contents.** I. Taken, not given · II. Who does my agent work for? · III. The investor nobody paid · IV. One ordinary life, priced · V. Dividend brackets · VI. Who can decide · VII. The cascade & the organizing · VIII. What to do today · Conclusion · Epilogue: the community we once were · Methods & assumptions · Sources

## I. Taken, not given

In the last two months, fifteen emails tried to win me back. A turtleneck, back in stock. Wedding guest dresses, restocked. A cutting board, "finally" available again. One cookware brand wrote five times in four weeks; its latest subject line begins "Trying Again." I opened none of them.

My phone tells the same story in text messages, from brands that somehow found my number and now check in like old friends. Each of them knew something about what I had looked at. Each of them kept paying to reach me. Somewhere, someone was paid for every one of those messages. It was not me.

Most of us never cared much where our data went, and the industry took that as permission. But not caring was never the same as not valuing. People did not care because nobody told them what their data was for. The intent was simply "forgotten."

Then someone asked. When Apple began requiring apps to request permission before tracking, somewhere between a third and a half of iPhone users said yes, depending on whose numbers you read. **The moment people were asked, roughly half the data disappeared.** And asking is becoming automatic: twelve US states now require companies to honor Global Privacy Control, a browser signal that says, in effect, do not sell my data, and some browsers send it by default.

There is a second signal hiding in the same numbers. Consent to tracking has risen since those first prompts, and analysts link part of the difference between apps to how clearly they explain the exchange. **The moment people were asked, half said no. The moment they understood why, more said yes.** That is not apathy. That is a market correcting itself, and asking to be told the intent.

## II. Who does my agent work for?

Agents are beginning to shop, compare and decide on our behalf. They read the reviews, weigh the prices, fill the basket. The question we never asked about our data becomes impossible to avoid about our agents: **who does it work for?**

I have spent years building the systems this question is about, and the honest answer today is: it depends on who is paying. An agent that steers you toward whatever a merchant paid to promote is not your assistant. It is a salesperson with your name on the badge.

Listen closely and you can already hear what the shopper says next:

> My agent works for me now. If you want my loyalty, stop trading on my data.

That is not a protest. It is a negotiating position, and it is the first time the shopper has ever had one.

> **The next shareholder revolt starts at the checkout.**

The fork is already visible. In February 2026, Anthropic promised that Claude would never carry ads. OpenAI was already selling them inside ChatGPT. Four months later, Anthropic became the more valuable company for the first time. I can't prove one caused the other; valuations move on revenue, compute and a hundred other things. But Anthropic is reportedly preparing to go public, and once both report to public markets, we can read the difference in approach quarter by quarter. **This is the race to watch: which way the money moves when someone states their intent.**

![Figure 1, Plate II](https://commonsenseinnovation.tech/paper/img/plate-2.png)

*Figure 1. The chain, reversed. Same shopper, same purchase, opposite direction of value.*

## III. The investor nobody paid

Here is the flip. You were never the product. Products are made, priced and sold. You were doing something else entirely: putting something in. Your attention, your searches, your return visits, your recommendations, the hundred small signals that taught a system what people like you want. You put that in every day, and nothing came back.

In any other part of the economy we have a word for someone who puts something in and expects a return. We call them an investor.

In 2024, Meta and Alphabet, the two companies that know the most about what we want, did something they had never done before. They paid their shareholders a dividend. The people whose searches, scrolls and purchases built those businesses received what they always had: more ads. In the last three months of 2025 alone, Meta reported earning about $16.56 from every person who used its apps each day, across 3.58 billion people.

Now picture the same year with one change. What if the people whose attention built those businesses had been counted among the investors? Not with a cheque that arrives as a rounding error, but with the things investors actually hold. A stake that grows the longer you stay. A quarterly statement showing what your data earned and what it was used for. A vote on what it may be used for next. A share of the dividend, however small, paid to the people who made it possible.

None of this would make anyone rich overnight. That is not the point. The point is who sits at the table. In my last essay I wrote about the 99 percent who were never asked. This is what asking them could look like: not a favour from the one percent, but a seat among them. **An economy that compounds for everyone who feeds it is not a utopia. It is simply what an investment was always supposed to be.**

None of this is as new as it sounds. In 1844, a group of mainly weavers in Rochdale, England, opened a small shop and returned its surplus to members in proportion to what each had bought. The shopper as shareholder is older than the light bulb. Today, Costco pays its Executive members 2% back on purchases, up to $1,250 a year, on one of the thinnest margins in retail. REI returns an estimated 10% a year to its co-op members. One app once paid shoppers in fractional stock of the brands they bought from; it closed in 2022. Others pay people a few dollars a month for their data. Each piece works on its own. Nobody has joined them, and the reason is not technology. We simply stopped thinking this way.

