---
title: "Why AI is just big data again"
url: https://modern-mba.com/case/ai-hoax
sector: "Technology"
sector_url: https://modern-mba.com/sector/technology
aired: 2024-05-26
published: 2026-08-26
updated: 2026-08-28
reading_time_minutes: 15
charts: 16
video: https://www.youtube.com/watch?v=pOuBCk8XMC8
author: "Yishan Lin"
author_url: https://www.linkedin.com/in/yishanlin/
publisher: Modern MBA
license: "© Modern MBA, all rights reserved. Quote and cite with credit and a link; republishing needs written permission."
---

# Why AI is just big data again

For a decade Silicon Valley told everyone that data was gold and that any company not mining it would be left behind. Thousands of startups staked their existence on it, the Fortune 500 spent billions catching up, and a generation of engineers built careers on it. Then the whole thing quietly stopped being mentioned. This case study goes through the accounts of the companies that made that promise — Groupon, Wayfair, Blue Apron, Opendoor, Stitch Fix, DoorDash, and two dozen more — and finds no business value anywhere in them. The only reliable winners were the vendors selling tools and the cloud providers renting the machines. AI is the same market structure with a new noun.

## Key figures

- **12 of 14** — Big-data consumer startups losing money at the 2016 peak
- **7,207%** — Shopify since IPO — the winners are all B2B vendors
- **$1B → $99B** — AWS revenue over the decade big data delivered nothing

## The argument

The tech sector runs on a story, and the story has to be replaced every few years because the last one never worked. Before AI there was crypto, web3, blockchain, VR, IoT, and — biggest of all — big data. Each was revolutionary while it was funded and dropped without ceremony the moment returns failed to arrive. Nobody pays a price for this, which is why it keeps happening.

Big data was the version that got everywhere: users generate data, data powers algorithms, algorithms produce innovation. Thousands of startups staked their existence on it, and it justified a decade of unprofitable growth as deliberate strategy. Groupon said it had 9 petabytes. Wayfair captured 4 terabytes a day and hired 1,900 data scientists. Every one is now worth a fraction of what it was, and not one produced a finding anybody can name.

So who got paid. Not the startups — twelve of the fourteen biggest lost money at the peak. Not the Fortune 500, whose 2023 margins are indistinguishable from 2013. The winners were the people selling shovels: enterprise vendors like ServiceNow and Snowflake, and above all the cloud providers, because everything ran on AWS. That is the exact market structure now reassembling around AI. The only thing that has changed is the noun.

## Charts

### All talk and no profit

Operating income of the largest venture-backed tech startups, 2022, in millions. Eleven of twelve lose money.

|  | US dollars |
| --- | --- |
| Uber | −$1,823M |
| Airbnb | $1,802M |
| Coinbase | −$2,710M |
| Snowflake | −$715M |
| DoorDash | −$1,124M |
| Rivian | −$6,856M |
| Lyft | −$1,459M |
| Opendoor | −$931M |
| Pinterest | −$102M |
| Robinhood | −$1,011M |
| Affirm | −$866M |
| Snap | −$1,395M |

Source: Modern MBA, “Why AI is just big data again”, aired May 2024. Chart: https://modern-mba.com/case/ai-hoax#chart-1 · Sources: https://modern-mba.com/case/ai-hoax#sources

### The hockey stick that started it

Groupon revenue, in millions. Nobody had to be argued into using the first movers — the adoption was organic.

|  | US dollars |
| --- | --- |
| 2008 | $0M |
| 2009 | $14M |
| 2010 | $313M |
| 2011 | $1,610M |
| 2012 | $2,334M |
| 2013 | $2,573M |
| 2014 | $3,192M |

Source: Modern MBA, “Why AI is just big data again”, aired May 2024. Chart: https://modern-mba.com/case/ai-hoax#chart-2 · Sources: https://modern-mba.com/case/ai-hoax#sources

### The margins that made the argument

Annual operating margins of the leading tech companies. If data is what produced these, the reasoning went, a startup should manage a tenth of it.

|  | 2009 | 2016 |
| --- | --- | --- |
| Facebook | 34% | 45% |
| Google | 35% | 26% |
| Apple | 27% | 28% |
| Microsoft | 35% | 24% |

