A 12 months in the past, Invoice D’Alessandro and I sat all the way down to report our predictions for 2025. We made calls on all the things from Trump’s tariff coverage to Google’s search dominance.
A few years, revisiting predictions is a humbling train. We’ve had some absolute stinkers up to now—predictions so off-base they’re virtually spectacular of their wrongness.
How did we fare for 2025? Learn on to seek out out.
—
Prediction #1: Trump = Tariffs + Tax Cuts
My Name: Trump would increase tariffs on China round 10%, and we’d see tax coverage like QBI extension and additional tax cuts.
What Occurred: I obtained the route spot-on, however badly underestimated the magnitude. China tariffs debuted at 150% earlier than settling round 30-40% efficient charges. My “believable 10%” name wasn’t even within the ballpark.
My “believable 10%” name wasn’t even within the ballpark. China tariffs debuted at 150% earlier than settling round 30-40% efficient charges.
On the tax entrance, I nailed it. The Large Stunning Invoice got here by way of with QBI extension and extra business-friendly tax cuts as predicted.
Grade: 8/10. Proper on the traits, mistaken on the size.
#2: Google Exhibits First Actual Weak point
My Name: Google would drop beneath 90% search market share for the primary time in 10 years as AI assistants like ChatGPT stole queries.
What Occurred: Google did dip below 90% in early 2025—the primary time in a decade they’d fallen beneath that threshold. For a short second, it seemed just like the prediction would play out completely.
However Google didn’t roll over. Gemini 3 turned out to be legitimately sturdy, and so they’ve achieved a formidable job integrating AI into search. The armor cracked, however they patched it quick.
Grade: 6/10. They stumbled, however recovered sooner than I anticipated.
#3: search engine marketing for AI LLMs Emerges
My Name: A brand new business can be born round rating inside ChatGPT, Claude, and different AI fashions—primarily search engine marketing for giant language fashions.
What Occurred: “GEO” (Generative Engine Optimization) is now an actual factor. Horrible identify, however the self-discipline exists. Instruments have launched, consultants are promoting companies, individuals are speaking about it at conferences.
Is it groundbreaking? Not but. Most of it nonetheless appears to be like like conventional search engine marketing carrying a masks. However the business has emerged precisely as predicted.
Grade: 10/10. Known as it.
#4: Bitcoin Strategic Reserve
My Name: The US would set up an express Bitcoin reserve stance and BTC would finish the 12 months round $150K.
What Occurred: We’re retaining seized Bitcoin as an alternative of liquidating it, which is one thing. However we’re not actively shopping for extra for the treasury.
Bitcoin touched $125K—tantalizing near my $150K name—earlier than crashing to round $90K the place it sits at present.
Half proper on the reserve stance. Manner mistaken on value.
Grade: 5/10. This one stings.
#5: Nuclear Energy Re-emerges
My Name: Huge energy demand from AI information facilities would push nuclear energy again into the highlight as a viable vitality answer.
What Occurred: 4 government orders supporting nuclear in Could. World nuclear funding up roughly 10% year-over-year.
This was the most secure guess of my 5 predictions—fairly apparent when you had been listening to AI’s vitality urge for food. However I known as it appropriately.
Grade: 9/10. Correct, if not significantly daring.
Invoice’s 2025 Predictions
Invoice D’Alessandro has been my partner-in-crime for making predictions for the previous decade. Right here’s a recap and evaluation of his predictions made a 12 months in the past:
#1: Offshoring Ramps Up, Staffing Mannequin Shifts
Invoice’s Name: The person-in-the-middle staffing company mannequin would give method to conventional placement for worldwide hires. E-commerce manufacturers would more and more rent international-first by way of direct placement as an alternative of ongoing staffing spreads.
What Occurred: Worldwide-first hiring is completely the default now. Invoice doesn’t know a single entrepreneur who doesn’t rent worldwide first for many roles.
However the man-in-the-middle companies are nonetheless extracting large income. The position mannequin is gaining floor, however it hasn’t swept the business the best way Invoice predicted. Motion in the proper route, simply slower than anticipated.
