Martyn Richard Jones

New York City, 18^th January 2017

Scope

I’ll make this short, sweetish and to the point.

These are the arguments that CEOs need to know about Big Data.

It was written for CEOs and those who provide independent advice to CEOs.

If you are someone who wants to be prepared for CEOs who are armed with the realities of Big Data, then maybe this is for you too.

Point One: Big Data is bullshit

What you have heard about Big Data is wrong. Most Big Data talk is cultivated babble. It is designed to get you and your organisation to part with your cash, time, and effort. This talk is full of bluster and bullshit.

Point Two: Heads up. Big Data bullshitters don’t do tangibles

Big Data gurus and bullshitter often use adjectives like amazing, awesome, and incredible. These adjectives fill the surfeit of Big Data bullshit articles they publish. You will be lucky to get one significant example of a tangible, coherent, and verifiable Big Data business success story.

Don’t be suckered by the Big Data bullshit. If Big Data pundits are so sure of their value generation abilities, put everyone at risk. Ensure the projects you commission involve risk for everyone. Confirm they offer rewards for everyone involved. Tie your vendors’ rewards to your own ROI from Big Data. No tangible outcome from Big Data, no tangible benefits for the vendor. Simples. Make sure your Big Data contract is heavily weighted to outcome based payments.

This measure will ensure solution providers receive payment only if a Big Data project offers tangible benefits. These benefits must be evident in your organisation.

So, don’t believe the Big Data bullshit that Big Data pundits are pushing. Be particularly cautious of the opinions expressed by vendor representatives. They could be selling offshored services, management consulting, hardware, software, or vapourware.

Point Three: Beware. The IT industry has vested interests

The IT industry is geared up to maximise advantages. They sell you and the people working for you the illusion that Big Data is the next big thing. It claims to revolutionise how we do business. It’s no coincidence that the hard-sell on Big Data is akin to modern digitalised mass marketing. It resembles campaigns for today’s snake-oil medicine. The IT industry is running short on margin. Big Data is being used to try and make up for increasing shortfalls.

Of course, there is nothing wrong with people trying to make a buck. Just make sure they don’t take you for a ride while they are doing so.

Point Four: FYI Big Data is technology, not business process

Big Data is a set of simple technologies designed to run on low-cost commodity technology products. It is a technology used to ingest and search unstructured data. This data comes from computer systems typically not central to your business operations. Unless of course you run your business via Facebook, twitter or LinkedIn.

It is a technical configuration not a business solution nor a silver-bullet.

Remember this. No matter what the IT people tell you, cheap hardware and open-source software does not equal cheap. They do not lead to beneficial and valuable project outcomes. A rogue Big Data project can inflict a lot of disruption and damage. It can sap your enterprise of time, effort, and money. In the process, it dislocates and frustrates day-to-day business operations and objectives.

Point Five: Vital. Make sure the eyes are on the important data

By far your most valuable data is stored in your businesses operational systems. The bread and butter systems of all business.

To add tangible business value, data is sourced. It is then packaged and provided to the business users. This is part of the Data Warehousing process. It delivers appropriate, adequate and timely data to where it is needed in the organisation. Not by magic, but my deliberate design that can be changed as rapidly as business change demands.

This data is provided in order to support decisions and to help determine actions. At best, Big Data can augment data provided by the Data Warehouse.

Big Data will not replace Data Warehousing. It will not modernise or revolutionise it either. However, claiming otherwise is the biggest and most misleading bullshit line in the field of IT.

Point Six: Good sense. If there is no tangible business benefit, it’s bullshit

Whatever it is, there must be a tangible business benefit. If not, it’s considered bullshit in business terms. It’s also bullshit in technical terms and good sense terms.

Which leads me to a key question. I invariably ask people this when I am brought in to create a new strategy. I may be there to architect a new solution. I might build a new business platform or validate business process reengineering, or something else entirely. Regardless, I ask, “to what ends?”

In my experience, Big Data initiatives generally fail to provide a coherent response. They lack a cohesive or realistic response to the challenge implicit in that question – to what ends?

