Buyers track one clock and Google tracks two, and opening the file 60 to 90 days out is what separates a 12 percent rate from a 25 percent one
Two calendars govern the deal. One of them is published quarterly on Form 8-K and the other sits in your contract folder. Only one side has consistently been reading both.
Prepared by Redress Compliance · August 17, 2026 · Google Cloud advisory. Redress Compliance advisory engagement file, 2024 to 2025.
Executive summary
Buyers open the file 60 to 90 days before expiry, and the failure mode is nearly universal. By that point Google already knows you have nowhere to go, and none of the structural levers work from that position.
Your clock is the EDP expiry or the Workspace anniversary. It is the only date most procurement teams track, and it is the less useful of the two.
Alphabet's clock is the calendar fiscal year, and it is public. Q4 closes on December 31 and the four reporting dates are filed on Form 8-K, so the pressure calendar is available to anyone who reads it.
Above $5 million annually on infrastructure, open the internal process nine months before expiry. That is the lead time the structural levers need, not a counsel of perfection.
The two clocks
One calendar sits in your contract folder. The other is filed publicly four times a year. The deal is decided by how they overlap.
| Clock | What it is | Who tracks it |
|---|---|---|
| Your clock | EDP term expiry, or the Workspace anniversary | Your procurement team, usually alone |
| Alphabet's clock | Calendar fiscal year, Q4 closing December 31 | Google, and anyone who reads the filings |
| The reporting dates | Four per year, filed on Form 8-K | Public, including the July 22, 2026 filing for the quarter ended June 30 |
| The overlap | Where your renewal lands in their quarter | Almost nobody, on the buyer side |
The asymmetry is entirely one of attention, not information. Alphabet's fiscal calendar is public and its quarterly reporting dates are filed. Nothing prevents a buyer from knowing exactly where their renewal falls in the seller's pressure cycle, and from choosing to move it. What actually happens is that the buyer tracks their own expiry date, opens the file when it approaches, and arrives inside a quarter that was chosen for them.
The levers all need lead time that a 60 day file does not have
The failure mode is nearly universal in our experience: buyers open the file 60 to 90 days before expiry. That timing is not slightly tight, it is structurally too late, because every lever that moves a Google Cloud number materially requires more runway than it leaves. Term length, tier tables, shortfall exposure, and commitment structure are all negotiable in principle. None of them are negotiable in a conversation that starts after Google already knows you have nowhere to go.
The gap this produces is not marginal. The practical difference between managing both clocks and managing one is the difference between a 12 percent Enterprise Discount Program rate and a 25 percent rate with Committed Use Discounts stacked on top of it, which on covered compute adds another 20 to 55 percent against list. That is two distinct mechanisms compounding: a better EDP rate on the whole spend, and a CUD layer that only pays if the commitment was structured with enough time to size it honestly.
For any infrastructure commit above $5 million annually, open the internal process nine months before expiry. That figure is not a counsel of perfection; it is what the sequence actually takes. Usage has to be analysed before it can be committed against, the alternative has to be built before it is credible, internal approval has to exist before an escalation means anything, and the commitment structure has to be designed rather than accepted. Nine months is the sum of those, not padding around them.
One sequencing point is worth stating separately for estates running both products. If you are negotiating infrastructure and Workspace together, treat the infrastructure clock as the master and pull Workspace into it. A Workspace seat block is a cheap concession for a Google team chasing a large cloud commitment and an expensive one for a Workspace rep negotiating alone. Running them separately means paying full attention twice and getting leverage neither time. The commitment mechanics sit in the CUD negotiation brief, the terms in contract terms, and the library in the Google Cloud practice.
- Your quote benchmarked against 500,000+ real closed deals, adjusted for size, region, and industry
- Commitment sized against measured usage rather than a forecast built to justify a rate
- Every risky clause flagged with the exact quote, the page, and the replacement language
The sequence
- Open the internal process nine months out above $5 million annual infrastructure spend, because that is what the lever sequence takes rather than a safety margin.
