Full narration of the briefing. Click a section heading to jump the player to that moment.
AWS does not run a hard sell. It runs a patient one, and the moves are consistent enough across accounts that you can name them while they are happening. That is the whole defence. Five moves, and the counter to each.
None of them are dishonest. All of them are designed to end with you committing more than you needed to.
Move one is the forecast that is not yours. AWS arrives with a growth model built on your telemetry, and the opening commitment typically lands fifteen to thirty percent above the customer's own number, often with a year one built on roughly twenty percent growth over your trailing six months. The counter is not to argue about their forecast. It is to refuse to negotiate on it at all.
Your floor is workloads already running plus migrations already under contract. Everything else is a hypothesis, and AWS is welcome to fund hypotheses with credits and ramp rather than have you guarantee them. Commit at eighty to ninety percent of your own model, never a hundred percent of theirs.
Move two is funding that buys your future baseline. Migration money is real: broadly a quarter of first year run rate, paid against tagged spend. It covers the double bubble while you run old and new estates at once, and it is worth having. But understand what it does to your next negotiation.
The credits move workloads onto AWS and lift your run rate permanently, while the credits themselves retire no commitment. So you carry a higher baseline into the renewal without having banked any progress for it, and your partner is typically paid a percentage of your post migration run rate, which means nobody in the room is paid to right size you. Price it honestly: a one off credit worth a quarter of year one is about five percent annualised across a five year workload. If accepting it costs you two points of permanent discount, you lost.
Move three is the first offer. It typically sits five to twelve points below what is achievable, and the tables in circulation are wide enough that buyers cannot tell. The counter is the term sheet you wrote in episode five, because a number you set before the meeting cannot be anchored by the number they bring to it. And ask the breakpoint question rather than the discount question.
Show me the rate at my number, plus a hundred thousand, plus a quarter million. It converts a haggle into arithmetic, and it exposes whether more commitment actually buys anything.
Move four is the scarcity story. This one is newly true and therefore more persuasive. AWS has said publicly it cannot meet demand this year, expects the same next year, and already has capacity reserved into the year after. Commit now or lose access is a real argument for accelerators.
It is not a real argument for general compute, storage, database and networking, which is most of your bill and where AWS is still losing share to faster growing competitors. So split the conversation. Take the scarcity seriously where it exists, and refuse to let it price the commodity eighty percent of your estate.
Move five is the renewal that reprices your discipline. Everything your FinOps team achieves after signature lowers the spend that retires your commitment, and at renewal AWS rebaselines on your current run rate with a discount percentage that can step down rather than up. So optimise before you commit, put the re-compete in the diary the day you sign, and treat every renewal as contested. A buyer who arrives with a refreshed alternative and three years of their own evidence gets a different conversation than one who arrives assuming continuity.
And one thing worth saying plainly, because it cuts against the fear most buyers carry. Shortfall is a plain gap true up. There is no penalty multiple, missing your number does not trigger a discount clawback, and the filed contracts show clawback only fires if AWS terminates you for cause. AWS restructures far more often than it enforces.
Which means the real risk is not catastrophe. It is quietly paying for capacity you never used, year after year, because nobody re-read the agreement. That is a much more common ending, and a much less dramatic one.
One last point. At Redress Compliance we sit across from this playbook for a living, on pure contingency: twenty five percent of what we save you, and nothing saved, nothing paid. Next episode: sizing and structuring the commitment itself, clause by clause.
Redress Compliance works on contingency: our fee is 25 percent of what we save you. Nothing saved, nothing paid. Independent, buyer side only, never vendor funded. Want Redress to contact you? Reach out and we respond the same day.
Talk to a AWS negotiator