I call it **common sense innovation**: ideas that should never have needed inventing, rebuilt for a world that forgot them. Every shopkeeper on a corner knows a loyal customer is worth more than a new one. We just built an internet that paid everyone in that relationship except the customer.

## IV. One ordinary life, priced

To make this concrete, picture one shopper. Tech-forward, early forties, lives on her phone, buys with intent: five food apps, two health apps, about ten subscriptions, four social platforms. Her year of groceries, eating out, entertainment, clothing, personal care and subscriptions comes to about $17,957, using US household averages.

**If the brands she buys from returned 2% of each purchase as equity, growing 7% a year, she would hold about $14,723 after twenty years. Same spending. Nothing extra.** Start five years later and it ends near $9,000: waiting costs her about $1,100 of future value a year.

![Figure 2, Plate III](https://commonsenseinnovation.tech/paper/img/plate-3.png)

*Figure 2. One shopper, twenty years: compounding, or none.*

Now put real companies against her year. Spread across fifteen she touches every week, from Amazon and DoorDash to Netflix, Uber and two AI assistants, about $7,280 of her spending supports roughly **$31,778 of market value every year**, using each company's market value per dollar of revenue. She owns none of it. The AI companies value her dollar most: her $240 a year at Anthropic supports about $3,563.

Looking back instead of forward: had 2% of that spending been paid in each company's shares every year since 2016, her $1,126 would be worth about $2,351 today. Apple and Microsoft led. Nike and Lululemon lost value. **A share is a stake, not a guarantee.**

![Figure 3, Plate IV](https://commonsenseinnovation.tech/paper/img/plate-4.png)

*Figure 3. Her year, priced by the market, with two lenses: what her spending supports, and ten years of stock back.*

The same model, run for households with less:

| Household | Yearly basket | 20 years at 2% | Value her spending supports, per year | 10 years of stock back |
| --- | --- | --- | --- | --- |
| Tech-forward persona | $17,957 | $14,723 | $31,778 | $1,126 → $2,351 |
| Below average, second fifth of incomes | $11,445 | $9,384 | $20,254 | $718 → $1,498 |
| Low income, lowest fifth of incomes | $8,013 | $6,570 | $14,181 | $503 → $1,049 |

For a household spending $35,000 a year, $6,570 is not a rounding error. It is an emergency fund, a deposit, a first investment account, for the people least likely to own shares today.

**Every assumption here is a dial. If you like the outcome, turn the assumption into a goal & a target.**

## V. Dividend brackets

For anyone reading this from inside a brand, here is the business case, in the language of the balance sheet. **Taken data depreciates.** Every opt-out, every blocked signal, every browser that says no on someone's behalf leaves a model guessing a little more. The guesses compound quietly, until the system is confidently wrong, like the dress that split the internet: a guess about the light, held so confidently it passed for fact. **Given data appreciates.** Data shared on purpose, with the intent understood, is richer and more honest, and it grows with every visit. Customer first is no longer charity. It is the better asset.

How much is fair? A flat 2% of revenue is easy for the giants and ruinous for the thin-margin brands. For Microsoft, Alphabet, Meta, Apple and Netflix, 2% of revenue is only about 4% to 7% of a year's profit. For Starbucks, DoorDash and Nike, which keep about five to seven cents of every dollar, it would eat roughly 30% to 39%.

So think of it like brackets: **the more a company keeps, the more it gives back.**

| Bracket: profit kept per dollar of revenue | Share of profit returned | Companies in our set |
| --- | --- | --- |
| Over 25¢ | 15% | Alphabet, Microsoft, Meta, Netflix, Apple |
| 10¢ to 25¢ | 10% | Spotify, Amazon, Uber, Lululemon, Instacart |
| Under 10¢ | 5% | Ulta, Etsy, Nike, DoorDash, Starbucks |

Not a tax. A dividend that scales, paid from the surplus, not the till, exactly as the Rochdale weavers did it.

Run across these fifteen companies, the brackets return about **$104 billion a year**, twice what a flat 2% of revenue would. Eighty-five percent of it comes from the five companies keeping more than a quarter of every dollar. Starbucks or DoorDash would give about 0.3% of revenue; Alphabet about 8%. The corner shop barely feels it. The giants feel it in proportion.