Source: Modern MBA, “Why AI is just big data again”, aired May 2024. Chart: https://modern-mba.com/case/ai-hoax#chart-3 · Sources: https://modern-mba.com/case/ai-hoax#sources

### Weak returns on data: content and social

Operating income of the largest data-driven consumer startups, 2016, in millions. Every one of them underwater at the peak of the story.

|  | US dollars |
| --- | --- |
| Zynga | −$114M |
| Pandora | −$319M |
| Pinterest | −$188M |
| Dropbox | −$194M |
| Roku | −$43M |
| Twitter | −$367M |
| Snap | −$520M |

Source: Modern MBA, “Why AI is just big data again”, aired May 2024. Chart: https://modern-mba.com/case/ai-hoax#chart-4 · Sources: https://modern-mba.com/case/ai-hoax#sources

### Weak returns on data: the marketplaces

Operating income of the largest data-driven consumer marketplaces, 2016, in millions. One of seven made money.

|  | US dollars |
| --- | --- |
| LendingClub | −$150M |
| Wish | −$147M |
| Groupon | −$100M |
| Yelp | −$2M |
| TrueCar | −$39M |
| GrubHub | $84M |
| Wayfair | −$196M |

Source: Modern MBA, “Why AI is just big data again”, aired May 2024. Chart: https://modern-mba.com/case/ai-hoax#chart-5 · Sources: https://modern-mba.com/case/ai-hoax#sources

### Nine petabytes and a downward slope

Groupon net revenue and operating income, in millions. The company that A/B tested every single feature.

|  | Net revenue | Operating income |
| --- | --- | --- |
| 2013 | $2,573M | $76M |
| 2014 | $3,192M | $31M |
| 2015 | $2,955M | −$71M |
| 2016 | $3,014M | −$100M |
| 2017 | $2,844M | $29M |
| 2018 | $2,637M | $54M |
| 2019 | $2,219M | $40M |
| 2020 | $1,417M | −$277M |
| 2021 | $967M | −$5M |
| 2022 | $599M | −$168M |
| 2023 | $515M | −$18M |

Source: Modern MBA, “Why AI is just big data again”, aired May 2024. Chart: https://modern-mba.com/case/ai-hoax#chart-6 · Sources: https://modern-mba.com/case/ai-hoax#sources

### 1,900 data scientists, one profitable year

Wayfair net revenue and operating income, in millions. Four terabytes a day and 40 billion customer actions a year.

|  | Net revenue | Operating income |
| --- | --- | --- |
| 2013 | $916M | −$16M |
| 2014 | $1,319M | −$148M |
| 2015 | $2,250M | −$81M |
| 2016 | $3,380M | −$197M |
| 2017 | $4,721M | −$235M |
| 2018 | $6,779M | −$473M |
| 2019 | $9,127M | −$930M |
| 2020 | $14,145M | $360M |
| 2021 | $13,708M | −$94M |
| 2022 | $12,218M | −$1,384M |
| 2023 | $12,003M | −$813M |

Source: Modern MBA, “Why AI is just big data again”, aired May 2024. Chart: https://modern-mba.com/case/ai-hoax#chart-7 · Sources: https://modern-mba.com/case/ai-hoax#sources

### Knowing every customer's taste profile

Blue Apron net revenue and operating income, in millions. Nine years, nine operating losses.

|  | Net revenue | Operating income |
| --- | --- | --- |
| 2014 | $78M | −$31M |
| 2015 | $341M | −$47M |
| 2016 | $795M | −$55M |
| 2017 | $881M | −$189M |
| 2018 | $668M | −$114M |
| 2019 | $455M | −$52M |
| 2020 | $461M | −$39M |
| 2021 | $470M | −$71M |
| 2022 | $458M | −$109M |

Source: Modern MBA, “Why AI is just big data again”, aired May 2024. Chart: https://modern-mba.com/case/ai-hoax#chart-8 · Sources: https://modern-mba.com/case/ai-hoax#sources

### Algorithms that price houses better than humans

Opendoor net revenue and operating income, in millions. Buying property at scale is a balance sheet, not an insight.