Also, remember that when it comes to Big Data, we are always inclusive. We remain exclusively exhaustive when questioning the raison d’être. Or, to put it more succinctly, there are no limits to the use of the question “why?” when it comes to Big Data.

Over the years I have had little problem in providing answers to the questions about business benefit. However, sometimes it requires effort to unearth and identify potential benefits. If identifying potential benefits isn’t possible for a Big Data initiative, it’s probably not worth the effort. No matter how attractive the attendant marketing guff appears to be, it’s not something worth giving the time of day.

Point Seven: Business ideas. If there is no business benefit stepping forward, then it’s bullshit

Trust and verify.

I have told you what I think. However, don’t take a word of this on face value. But don’t ignore it either. I suggest getting someone you trust to verify the factual elements of this brief article. It may require more than one subject matter expert whom you trust. Then make up your own mind.

Of course, not everything we do needs to drive cash value. Big Data is definitely a situation to consider carefully. If there are no tangible business benefits, it should be a sign that it’s not worth contemplating. If the Big Data business case cannot satisfactorily answer the question “to what ends?” then the Big Data business case sucks and should be dumped.

I am personally convinced that what I state here is actual, relevant and verifiable. So my advice is to trust, but, to verify for yourself.

And so it came to pass

There you have it. Here’s everything you need to know to align with the business angels. This information is crucial when it comes to Big Data.

Treat all Big Data as bullshit. You won’t go far wrong. This is true even if your business eventually finds a way to make money from Big Data. You’ll thank me for it.

Before I conclude. Two final pieces of counsel.

Do not let anyone in your organisation augment their pension fund. Prevent them from starting any Big Data, MIS, or Business Intelligence project. Avoid initiating any Enterprise Data Warehousing, Artificial Intelligence, Machine Learning, or Data Science programme. You don’t want that sort of conflict of interest to be inflicted upon your business.

Avoid outsourcing any aspect of your business IT that has a substantial link to core business processes. Never offshore anything that is inextricably linked to the value propositions and intellectual capital of the business. This includes Data Warehousing, Business Intelligence, Analytics, and Big Data.

I am available to offer additional advice. This includes Big Data, Data Warehousing, or any aspect of Data Architecture and Management in business. Feel free to give me a call if needed. I look forward to doing business.

Many thanks for reading.

I can be contacted on this and other subjects via my email address: goodstrat2017@gmail.com

You may also be interested in joining The Big Data Contrarians group on LinkedIn

Bill Inmon has written a great piece on Big Data value which can be found here: https://www.linkedin.com/pulse/fishing-where-fish-simone-molenaar

FISHING WHERE THE FISH ARE

Retrospective

A Retrospective on Martyn Jones’s “Big Data is Bullshit” Memo.

This is a look back from February 7, 2026. Martyn Jones
Carmarthen, Wales – February 7, 2026

Nearly a decade has passed since I fired off that blunt, CEO-targeted memo in the mid-2010s. This timeframe was during the 2015–2017 era, based on the style and references. I declared Big Data to be mostly “cultivated babble, bluster and bullshit.” I warned executives: Treat the hype with suspicion. Tie vendor pay to real ROI. Focus on core operational data via data warehousing rather than chasing unstructured social media noise. Always ask “to what ends?” I highlighted the IT industry’s vested interests in promoting ineffective solutions. I emphasized the need for verification over blind trust. I also cautioned against conflicts of interest in internal projects.

From today’s vantage (February 7, 2026) how did that contrarian rant hold up? The data landscape has evolved dramatically, but my core scepticism about hype-driven, vendor-fueled initiatives remains remarkably prescient. Big Data as a standalone buzzword has largely faded. It has been absorbed into broader data/AI ecosystems. Yet, the patterns of overpromising, underdelivering, and misplaced priorities persist—now amplified by the AI boom. Let’s revisit the key points with 2026 hindsight.

