Having worked at Amazon for well over a decade, here’s more or less what happened:<p>1. Product, in alignment with directors and VPs, have an overarching 3-5 year vision document<p>2. The document talks about “customer obsession” and a 2025 vision for “delighting customers”<p>3. Document has mostly fluff ideas, written by some MBA grad, on an H1B visa, who is feeling pressure to deliver a grand vision. Both to get promoted (eventually) and mainly to keep their job. If they get fired, they have 30 days to find a new company that will employ them with visa sponsorship. Or they have to move back to Hyderabad or whichever Indian city.<p>4. Document goes through several rounds of review, where this poor PM gets ripped a new one, many many times. In 1:1 with their manager, they are coached on having a thick skin and being able to handle verbal abuse and being made to feel small. Keep in mind, this person doesn’t really use Amazon Music themself. They just want to turn in their work, get paid, and go home.<p>5. Eventually, the product team aligns on increasing subscriptions X% YoY and increasing revenue Y%, by end of 2023 or mid 2024. Since we have some agreement already about to go in effect which will expand the music catalog, let’s leverage that. We’ll restrict the non-paid streaming option, and sell it as offering our premium customers (since that’s the metric we care about) more variety. Indeed we are “delighting customers” and “streamlining the music listening and discovery journey” <—- I shit you not. This is exactly how these docs are written.<p>6. This is reviewed and signed off by layers of directors and VPs. Keep in mind none of these people use Amazon Music. If anything, some of them might even have disdain for music.<p>7. The changes roll out. There’s an A/B test, designed in a way to not completely obscure reality but still give us what we want. End customers unhappy.<p>8. Product team works with technical team to look at specific A/B test results. Results show subscriptions are actually up. The measured increase is statistically significant. Music listening time is overall flat or increased slightly as well, but definitely not a regression.<p>9. Product team celebrates. Product team knows the product is worse off, customers are unhappy, etc. but the “metrics” we have show a win.<p>10. PM hangs out for several years. Realizes they can’t get promoted to a principal and leaves for another company, or they get annoyed with the whole visa back and forth process. They move to Canada and find some other Amazon team there, or they just go back to India once the initial stocks vest.<p>11. Cue the next round of MBA talent, looping through step 1.