Invoice’s Grade: 50%. Directionally proper, velocity mistaken.
#2: Margins Rise, Advert Prices Increase to Eat Them
Invoice’s Name: AI and offshoring would make groups leaner and elevate margins. However as a result of Meta and Amazon run aggressive auctions, elevated profitability would push up CPAs as sellers bid extra aggressively.
What Occurred: This performed out precisely as predicted. CAC up double digits in 2025. CPMs up double digits throughout Meta and Amazon.
Invoice known as this “essentially the most sensible enterprise mannequin on earth” and it’s proving diabolical. Any margin positive factors companies squeeze out get instantly plowed again into advert spend. The public sale mannequin dynamically expands to seize surplus.
Invoice known as this ‘essentially the most sensible enterprise mannequin on earth’ and it’s proving diabolical. Any margin positive factors get instantly plowed again into advert spend.
This development will speed up into 2026 and past.
Invoice’s Grade: A+ / 10 out of 10. Nailed it.
#3: Proudly owning Manufacturing Turns into Essential Moat
Invoice’s Name: Vertical integration, particularly US-based manufacturing, would develop into important for higher margins, sooner innovation, stronger IP, and provide chain management.
What Occurred: Everybody in Invoice’s community is constructing factories or aggressively transferring towards vertical integration. Tariffs accelerated this development, however the actual drivers are management, pace, and strategic benefit.
With AI commoditizing advertising and marketing and content material creation, product is more and more turning into the one actual moat. Proudly owning your manufacturing offers you the flexibility to iterate sooner and shield your differentiation.
With AI commoditizing advertising and marketing and content material creation, product is more and more turning into the one actual moat.
Invoice’s Grade: A- to 10/10. Spot on.
#4: Amazon Haul Drives Two-Tier Ecosystem
Invoice’s Name: Amazon Haul (their mobile-only, ultra-cheap, direct-from-China part) would succeed and develop into the house for unbranded cut price merchandise. Amazon.com would tighten insurance policies on non-branded abroad sellers and deal with vetted, brand-registered merchandise.
What Occurred: Tariffs kneecapped this prediction earlier than it may achieve momentum. De minimis reform additionally gutted the direct-from-China transport mannequin that Haul relied on.
There are early indicators of a crackdown—Amazon now experiences Chinese language sellers to the Chinese language authorities for taxation, which is an enormous win for US-based sellers. However general, this didn’t play out the best way Invoice anticipated.
Invoice’s Grade: 2-3/10. Clear miss, principally as a result of coverage modifications Invoice didn’t foresee.
#5: Crypto and Fairness Markets
Invoice’s Name: Bitcoin would contact $150K, however finish beneath $100K after hype exceeded coverage follow-through. The S&P can be up 20% mid-year on animal spirits and M&A exercise, then give again positive factors to complete flat if Doge effectivity efforts really materialized.
What Occurred: Bitcoin hit $125K earlier than crashing to round $87K at present. The hype-then-crash sample was dead-on, even when he didn’t nail the precise numbers.
The S&P is up 16-17% year-to-date and holding sturdy, not flat as predicted. Why? As a result of the Doge austerity efforts turned out to be a “large nothing burger”—zero actual cuts, huge spending invoice, no affect on danger property.
Invoice’s logic was sound: no austerity means danger property keep elevated. He simply guess on austerity really taking place.
Invoice’s Grade: A-. Received the Bitcoin sample proper, partially proper on markets.
—
This was legitimately our greatest 12 months for predictions. Can we do it once more?
This Friday on the podcast: Invoice and I are recording our 2026 predictions, together with a quite ludicrous guess to see who’s extra correct this 12 months. Loser buys the winner a steak—as graded by AI.
Need to see what’s coming BEFORE it hits?
One of the best ways to foretell the longer term isn’t crystal balls—it’s tapping into 1,000+ savvy 7- and 8-figure operators who’re seeing traits emerge in real-time.
When tariffs began hitting, our members had been sharing methods weeks earlier than the headlines. When advert prices spiked, they had been already testing options.
Be part of the eComFuel neighborhood to get forward of what’s subsequent, not scrambling to catch up.