|  | Net revenue | Operating income |
| --- | --- | --- |
| 2015 | $53M | −$11M |
| 2016 | $339M | −$33M |
| 2017 | $711M | −$61M |
| 2018 | $1,838M | −$164M |
| 2019 | $4,741M | −$248M |
| 2020 | $2,583M | −$186M |
| 2021 | $8,021M | −$568M |
| 2022 | $15,567M | −$931M |
| 2023 | $6,946M | −$386M |

Source: Modern MBA, “Why AI is just big data again”, aired May 2024. Chart: https://modern-mba.com/case/ai-hoax#chart-9 · Sources: https://modern-mba.com/case/ai-hoax#sources

### Growth was never the part anybody doubted

DoorDash net revenue and operating income, in millions. Thirty-fold revenue growth and a loss in every year of it.

|  | Net revenue | Operating income |
| --- | --- | --- |
| 2018 | $291M | −$210M |
| 2019 | $885M | −$616M |
| 2020 | $2,886M | −$436M |
| 2021 | $4,888M | −$452M |
| 2022 | $6,583M | −$1,124M |
| 2023 | $8,635M | −$579M |

Source: Modern MBA, “Why AI is just big data again”, aired May 2024. Chart: https://modern-mba.com/case/ai-hoax#chart-10 · Sources: https://modern-mba.com/case/ai-hoax#sources

### During a gold rush, sell shovels

Tech stocks with the greatest appreciation in valuation since IPO, 2012 to 2024. The winners are almost entirely enterprise vendors.

|  | Percentages |
| --- | --- |
| ServiceNow | 6,581% |
| Trade Desk | 3,809% |
| MongoDB | 1,268% |
| Palo Alto | 3,318% |
| Arista | 3,483% |
| Block | 1,348% |
| Shopify | 7,207% |
| Workday | 1,388% |

Source: Modern MBA, “Why AI is just big data again”, aired May 2024. Chart: https://modern-mba.com/case/ai-hoax#chart-11 · Sources: https://modern-mba.com/case/ai-hoax#sources

### Selling table stakes

Revenue of the two leading telemetry vendors, in millions. Neither needed anyone downstream to derive a single insight.

|  | Splunk | Datadog |
| --- | --- | --- |
| 2014 | $303M | $0M |
| 2015 | $451M | $0M |
| 2016 | $668M | $0M |
| 2017 | $944M | $101M |
| 2018 | $1,309M | $198M |
| 2019 | $1,803M | $363M |
| 2020 | $2,359M | $603M |
| 2021 | $2,229M | $1,029M |
| 2022 | $2,674M | $1,675M |
| 2023 | $3,654M | $2,128M |

Source: Modern MBA, “Why AI is just big data again”, aired May 2024. Chart: https://modern-mba.com/case/ai-hoax#chart-12 · Sources: https://modern-mba.com/case/ai-hoax#sources

### All those petabytes had to go somewhere

Snowflake revenue, in millions. A warehouse for data nobody proved was worth warehousing.

|  | US dollars |
| --- | --- |
| 2019 | $97M |
| 2020 | $265M |
| 2021 | $592M |
| 2022 | $1,219M |
| 2023 | $2,066M |
| 2024 | $2,806M |

Source: Modern MBA, “Why AI is just big data again”, aired May 2024. Chart: https://modern-mba.com/case/ai-hoax#chart-13 · Sources: https://modern-mba.com/case/ai-hoax#sources

### Being B2B is not the same as being profitable

Operating income of the leading B2B enterprise tech companies, 2024, in millions. Resilient valuations, unresolved economics.

|  | US dollars |
| --- | --- |
| Atlassian | −$345M |
| Asana | −$270M |
| Datadog | −$33M |
| Snowflake | −$1,095M |
| MongoDB | −$234M |
| Twilio | −$877M |
| Shopify | −$1,418M |
| Palo Alto | $387M |

Source: Modern MBA, “Why AI is just big data again”, aired May 2024. Chart: https://modern-mba.com/case/ai-hoax#chart-14 · Sources: https://modern-mba.com/case/ai-hoax#sources

### The people who actually got paid

Annual revenue of the public clouds, in billions. It does not matter whether the trend is big data, IoT, or AI, as long as it drives usage.