One honest limit. The giants' profit comes mostly from attention, and attention pays little per person: Meta's bracket share works out to about $2.85 a year for each of its daily users. **Alone, that is a rounding error. Pooled, it is leverage.** That is where Part VII picks up.

![Figure 4, Plate V](https://commonsenseinnovation.tech/paper/img/plate-5.png)

*Figure 4. The more you keep, the more you give back: brackets, not a tax.*

## VI. Who can decide

The obstacle is not permission from the world. It is a yes from a surprisingly small number of people. In several of the companies that know us best, voting control sits with one founder or a handful.

| Company | Share structure | Who controls the vote |
| --- | --- | --- |
| [Meta](https://www.sec.gov/Archives/edgar/data/0001326801/000121465926005093/z424260px14a6g.htm) | Dual class | Mark Zuckerberg, about 61% of voting power |
| [Alphabet](https://www.sec.gov/Archives/edgar/data/0001652044/000130817926000342/goog-20260424.htm) | Three classes | Larry Page & Sergey Brin, about 52.7% combined |
| [DoorDash](https://www.sec.gov/Archives/edgar/data/0001792789/000114036126034688/ny20079824x2_def14c.htm) | Dual class | Founders, about 54.2% |
| [Spotify](https://www.fool.com/investing/how-to-invest/stocks/which-company-owns-spotify) | Shares plus founder certificates | Daniel Ek & Martin Lorentzon, about 73% (reported) |
| [Nike](https://www.sec.gov/Archives/edgar/data/0000320187/000032018726000089/nke-20260715.htm) | Dual class | Class A, mainly Knight family's Swoosh LLC, elects 8 of 11 directors |
| [OpenAI](https://techcrunch.com/2025/10/28/openai-completes-its-for-profit-recapitalization) | Public benefit corporation | Board appointed by the nonprofit OpenAI Foundation |
| [Anthropic](https://www.anthropic.com/news/narasimhan-board) | Public benefit corporation | Directors appointed by its Long-Term Benefit Trust now form a majority |
| Amazon, Apple, Microsoft, Netflix, Starbucks, Uber, Lululemon, Etsy, Instacart, Ulta | Single class | Widely held; the board decides. Jeff Bezos holds about 9% of Amazon |

What each kind of return would take, in broad terms (general picture, not legal advice):

- **Cash or credit rewards, like Costco's 2%**: a management decision. It already exists.
- **Shares bought on the market and given to customers**: management and board, funded from marketing.
- **New shares issued to customers**: the board, sometimes shareholders, and securities lawyers to structure it.
- **A permanent promise written into the charter**: a shareholder vote, which in a controlled company means a handful of people.

The two AI companies deserve a closer look. As public benefit corporations, both may legally weigh stakeholders beyond shareholders. And with an IPO reportedly ahead for Anthropic, **the months before a listing are exactly when such a promise is easiest to write in.** The concrete, before it sets.

## VII. The cascade & the organizing

It only takes one. A small trading app spent years making free trades feel normal. Then, in October 2019, the largest brokers dropped their commissions to zero within about ten days of each other. None of them wanted to. None of them could afford to be the last one still charging. Ownership could spread the same way: the first brand that treats its shoppers as investors makes every other offer look like a bad deal.

Social platforms will be slower to move. You cannot take your friends with you. But you can take your wallet. You can stay for the people and still choose where you buy. That is a revolution almost anyone can join, and it asks you to give up nothing. **It doesn't raise incomes. It spreads ownership.** And the loop closes on its own: a shopper's share is already invested in the brands they love. Every purchase buys a little more of what they believe in.

Big tech organized long ago. Now everyone else is learning to. In September 2026, the European Union offered Canada something new: to become its first associate member. When the biggest players organize, the rest of us find a bigger table too. One person's attention is worth $2.85 a year to a platform. A neighbourhood's attention, pooled, is a negotiating position.

What holds most of us back is not knowledge. It is the quiet feeling of defeat: what can one person do? More than it seems. **Take out the middleman. Buy directly from the maker. Start organizing your attention.**

## VIII. What to do today

Most writing on this ends with the right words and no first step. So here are first steps, small enough to start this week.

