> even if it got 50%
Good analysis but a 50/50 split after costs would be unusually high. That level of back-end usually requires investing significant capital in a film. It's a high-risk business like VC where most releases lose money which is (hopefully) offset by a few outsized hits.
The shape of a pure IP license deal is typically more like ~10% share after costs. But that's generally for a property like Fallout with at least some known value. Beyond their hits, MSFT has a back catalog of game titles with unknown cross-media potential large enough to make the likelihood of a break-out like KPop fairly high. I suspect the deal stalled at $500M offer / $800M ask over theatrical back-end with a minimum production guarantee.
But this isn't a normal IP license deal nor just a streaming rights or even a global distribution deal, it's an exclusive IP library production deal which has different dynamics. MSFT spent over $100B acquiring game studios in an ill-advised strategy to turn Game Pass into the "Netflix of Games". While that seemed like a reasonable high-risk, high-reward bet in 2018-19, it's now clear no one can be the Netflix of games. The unit economics don't work because people watch a lot more movies and shows in a year than they buy games. MSFT bet $100B that a large number of gamers would pay >$20/mo but after the COVID shutdowns ended, ASP growth stalled below $20 and the higher the ASP got, the more the churn went vertical. Oops.
Now Sharma's job is to recoup as much of that $100B as possible by 1) Selling the shit out of all the evergreen hits they own like CoD, and 2) Milking that huge library of non-hit titles to harvest as much revenue as she can while spending as little as possible. Netflix sees this and thinks "Let's toss in a low-ball offer and see how hungry they are for quick cash." But Sharma has already modeled out a revenue projection for her evergreen titles and knows what they'll make over the next 5-6 years (to within +/- ~10%). CoD gonna keep CoD-ing, she just needs to turn the crank competently and not fuck it up. Her only upside here is to find a way to harvest more than the 'expected value' from the pool of unknown-value IPs, which are essentially lotto tickets. In terms of cross-media value, most of the unknowns will be losers and the tricky part is it costs money to audience test each IP to get an educated guess on if it has cross-media potential and then a lot more money to produce any that look good.
My guess is the deal fell apart over minimum production guarantees (MPG). Netflix says "We'll give you $500M and cover all the development, production and distribution costs but we also keep ALL the upside." Sharma says "Do I look as dumb as Sony Pictures? :-) You'd be tying up exclusive rights to all my IP and you can control your downside costs by simply not testing or producing any but the safest-bets. If any of those become the next KPop, I'd like a little back-end upside and I want some guarantee you're going test at least half the back catalog and actually produce a couple of the unknowns so I at least have a chance one becomes the next KPop." Netflix does the math and says "You want us to guarantee spending ~$20M in development testing plus ~$100M in production on unknowns on top of $500M for rights? Hmmm, what's the price to just test, develop and produce as much or little as we want and still own all the upside?" MSFT comes back with "If promising to spend $620M feels 'too rich' then a full buyout to tie up our library, develop all or none, and keep all the upside is $800M. Alternatively, let's explore actually partnering. If you want a lower front-end price, cut us in on some back-end upside or put some more skin in the game with MPG. Or lets split the known and unknown IPs into two pools with different back-end/MPGs, or maybe a sliding scale, like you pay no back-end on anything until after you've earned 3X its prod cost?" Netflix: "Yeah, nah. $500M, take it or leave it."
In reality, straight buyouts are very rare even on single-title, distribution-only deals. They were probably already deep in negotiating pool splits, back-end rates and sliding MPG scales with Netflix hoping MSFT would make a Sony-sized IP mistake and MSFT trying not to become the industry's next cautionary tale. Even the infamous Sony/Netflix KPop deal was something like $125M to buy a 'first-look' right on just a sub-pool of unknown IPs in development with a buyout option on any that they liked for exclusive worldwide distribution (theatrical & streaming) with no back-end - and not just on the first release but all derivatives, forever. IIRC, exercising the distribution buyout on a movie required an additional upfront payment based on a percentage of the production cost Sony had spent, which was around $20M for KPop. In the midst of COVID shutdowns, Sony thought they were turning long-shot option rights on a pool of unknown, low-budget probable 'dogs' into a little guaranteed cash. I mean a niche animated musical targeted at teen girls in Asian markets? Some (now-former) junior Sony exec is like, "Sure, Netflix. If you can sell that to your boss and get a check here by the end of the quarter, we'll let you own ALL that upside (muffled giggling)." For over a century now, the lesson to never give up ALL the back-end keeps being re-learned in Hollywood.
I think Sony cutting that deal with NetFlix was a special brand of dumb because Netflix has compiled more and better viewing habit data for longer than anyone ever. Giving NetFlix a 'first look' deal that lets them test your pre-release movies in their secret 'Magic Analytics Black Box' is a great idea, IF you keep some back-end on the first release AND all the sequels and ancillary. You share a little in Netflix's upside on a mega-hit but, most important, NetFlix activating their option is an invaluable signal that lets you double down early on any property the NetFlix 'Magic Box' thinks is a winner. So... all we really know here is MSFT is dumb enough to make a huge $100B strategic mistake but that's STILL not quite as dumb as Sony Pictures. :-)