Point One & Two: Big Data as Bullshit and Lack of Tangibles. I claimed most Big Data talk was designed to extract cash without delivering verifiable successes. I urged outcome-based contracts. Reality check: The early 2010s–mid-2020s hype cycle did produce plenty of flops. Think of Hadoop-heavy projects that became expensive data graveyards with low ROI. But tangible wins emerged and scaled massively. Netflix’s recommendation engine (powered by Big Data + AI) still saves ~$1 billion annually and drives 80% of views. Amazon’s predictive inventory and Walmart’s demand forecasting cut waste and boosted efficiency dramatically. In healthcare, predictive analytics saved billions (e.g., Kaiser Permanente’s $1B+ gains). By 2025–2026, organizations reporting measurable value from data/analytics investments hit 91.9%, with average revenue boosts of 8% and cost reductions of 10%. High performers see 10–15% revenue lifts and even 127% ROI on BI implementations over three years. Yet, the bullshit filter still applies. Gartner notes 65% of AI projects, which are Big Data’s successor, risk abandonment due to poor data readiness. Many “Big Data” initiatives failed precisely because they ignored “to what ends?”, focusing on volume over value. My advice to tie rewards to ROI? Spot-on; outcome-based models are now standard in mature deals.Point Three: IT Industry Vested Interests

The industry pushed Big Data to offset margin squeezes—classic snake-oil marketing.In 2026: This critique aged like fine wine. The shift from Big Data to AI/genAI saw similar hype: massive investments ($1.5T+ AI spend projected for 2025, rising to $2T+ in 2026), yet many face a “hype correction.” 2025 brought an AI reckoning, with bubbles deflating and agentic AI overhyped. Vendors still sell compute-heavy stacks, but successful orgs now demand proof—integrated transformations yield 10x ROI vs. fragmented ones. The caution against vendor opinions holds: trust but verify remains essential.

Point Four: Big Data as Technology, Not Business Process. I stressed it’s just cheap commodity tech for ingesting and searching unstructured data. It’s not a silver bullet. Rogue projects can disrupt operations.Evolution: This nailed it. Pure “Big Data” (Hadoop-era lakes) transformed into data lakehouses. These unify lakes’ scale with warehouses’ governance via formats like Delta Lake, Apache Iceberg, and Hudi. By 2025–2026, lakehouses matured into the dominant architecture: open, transactional, AI-ready, with real-time streaming and ACID support. It’s not revolutionary magic; it’s deliberate engineering. Failures still happen when tech leads without business alignment—disrupting ops as I warned.Point Five: Eyes on Important (Operational) DataCore systems house the most valuable data; Big Data augments warehouses, not replaces them.2026 verdict: Vindicated emphatically. Modern architectures prioritize governed, high-quality operational data as the foundation. Brick Lakehouses build on this, augmenting warehouses with unstructured sources for AI/ML. Data warehousing evolved, not died; it’s integral to “data-ready” orgs pulling ahead in AI. Pretending Big Data “revolutionizes” warehousing was indeed misleading hype.

Points Six & Seven: No Tangible Benefit = Bullshit; Trust but VerifyIf no clear business ends, dump it. Verify claims independently.This is timeless. In 2026, data maturity varies wildly—only ~38% of Fortune 1000 are truly data-driven despite massive spends. Successful firms quantify gains (e.g., 11–20% profitability boosts). The “to what ends?” question is now central to AI governance and value chains. My call to avoid internal pension-padding via pet projects? Still wise amid AI conflicts.Final Counsel: Avoid Outsourcing Core Data CapabilitiesDon’t offshore intellectual capital like data warehousing/BI/analytics.Spot-on in hindsight: Data sovereignty, quality, and AI readiness demand in-house control or trusted partners. Outsourcing core processes led to many regrets; hybrid models prevail where value props are at stake.

In sum, my 2010s memo was a necessary antidote to hype. Big Data wasn’t total bullshit. Mature applications deliver enormous value. However, the bullshit parts (vendor overpromising, tech-first thinking, ignoring business ends) were real and costly. Today, we’ve moved to lakehouses and outhouses, AI-native platforms, and real-time analytics. The discipline I advocated defines winners. Scepticism, outcome focus, and verification are essential. Treat unproven initiatives as suspect, and you’ll still avoid the snake pits.Europe—and global business—thrives on evidence-based unity, not hype. Thanks for the read; some things never change.

Many thanks for reading.