**If you are a shopper**

1. **Run the five-purchase audit.** Look at the last five things you bought. For each one, ask who learned something from it, and what came back to you.
2. **Say no by default.** On an iPhone, turn off "Allow Apps to Request to Track" in your privacy settings. In your browser, switch on Global Privacy Control, or use a browser that sends it for you.
3. **Prune the ones you never open.** Unsubscribe from every brand you have not opened in a month. Keep the few you love, and reply to tell them why.
4. **Ask your agent who it works for.** Before an AI assistant buys anything for you, ask where its recommendations come from, and whether anyone paid to be there.
5. **Buy direct.** See it anywhere you like. Then buy it from the maker: their own site, their own door. The shop still wins, and the relationship lives where a share could one day exist.
6. **Organize your attention.** Give it to those who deserve it, directly: the makers you love, on their own channels, not through the feed. Then hand your agent a standing instruction, like the one below.

**Copy this into your AI assistant:**

> From now on, when you help me shop:
>
> 1. Tell me if any result is sponsored, or if anyone pays you or your maker to show it to me.
> 2. Show me the maker's own site or shop first, and buy direct whenever you can.
> 3. When quality and price are close, choose the independent maker or the local shop.
> 4. Use what you know about me for this task only, never to target me later.
> 5. Once a month, list the brands I keep ignoring and offer to unsubscribe me.

One person typing this changes nothing. Millions typing it becomes the most requested feature in the industry, and the companies building agents will have to answer it. Some assistants will answer every line. Some will tell you they cannot say who paid to be in front of you. **That answer is the proof. Question the ones that give it, every single time.**

**If you run a brand or an enterprise**

1. **Write your intent in one sentence.** What you collect, why, and what the customer gets back. Put it where they buy, not in the terms.
2. **Sort taken from given.** List every source of customer data you use and mark each one taken or given. Track that ratio every quarter, like any other asset.
3. **Stop paying to chase silence.** Count the messages your customers never open. Each one is a cost with no return, on both sides.
4. **Move that budget.** Take what you spend chasing silence and fund a reward for the customers who actually come back.
5. **Pick your bracket.** Look at how much of each dollar you keep, and decide what share of profit you would return. Start with your best customers, not all of them.
6. **Pilot one return.** Give those customers something that grows with every visit, not points that expire. Measure who comes back against a group that got nothing.
7. **Make yourself readable to agents.** Clear prices, honest delivery dates, plain policies. When an agent shops for someone, transparency is how you get picked.
8. **Answer the shopper's five questions before they ask.** If a customer's agent asked your store, your app or your own agent the five questions above, would every answer be yes?

**Put this in your next brief:**

> Every agent, storefront and campaign we run must be able to say, in plain words:
>
> 1. Who it serves, who pays it, and what it optimizes for.
> 2. Which data we were given, and which we took.
> 3. What the customer gets back, and when.

The brands that can answer all three will be the ones agents choose. **The ones that cannot are telling you something too. That silence is the proof, and your customers' agents will hear it first.**

## Conclusion: whoever builds it first

The infrastructure for agents is being poured right now: the protocols agents use to talk to stores, the rules for how they pay, the defaults nobody will revisit. Concrete is cheap to pour and brutal to move once it sets. This time, intent can be built in from the start. Every agent can declare who it serves, who pays it, and what it optimizes for. Not buried in terms of service. Visible, like the label on a bottle.

We already know how to do this. Lawyers and financial advisors are bound to act in their client's interest; we call it fiduciary duty. Consumer Reports has proposed "loyalty by design" for AI agents, with agency law as the starting point. **Common sense says a machine spending your money should meet the same standard as a person spending it.** And once an agent is truly on your side, it can collect what your behaviour is worth and bring it back down the chain. A royalty. And royalties, accumulated, become a share.

A sharp reader will ask: says who? Where is the proof that given data outperforms taken data, that shoppers who hold a share come back more often, that a loyal agent sells better than a paid one? Nobody has it yet. **The proof doesn't exist. It belongs to whoever builds it first.**

That is the invitation to every brand reading this. Not to give something away, but to be first to the one position that cannot be copied later: the company its customers own a piece of. It is not radical. It is common sense innovation, and it is waiting for someone to stop calling it impossible.

Everything moves up the chain. What if, this time, something came back down?

## Epilogue: the community we once were

Before there were platforms, there were neighbourhoods. The baker knew your order. The pharmacist knew your mother. When someone fell ill, the street brought soup, and when a family hit hard times, the shopkeeper let the tab run a little longer. Nobody called it a loyalty program. It was simply what a community did for the people who kept it alive.

The Rochdale weavers did not invent a financial product. They wrote that instinct down. Somewhere between the corner shop and the checkout page, we lost the habit, and we have been paying for it ever since: in attention, in data, in the quiet loneliness of being a target instead of a neighbour.