- Locate your renewal inside Alphabet's fiscal calendar, which closes Q4 on December 31 and files four reporting dates publicly.
- Treat the infrastructure clock as the master and pull Workspace into it, since a seat block is a cheap concession inside a large cloud deal and an expensive one alone.
- Analyse usage before committing against it, so the CUD layer is sized on what actually runs rather than on a forecast built to justify a rate.
- Build the alternative before you need it, since credibility is a function of preparation time and cannot be assembled inside a 60 day window.
- Secure internal approval to escalate early, because an escalation you have not pre-agreed is a bluff and gets treated as one.
What the timing pattern showed
From the Redress Compliance advisory engagement file, 2024 to 2025:
When buyers open the file. Nearly universal, and structurally too late for any lever that requires a credible alternative.
The practical difference between managing one clock and managing both, before Committed Use Discounts stack on covered compute.
Stacked CUDs add a further 20 to 55 percent against list on covered compute, which only materialises where the commitment was sized with enough lead time to be honest about usage.
Alphabet reports on a calendar fiscal year with Q4 closing December 31 and four public reporting dates filed on Form 8-K, including the July 22, 2026 filing covering the quarter ended June 30.
Watch the briefing · 6:33Is There Leverage in a Google Cloud Deal? Five TacticsWhere leverage actually comes from, and how much lead time each source needs.
Your first five moves
- Put your EDP expiry and Alphabet's quarter ends on one calendar, so the overlap is a decision rather than an accident.
- Open the internal process nine months out for any infrastructure commit above $5 million a year.
- Run the usage analysis first, since the CUD layer cannot be sized honestly without it and it is the larger of the two discounts.
- Consolidate Workspace into the infrastructure negotiation, where a seat block is a cheap concession rather than the whole deal.
- Build the alternative and the escalation path early. The Google Cloud practice runs the sequence with you.
Frequently asked questions
When should a Google Cloud negotiation start?
Nine months before expiry for any infrastructure commit above $5 million annually. That is the time the lever sequence takes, not a safety margin added on top of it.
When do buyers actually start?
60 to 90 days before expiry, and the pattern is nearly universal. By then Google already knows you have nowhere to go, which removes the credibility every structural lever depends on.
What are the two clocks?
Yours is the EDP term expiry or the Workspace anniversary. Alphabet's is the calendar fiscal year, with Q4 closing December 31 and four reporting dates filed publicly on Form 8-K.
Why does the second clock matter?
Because it tells you where your renewal falls in the seller's pressure cycle. The information is public. The asymmetry is one of attention rather than access, which makes it unusually cheap to close.
How much is the timing worth?
The practical gap between managing both clocks and managing one is a 12 percent EDP rate versus a 25 percent rate with CUDs stacked on top, which add another 20 to 55 percent against list on covered compute.
Why do CUDs depend on lead time?
Because they only pay if the commitment is sized against usage that actually runs. Sizing honestly requires analysing usage first, and that analysis cannot be produced inside a 60 day window alongside the negotiation itself.
Should Workspace and infrastructure be negotiated together?
Yes, with infrastructure as the master clock. A Workspace seat block is a cheap concession for a Google team chasing a large cloud commitment and an expensive one for a Workspace rep negotiating alone.
What has to happen before the quote?
Usage analysed, alternative built, internal approval to escalate secured, and commitment structure designed. Each takes time independently, and the nine month figure is their sum rather than padding.
Does an escalation threat work late in the process?
Not if it has not been pre-agreed internally. An escalation you cannot actually execute is a bluff, it is read as one, and using it late tends to confirm rather than improve your position.
Can the renewal date itself be moved?
Often, and it is worth asking early. Choosing which of Alphabet's quarters your renewal lands in is a lever available to any buyer who is looking nine months ahead and unavailable to one looking sixty days ahead.
Negotiating Google 7: Google's Playbook, and the Counters
Five moves Google runs in almost every account: land and expand credits, quarter end pressure, the engineered renewal cliff, deadline theater, and eligibility policing, each with the counter that works.