|  | AWS | Google Cloud |
| --- | --- | --- |
| 2011 | $1B | $0B |
| 2012 | $2B | $0B |
| 2013 | $4B | $0B |
| 2014 | $5B | $0B |
| 2015 | $8B | $0B |
| 2016 | $12B | $0B |
| 2017 | $17B | $4B |
| 2018 | $25B | $6B |
| 2019 | $35B | $9B |
| 2020 | $45B | $13B |
| 2021 | $62B | $19B |
| 2022 | $80B | $26B |
| 2023 | $99B | $33B |

Source: Modern MBA, “Why AI is just big data again”, aired May 2024. Chart: https://modern-mba.com/case/ai-hoax#chart-15 · Sources: https://modern-mba.com/case/ai-hoax#sources

### No evidence of impact

Operating margin of the leading data-driven Fortune 500 companies, 2013 against 2023. A decade of investment, no visible return.

|  | 2013 | 2023 |
| --- | --- | --- |
| Walmart | 5% | 3% |
| Costco | 3% | 3% |
| Kroger | 3% | 2% |
| Target | 6% | 3% |
| PepsiCo | 13% | 12% |
| Kellogg's | 18% | 8% |
| General Mills | 16% | 17% |
| Unilever | 14% | 17% |
| Chase | 27% | 36% |
| Citi | 26% | 25% |
| Wells Fargo | 40% | 20% |
| Goldman Sachs | 37% | 28% |

Source: Modern MBA, “Why AI is just big data again”, aired May 2024. Chart: https://modern-mba.com/case/ai-hoax#chart-16 · Sources: https://modern-mba.com/case/ai-hoax#sources