Nowhere is that clearer than in health. In 2025, 26.7 million Americans went the whole year without insurance. A family plan through an employer now costs about $27,000 a year. And most of us agree on the goal, even when we argue about the route: two in three Americans say everyone should have coverage, though the parties split sharply on how to get there.

So picture the same machinery pointed at care. A shopper chooses to send her dividend into her own health savings account. A neighbourhood pools its share into a community health fund for those who cannot open one. A health company returns part of its surplus to the people it serves, the way the weavers did. Insurers already do a version of this by law: when they spend too little of our premiums on care, they must give the difference back, about $1.6 billion in 2025.

Run the brackets from Part V across just fifteen companies and point the pool at the uninsured, and it comes to about $3,900 per person, every year: roughly 40% of a single employer plan. Not coverage for everyone. A contribution toward it, from the companies whose growth we fed. Nobody has to win the argument about government first. It only asks companies, and the rest of us, to act like the community we once were.

**Because if we all do what is best for each other, we could heal our world in new ways.**

## Methods & assumptions

Every figure in this paper is illustrative. Each assumption below is a dial, not a finding.

| Dial | Value used | Basis |
| --- | --- | --- |
| Household spending | $17,957 a year across groceries, eating out, entertainment, clothing, personal care & subscriptions | US Bureau of Labor Statistics, Consumer Expenditures 2024, all-household averages; likely conservative for a Bay Area household |
| Subscriptions | $1,200 a year | Top quarter of US subscribers spend over $100 a month (Bango, 2025) |
| Share returned | 1% to 2% of purchases, as equity | Range once paid by stock-for-shopping apps; Costco pays 2% in rewards today |
| Growth | 7% a year | An assumption, not a promise |
| Spending per company | Illustrative split of the averages across fifteen companies; Uber added as an estimate; subscriptions at list prices | Author's estimate |
| Value supported | Spend × market cap ÷ trailing revenue | stockanalysis.com, 15 October 2026; OpenAI and Anthropic from press-reported valuations and revenue run-rates |
| Stock back | 2% of spend bought at each year-end close since 2016 (or first listing), valued at the 14 October 2026 close | StatMuse and Macrotrends; some older prices include dividend adjustments |
| Brackets | 5%, 10% or 15% of profit by net margin under 10%, 10% to 25%, over 25% | Trailing twelve months to mid-2026; Amazon and Alphabet profits include unusually high recent quarters |
| Lower-income households | Second and lowest fifths of incomes, spending 64% and 45% of average | BLS 2024; spending mix assumed proportional, though lower-income households spend more on groceries |
| Attention | Meta revenue per daily active person, worldwide | Meta's trailing revenue ÷ 3.58 billion daily active people |

Read the numbers as a sketch of direction and scale, not a forecast. Change any dial and the conclusion holds: today, the shopper's stake is zero.

## Sources

**Data & consent**

- [IDFA in 2026: What Has Actually Happened Since App Tracking Transparency Launched (Moburst)](https://www.moburst.com/blog/idfa-att-update/)
- [What Is Global Privacy Control, the Opt-Out Signal 12 US States Now Require You to Honor (Clym)](https://www.clym.io/blog/what-is-global-privacy-control-the-opt-out-signal-12-us-states-now-require-you-to-honor)

**Agents & AI companies**

- [Anthropic pledges ad-free Claude, with a Super Bowl shot at ChatGPT (Axios)](https://www.axios.com/2026/02/04/anthropic-ads-openai-super-bowl)
- [Anthropic Raises at $965 Billion Valuation, Eclipsing OpenAI (Bloomberg Law)](https://news.bloomberglaw.com/private-equity/anthropic-raises-at-965-billion-valuation-eclipsing-openai-1)
- [Anthropic revenue run-rate and IPO plans (Axios)](https://axios.com/2026/08/17/anthropic-revenue-run-rate-ipo-openai)
- [OpenAI valued at $852B after new funding round (AI Business)](https://aibusiness.com/generative-ai/openai-valued-852b-after-new-funding-round)
- [OpenAI completes its for-profit recapitalization (TechCrunch)](https://techcrunch.com/2025/10/28/openai-completes-its-for-profit-recapitalization)
- [The Long-Term Benefit Trust (Anthropic)](https://www.anthropic.com/news/the-long-term-benefit-trust)
- [Engineering "Loyalty By Design" in Agentic Systems (Consumer Reports)](https://innovation.consumerreports.org/engineering-loyalty-by-design-in-agentic-systems/)

