CʳAᵖGR: The Cynical Joke Market Research Firms Play at the Expense of Gullible Executives

Every market research report ends the same way. The market, valued at $X billion today, is poised to reach $Y billion by 2030 at a CAGR of Z%. Nobody ever goes back to check. In 2018, BCC Research published a report projecting the global VR and AR market would reach $142.4 billion by 2023, from a base of $13.4 billion, at a CAGR of 60.4%. That is not a misprint###-###-#### percent, compounded, for five consecutive years. The actual VR market in 2023? Somewhere between $13 billion and $37 billion, depending on which current report you consult. The most optimistic estimate implies a realized CAGR of roughly 18%. The most conservative puts 2023 revenues approximately where they were when BCC made the forecast - a CAGR near zero. BCC Research is a legitimate Boston-based firm with a decades-long track record. Not top-tier, but also not complete nonsense. They were not making things up. They were doing exactly what the industry does: Selling... Sorry, I meant - extrapolating from early hype signals in a market that had not yet encountered hardware cost realities, distribution economics, or the particular challenge of making people strap displays to their faces for extended periods without becoming nauseous. CHART 1 · PREDICTED VS. ACTUAL VR/AR in 2023: predicted at $142B, arrived somewhere between $13B and $37B BCC Research 2018 forecast vs. current market estimates. Gap: 40+ percentage points of annualized CAGR. Interesting postscript: every firm that missed the 2023 VR/AR number has now published a fresh 5-to-10-year forecast for the same market, with CAGRs ranging from 19% to 39%. The prior miss goes unacknowledged. There is no mechanism that requires accountability. The reports are sold for $4k a pop, gullible execs still buy them, the horizon resets, the ecosystem rolls forward. The market will not stop producing these numbers. The least anyone can do is stop pretending to believe them. What CAGR Is (vs. What People Think It Is) CAGR as a retrospective metric is fine. If your revenue was $5M in 2022 and $8M in 2026, your CAGR is 12.5%. Nobody disputes the arithmetic. The problem is the predictive use: taking a 3-to-5-year historical growth window, applying it forward for 5 to 10 years, and presenting the result as a market forecast dressed in two decimal points of false precision. What you have is a linear/exponential extrapolation wearing a compound-growth coat. It assumes that whatever was true yesterday will continue, smoothly and uninterrupted, into a future that does not contain recessions, regulatory shifts, competitive disruption, technological substitution, AI disruption, Quantum disruption or the occasional global pandemic. That is not a model of the future. That is the future with all the interesting parts removed. The Track Record Nobody Talks About The CAGR industrial complex does not like to discuss retrospective accuracy, for an obvious reason: the retrospective accuracy is embarrassing. The CXO Advisory Group spent seven years collecting and grading 6,584 public market forecasts from 68 widely cited financial experts. Average forecast accuracy across the full sample: 47%. Statistically speaking, slightly worse than a coin flip. The bell curve of accuracy scores looked exactly like what you would expect from random outcomes. Among the notable participants: Abby Joseph Cohen of Goldman Sachs, 35%. Gary Shilling of Forbes, 38%. CHART 2 · FORECASTER ACCURACY Of every 100 expert market forecasts studied, 53 were simply wrong Each cell = 1 forecast. Based on CXO Advisory Group study of 6,584 forecasts by 68 professional experts, ###-###-#### . Source: CXO Advisory Group - Guru Grades It gets worse. A Berkeley Haas study analyzing 16,559 forecasts from the Federal Reserve's Survey of Professional Forecasters found that senior economists at major banks were 53% confident in their predictions but correct only 23% of the time. Not slightly miscalibrated. Wrong in all directions, with high confidence. And if you believe the Federal Reserve's own economists are doing better: Deutsche Bank's Torsten Slok found that Wall Street forecasts for 10-year Treasury rates had been "consistently wrong" for over a decade and a half, averaging 60 basis points of error per year in the same direction. The Federal Reserve's own research found that its one-year economic forecasts performed no better than simple benchmark predictions. JPMorgan's Jamie Dimon called Wall Street's recent forecasts "100 percent dead wrong" at a public summit, then immediately said to be cautious about next year's forecasts, without registering any apparent irony. Warren Buffett, in his 2013 shareholder letter: "Forming macro opinions or listening to the macro or market predictions of others is a waste of time. Indeed, it is dangerous, because it may blur your vision of the facts that are truly important." This is, notably, a man who has spent seven decades analyzing businesses and markets saying that macro forecasting is a waste of time. We continue to produce it at industrial scale. You want another case study? Here you go... In December 2018, MarketsandMarkets projected the global blockchain market would grow from $1.2 billion to $23.3 billion by 2023, at a CAGR of 80.2%. What actually happened to blockchain by 2023 is, depending on which firm you ask: $7.5 billion (SkyQuest), $12.4 billion (GlobalData), or $17.46 billion (CoinLedger). Note that none of those estimates reach $23.3 billion. Also note that they span a range of more than 2x. In other words, a market that unambiguously existed and had been operating for years could not be measured retroactively within a factor of two. CHART 3 · THE MEASUREMENT PROBLEM Blockchain 2023: one forecast, three 'actuals,' none agreeing MarketsandMarkets 2018 forecast vs. recent estimates for 2023 from three independent firms. Same market, same year. And trust me, there are many many more examples, thousands probably, here are... ...Three more from Perplexity ...And 10 more from Gemini And I think you got the idea… Not All CAGR Is Equal There is meaningful variance in how market research is produced, and pretending otherwise would be