## Takeaways

1. Then Marc Andreessen wrote that software was eating the world and the pitch generalized. Data was gold, data could be applied to any industry, users generated data, data powered algorithms, algorithms produced innovation. **Chegg's** leadership summarized it exactly: take a giant category, use technology to deliver at scale, own the customer and the channel, keep the data, and build a moat nobody can cross.
2. The moat argument rested on companies that were already winning. **Facebook**, **Google**, **Apple**, and **Microsoft** posted operating margins of **34%**, **35%**, **27%**, and **35%** in 2009, and **45%**, **26%**, **28%**, and **24%** by 2016 — figures the rest of the private sector had never seen. If data was what made FAANG work, a nimbler startup applying it to a smaller industry should manage a tenth of that. That was the whole syllogism.
3. By 2016 the results were in and they were uniform. Of the fourteen largest data-driven consumer startups, twelve lost money: **Snap** **−$520M**, **Twitter** **−$367M**, **Pandora** **−$319M**, **Wayfair** **−$196M**, **Dropbox** **−$194M**, **Pinterest** **−$188M**, **LendingClub** **−$150M**, **Wish** **−$147M**, **Zynga** **−$114M**, **Groupon** **−$100M**, **Roku** **−$43M**, **TrueCar** **−$39M**. Only **GrubHub** at **+$84M** made money.
4. Silicon Valley's response was to move the goalposts rather than concede. Data was still valuable — you simply did not have **enough** of it, or the means to interpret it. Basic analytics and personalization were no longer sufficient. What you needed was terabytes, sophisticated tooling, and data scientists. That relabeling is what got called Big Data.
5. The quotes from the peak have not aged well. **Groupon**: *we have more than 9 petabytes of data and we A/B test every single feature — how many companies do this?* It bottomed out within five years. **Wayfair**: *4 terabytes a day, 40 billion customer actions a year, 1,900 engineers and data scientists, multiple platforms at a strong ROI.* It still loses close to a billion dollars a year.
6. **Groupon's** own numbers are the clearest arc in the industry: revenue **$3,192M** in 2014 falling to **$2,573M**, **$2,955M**, **$3,014M**, **$2,844M**, **$2,637M**, **$2,219M**, **$1,417M**, **$967M**, **$599M**, and **$515M** by 2023 — with an operating loss in seven of those eleven years.
7. **Blue Apron** claimed it knew every individual's taste profile and could algorithmically predict demand for any meal kit. Revenue peaked at **$881M** in 2017 and fell every year after to **$458M**, with an operating loss in all nine years — **−$189M** at the peak of the data story.
8. The Fortune 500 joined out of fear, not merit. **Walmart**, **ExxonMobil**, **Home Depot**, **AT&T**, **Comcast**, **Disney**, **PepsiCo**, **Chase**, and **Citigroup** each committed hundreds of millions — some billions — because no chief executive wanted to be the one caught saying their teams were not data-driven. **Under Armour** is the small-cap version, spending over half a billion on fitness apps chasing FAANG margins.
9. A decade later there is nothing in their accounts to show for it. Operating margins in 2013 against 2023: **Walmart 5% → 3%**, **Costco 3% → 3%**, **Kroger 3% → 2%**, **Target 6% → 3%**, **PepsiCo 13% → 12%**, **Kellogg's 18% → 8%**, **General Mills 16% → 17%**, **Unilever 14% → 17%**, **Chase 27% → 36%**, **Citi 26% → 25%**, **Wells Fargo 40% → 20%**, **Goldman Sachs 37% → 28%**. If big data had produced efficiencies, this is where they would appear.
10. During a gold rush, sell shovels. Of the tech stocks with the greatest appreciation since IPO between 2012 and 2024, the winners are almost entirely enterprise: **Shopify** at **7,207%**, **ServiceNow** at **6,581%**, **The Trade Desk** at **3,809%**, **Arista** at **3,483%**, **Palo Alto Networks** at **3,318%**, **Workday** at **1,388%**, **Block** at **1,348%**, and **MongoDB** at **1,268%**.
11. **Splunk** grew from **$303M** to **$3,654M** and **Datadog** from **$101M** to **$2,128M** selling table-stakes software for capturing and monitoring data. **Snowflake** went **$97M**, **$265M**, **$592M**, **$1,219M**, **$2,066M**, **$2,806M**. None of that revenue required anyone downstream to derive a single insight.
12. Being B2B is not the same as being profitable, though. In 2024 the leading enterprise vendors posted operating income of **−$1,418M** at **Shopify**, **−$1,095M** at **Snowflake**, **−$877M** at **Twilio**, **−$345M** at **Atlassian**, **−$270M** at **Asana**, **−$234M** at **MongoDB**, and **−$33M** at **Datadog**, with only **Palo Alto Networks** positive at **$387M**. The advantage is resilience: enterprise deals are large, churn is slow, and a tool embedded at a big company is hard to rip out.
13. But the biggest winners were the clouds, and it was not close. **AWS** went from **$1 billion** in 2011 to **$99 billion** in 2023 and **Google Cloud** from **$4 billion** to **$33 billion** — because the consumer startups burned venture money on cloud, the Fortune 500 used big data as the forcing function to migrate, and the B2B vendors built their tools there too. It does not matter whether the trend is big data, IoT, AR, or AI, as long as it drives usage.
14. Engineers had their own reasons to keep it alive. The technology was new, so there were no best practices; everyone worked it out in the open on GitHub, and naming the right tools on a résumé converted into a higher-paying job within months. Big data industrialized résumé-driven development, and developers stopped being implementers and became the people choosing the clouds, tools, and vendors on their employer's behalf.
15. And the AI premise is stranger than the one it replaced. Big data said the insights were in there and you needed better tools. AI says the insights are in there and **humans cannot get them out** — so trust a system almost nobody can inspect, accept the answer instead of seeking it, and value the packaging and the speed above the accuracy. That is a retreat dressed as a breakthrough.
16. The accounting is the same as it ever was. All talk, no show, all sizzle, no steak. The winners from big data were the founders, executives, and venture capitalists who liquidated at IPO and the people at the B2B vendors. The losers were public investors and the employees. Nothing about the AI cycle has changed who sits in which seat.

## Common questions

### What happened to big data?

Nothing, which is the point. It was sold for roughly a decade as the technology that would unearth hidden insights, predict demand before it existed, and give any company a moat. Thousands of consumer startups staked their existence on it, the Fortune 500 committed billions, and by the early 2020s it had quietly stopped being mentioned. No consumer startup demonstrated a business result from it, no Fortune 500 shows a margin improvement in its filings, and nobody can point to a replicable method for extracting value from it. It was not discredited or debated — it was simply replaced by AI as the thing everyone was supposed to be doing.