**Dividends, filings & control**

- [Meta Platforms initiates quarterly dividend of $0.50 (Seeking Alpha)](https://seekingalpha.com/news/4061622-meta-platforms-inc-declares-missing-dividend)
- [Alphabet issues first-ever dividend, $70 billion buyback (CNBC)](https://www.cnbc.com/2024/04/25/alphabet-issues-first-ever-dividend-70-billion-buyback.html)
- [Meta Q4 2025 earnings: key metrics (Finviz)](https://finviz.com/news/290992/meta-platforms-meta-q4-earnings-taking-a-look-at-key-metrics-versus-estimates)
- [Company financials, trailing twelve months (stockanalysis.com)](https://stockanalysis.com/stocks/amzn/financials/?period=ttm)
- [Marketing expense, company filings (SEC XBRL data)](https://data.sec.gov/api/xbrl/companyconcept/CIK0001018724/us-gaap/MarketingExpense.json)
- [Yearly closing share prices (StatMuse)](https://www.statmuse.com/money/ask/amazon-yearly-closing-price)
- [Bezos ownership of Amazon (TheWrap)](https://www.thewrap.com/jeff-bezos-amazon-ownership-falls-below-10)

**Precedents**

- [Charles Howarth and the Rochdale dividend on purchases (Wikipedia)](https://en.wikipedia.org/wiki/Charles_Howarth)
- [Rochdale Equitable Pioneers Society (Fondazione Don Guetti)](https://www.fondazionedonguetti.org/wp/?p=2794)
- [Executive rewards (Costco)](https://www.costco.com/executive-rewards.html)
- [Co-op member rewards (REI)](https://www.rei.com/rewards)
- [What happened to Bumped (Grifin)](https://www.grifin.com/post/what-happened-to-bumped-and-is-grifin-the-replacement)
- [Bumped rewards program (The Points Guy)](https://thepointsguy.com/guide/bumped-rewards-program)
- [TD Ameritrade cuts to zero, matching Schwab (CNBC)](https://www.cnbc.com/2019/10/02/the-end-of-commissions-for-stock-trading-is-near-as-td-ameritrade-cuts-to-zero-matching-schwab.html)
- [Fidelity joins the stampede to eliminate fees (CNBC)](https://www.cnbc.com/amp/2019/10/10/fidelity-joins-the-stampede-to-eliminating-fees-for-online-trading.html)

**Households & organizing**

- [Consumer Expenditures 2024 (US Bureau of Labor Statistics)](https://www.bls.gov/news.release/archives/cesan_12192025.htm)
- [One in four Americans spend over $100 a month on streaming and subscriptions (Bango, via Nasdaq)](https://www.nasdaq.com/press-release/one-four-americans-spend-over-100-month-streaming-and-subscriptions-2025-06-05)
- [Commission proposes Canada as EU's first associate member (LexisNexis)](https://www.lexisnexis.co.uk/legal/news/commission-proposes-canada-as-eus-first-associate-member)

**Health**

- [Health Insurance Coverage in the United States: 2025 (US Census Bureau)](https://www.census.gov/library/publications/2026/demo/p60-291.html)
- [2025 Employer Health Benefits Survey (KFF)](https://www.kff.org/health-costs/press-release/2025-employer-health-benefits-survey/)
- [Most Americans say government has a responsibility to ensure health care coverage (Pew Research Center)](https://www.pewresearch.org/short-reads/2025/12/10/most-americans-say-government-has-a-responsibility-to-ensure-health-care-coverage/)
- [Medical loss ratio rebates (KFF)](https://www.kff.org/private-insurance/medical-loss-ratio-rebates/)
- [HSA contribution rules (IRS VITA training guide)](https://vita.business.uconn.edu/wp-content/uploads/sites/1690/2022/01/HSA-Pages-from-Pub-4491-1.pdf)

## About the author

**Debora den Iseger** is SVP of AI Technology & Solutions at Monks, where she leads the development of agentic AI systems and partnerships with major technology platforms. Her career spans creative production, integrated content and AI strategy, at KesselsKramer, DDB Chicago, Uber and Monks.

She is a three-time returning judge for the Runway Festival, has spoken on panels at Cannes Lions, and has been quoted in Adweek and The Wall Street Journal. Her previous essay, *Nobody Asked the Other 99%*, asked who AI is really being built for. This paper is its sequel.

Dutch-born, she is based in the San Francisco Bay Area.

*The views in this paper are her own.*