intellectually dishonest. The ecosystem runs from institutional-quality work to something that occupies a category best described as decorative, and more likely - just cynical. At the credible end, firms like IBISWorld, Wood Mackenzie, and Definitive Healthcare use proprietary primary data, named macro drivers, and regression analysis - and they acknowledge uncertainty explicitly. They are not selling certainty; they are selling analysis, which is a different product. At the other end, a cohort of high-volume publishers - Mordor Intelligence, Grand View Research, MarketsandMarkets, Technavio - produce forecasts at industrial scale. These reports are not peer-reviewed, their methodologies are typically not disclosed, and their accuracy track records are not published. What they are is cheap, fast, and formatted to drop directly into slide 3 of your pitch deck. They still charge you $5,000 a pop though, because they can, and because you're gullible. Reviews for these publications are scarce but negative, and for some obscure reason never actually come from end consumers in America, Europe or the Far East. Generally, they have data that does not match independently verifiable figures, CAGRs that differ between executive summary and body text, and a refund policy that is technically present but functionally inaccessible. Or as one Reddit user eloquently summarized: Article content CHART 4 · PUBLISHER LANDSCAPE Where major CAGR publishers sit on rigor vs. optimism An entirely subjective but defensible placement of major market research publishers. Did you notice there is no bottom left here? (low Analytical Rigor + low CAGR optimism) – the answer is simple – pessimism doesn’t sell. C’est Tout And here is a different, tabled, visualization of those tiers: I want to be fair to the credible end of this spectrum, because even firms with rigorous methodology face the same fundamental problem: complex adaptive systems do not compound smoothly. They lurch. IBISWorld's 5-year CAGRs will miss more often than their 2-year CAGRs regardless of methodology quality. BCC Research uses real analysts and real data. They still missed VR/AR by 40 percentage points, annualized, over five years. The problem is not just the researchers. It is the horizon. I also want to be fair to all your buyers and remind you that even the most rigorous of these publications, ultimately are trying to sell a service - and that having the CAGR number in there, makes it a little easier to. For this reason, you can’t help but admire outlets like MarketReader that don’t even try to predict, but rather rigorously explain real-time movement. A Note on Government Sources ------------------ Government publications are not exempt from this problem, and they occupy a category worth discussing separately. The EIA's Annual Energy Outlook, which shapes federal energy policy, projects energy production and consumption through###-###-#### The EIA itself states in the AEO 2025 that "these results are not predictions of what will happen." That honest disclaimer rarely survives the citation chain: by the time the AEO appears in an investor deck or a congressional hearing, the scenario label has been dropped and the projection has become a forecast. The Bureau of Labor Statistics publishes 10-year employment projections by occupation and sector, used extensively by workforce planning teams and universities. The BLS methodology is transparent and the data is real, but predicting which occupations grow over a decade requires predicting technology adoption rates, regulatory changes, and macroeconomic conditions that are themselves unpredictable. The CBO, IMF, and World Bank all publish multi-year economic projections that carry institutional credibility without the methodological limitations disappearing. Government projections are, on balance, more trustworthy than market research reports. They are not forecasts in the predictive sense. The best ones say so. Use them accordingly. -------------------- The CRVI: A Metric Worth Naming And to show you the absurdity of prediction, I created a metric: CRVI, CAGR’s Real Value Index. The numerator is perceived legitimacy, which is high. A CAGR from a named source with two decimal places looks authoritative. The denominator is actual predictive value. Across firms, tiers, and forecast horizons, it always rounds to zero, whether your a 17 year old research intern from Pune or a McKinsey veteran. The ratio remains: CRVI = Perceived Legitimacy / Actual Predictive Value = ∞ The metric holds regardless of publication tier. The denominator rounds to zero at every price point. Why It Persists Anyway CAGR persists for the same reason that nautical miles survived long after GPS: it became the unit. When your investor asks "how big is this market?", they expect a dollar figure and a growth rate. If you respond that you do not have a CAGR because it would represent fabricated precision on an unknowable question, you do not close the round. You get a follow-up email that never comes. The ecosystem is self-sustaining. Pune firms produce the most optimistic numbers. Reputable firms produce more conservative but still compounding numbers. Founders cite the most favorable version on slide 3. Investors do not verify it. The economy continues to refuse to compound smoothly. No one is penalized for the miss because no one is tracking the miss. CHART 5 · FORECAST RELIABILITY DECAY The longer the horizon, the more methodology stops mattering Estimated forecast reliability by horizon and publisher tier. Illustrative framework based on documented accuracy research. Source: Berkeley Haas forecaster accuracy study; St. Louis Fed recession forecast accuracy; author's framework What to Do With a CAGR When You See One If the source is rigorous, the industry is not technology, and the horizon is one to two years: Respect the rigor, toss the prediction. If the source is from Pune, the industry is technology, and the horizon is 5 to 10 years: Don't respect the rigor, toss the prediction. ---------------------- If you did get all the way here, I will post a second part to this, in the next week or two, discussing cynicism relates to your direct search, valuation and broker engagements.