### Is AI a bubble?

The market structure is identical to the one that produced big data. Thousands of consumer startups arriving with AI products and only the promise of business value; enterprise vendors selling AI tooling to those startups; the Fortune 500 running scared and buying in for fear of being caught behind; the cloud providers and chipmakers collecting rent from all of them; engineers adopting it to improve their own career prospects; and Wall Street treating tech as the story again. Whether the underlying models are useful is a separate question from whether the valuations built on them can be earned back, and the last three cycles say they cannot.

### Did any company make money from big data?

The vendors and the clouds. Of the tech stocks with the greatest appreciation since IPO between 2012 and 2024, almost all are enterprise: Shopify at 7,207%, ServiceNow at 6,581%, The Trade Desk at 3,809%, Arista at 3,483%, and Palo Alto Networks at 3,318%. Splunk grew to $3.65 billion of revenue and Datadog to $2.13 billion selling table-stakes tooling. And above them AWS went from $1 billion in 2011 to $99 billion in 2023, with Google Cloud from $4 billion to $33 billion. Everyone selling the means to do big data got paid. Almost nobody doing big data did.

### Why were so many data-driven startups unprofitable?

Because they were structurally unprofitable and had a story that made it sound deliberate. The genuine first movers — Yelp, Pandora, Groupon, GrubHub — had organic adoption and network effects. The thousands who followed had neither, so they bought users with advertising and subsidies. Broken unit economics were reframed as investment: these companies were unprofitable by design, acquiring the data that would eventually supercharge the business. The data was acquired. The supercharging was not. In 2016, twelve of the fourteen largest of them were losing money, led by Snap at −$520 million and Twitter at −$367 million.

### Did big data actually improve the Fortune 500?

Not in any way that reaches the accounts. Compare operating margins in 2013 against 2023: Walmart 5% to 3%, Kroger 3% to 2%, Target 6% to 3%, PepsiCo 13% to 12%, Kellogg's 18% to 8%, Wells Fargo 40% to 20%, Goldman Sachs 37% to 28%. A few rose — Chase from 27% to 36%, Unilever 14% to 17% — for reasons that have nothing to do with analytics. These companies are still doing layoffs, cost cuts, shrinkflation, and buybacks to hit their numbers, which is exactly what they would not have to do if big data had delivered the efficiencies it promised.

### Why do software engineers keep adopting hyped technologies?

Because it pays. When a technology is new there are no best practices, so proficiency is scarce and claiming it on a résumé converts into a higher-paying job within months. That gave engineers a direct financial interest in keeping big data trendy regardless of business results, and it is why enterprise vendors and cloud providers spend millions a year on conferences and developer relations — a converted engineer is the cheapest enterprise sales channel available. It also made engineering tribal: React against Angular, Kubernetes against ECS, and now TensorFlow against PyTorch and one model against another.

### How is the AI pitch different from the big data pitch?

It concedes more while promising the same thing. Big data said the insights were sitting in your data and you needed better tools and better people to reach them. AI says the insights are still there but humans cannot extract them at all — so you should trust a system very few people understand, accept the answer rather than seeking it, and prioritize digestibility and speed over accuracy. That is a weaker claim wearing a bigger one's clothes, and it is being made by many of the same people who were selling crypto and NFTs three years ago.

## Discussion

- Crypto, web3, blockchain, VR, IoT, and big data were each revolutionary while funded and dropped without ceremony when returns failed to arrive. Nobody paid a price. What would a functioning accountability mechanism even look like here?
- Big data justified a decade of unprofitable growth as deliberate strategy. Which current narrative is doing that same work today, and for whom?
- Groupon reported 9 petabytes and Wayfair captured 4 terabytes a day and hired 1,900 data scientists. What question should a board have asked before approving that spend?
- If the story has to be replaced every few years because the last one never worked, who is the story actually for?
- You sit on a board being pitched an AI transformation. What three questions do you ask, and what answers would satisfy you?

## Sources

Revenue and operating income from company annual reports, 10-K filings, and S-1s for the years shown; venture-backed startup figures are the most recent full year disclosed before or at the date given; Fortune 500 operating margins calculated from reported operating income over revenue; cloud revenue as segment-reported by Amazon and Alphabet

---

From [Modern MBA](https://modern-mba.com/). Read this case in full at [https://modern-mba.com/case/ai-hoax](https://modern-mba.com/case/ai-hoax), or watch the episode at https://www.youtube.com/watch?v=pOuBCk8XMC8.