NUTANIX DAILY

A month ago I told you to treat the Broadcom renewal letter as your discovery document. Read it with the customer, walk through the 72 core minimums and the bundle math, and let the letter do the qualifying for you. Since then something interesting has happened: the letters stopped making news. Nobody is forwarding you a quote with a 1,200 percent increase circled in red anymore. The renewal season got quiet.

So let me say the thing first: yes, the storm is calming. I am not going to pretend otherwise, because your customers can feel it too. But calm is doing a lot of work in that sentence, and if you walk into your next meeting thinking calm means over, you are going to misread every signal in the room. The storm did not pass. It priced in. The market learned the new rules, built them into the budget, and moved the fight from the headlines to the negotiation table. Here is what that means for the four conversations you are about to have.

"We Already Negotiated. We Are Fine for Three Years."

What happened: the negotiability of Broadcom quotes went from rumor to benchmark. Licensing consultancies that have now sat through 30 to 40 of these renewals report that opening quotes of 2x to 5x the prior spend are routinely settling at 1.3x to 2x. Customers who pushed back hard, brought an exit model to the table, and were willing to walk got real concessions.

The objection you will hear: we got our number down, we signed for three years, this problem is solved. Why are we talking about migration?

How to respond: congratulate them, genuinely, because they did the work. Then ask one question: what did the negotiation cost you? The benchmark data shows the customers who got the best settlements were the ones who modeled moving 20 percent or more of the estate to an alternative. Broadcom prices against the risk of you leaving. Which means their good deal exists because an exit plan exists, and that exit plan has a shelf life. If they let it go stale, the next renewal opens at 2x to 5x again and they will have nothing to price against.

The caveat to be upfront about: a 1.3x to 2x settlement is a real win versus the opener, and for some estates it genuinely beats a rushed migration. Say that out loud. It builds the credibility you need for the next sentence.

What to tell your customer: the leverage that won this renewal needs to be real by the next one. Keep the exit model funded and current, because Broadcom rewarded you for having it, not for loving them.

"The Horror Stories Faded. This Seems Manageable Now."

What happened: the scary anecdotes dried up, but the structural number underneath is louder than any anecdote. Multiple analyses of Broadcom's VMware business point to the same pattern: revenue grows around 13 percent while the average customer renews only about 25 percent of their previous VMware estate. Read those two numbers together. Broadcom is making more money from less footprint.

The objection you will hear: those 1,200 percent stories were last year. Everyone we talk to seems to have worked it out.

How to respond: they have worked it out, and this is what worked out looks like. The average customer kept a quarter of the estate on VMware and is moving, or planning to move, everything else. The horror stories faded because the market stopped being surprised, not because the economics improved. Manageable and shrinking are not opposites. They are the same strategy seen from two angles.

The caveat: some workloads will stay on VMware for years and should. Deeply integrated estates with regulatory support requirements are renewing on purpose, not from inertia. Do not pitch a 100 percent migration to a customer who only needs a 75 percent one.

What to tell your customer: the question is not whether to leave VMware. It is which quarter of your estate earns the right to stay, and what the plan is for everything else.

"Moving 500 VMs Is Riskier Than Overpaying. We Will Ride It Out."

What happened: the migration wave is real but slower than the headlines suggested. Most organizations are 12 to 24 months into planning and not yet into execution. That gap between planning and doing is exactly where the ride it out logic lives.

The objection you will hear: migration risk beats cost risk. We can absorb the premium for another cycle.

How to respond: riding it out is a strategy with a hard floor under it, and the floor has a date. vSphere 7 support ends entirely on April 2, 2027. That is less than 23 months out. Back in April I walked through the cost conversation in three meetings, and the same discipline applies here: model the premium they are absorbing per year, then put it next to the migration cost spread over the same window. Absorbing the premium is a choice. Absorbing it without a model is just drift. And the second migration wave, the one happening right now, is made of customers who said ride it out at their last renewal and started moving the day after they signed.

The caveat: they are right that a 500 VM migration carries real risk, and anyone who tells them otherwise is selling. The honest framing is that risk managed on your own calendar beats risk forced by a support deadline.

What to tell your customer: count backward 18 months from your next renewal date or from April 2027, whichever lands first. That is the month your migration planning has to start for the decision to still be yours.

"You Vendors Manufactured This Panic. The Renewal Went Fine."

What happened: enough renewals have now closed at survivable numbers that the panic of the last two years looks, in hindsight, overcooked. Some of it was. Vendors on every side of this fight, ours included, ran hard at the fear.

The objection you will hear: the urgency was sales theater. Our renewal went fine, so the rest of the pitch is suspect too.

How to respond: concede the first half. The 1,200 percent anecdotes were real but rare, and they did more work in vendor decks than they did in most budgets. Then pivot to where the data comes from now: the 2x to 5x openers, the 1.3x to 2x settlements, and the 25 percent estate renewals are not Nutanix numbers. They come from licensing consultancies and industry analyses whose business is advising VMware customers on renewals, including plenty who stayed. The case for planning an exit stopped depending on horror stories about a year ago. The boring benchmark data makes it better than the anecdotes ever did.

The caveat: if your own deck still leads with the scariest quote you ever saw, retire that slide. It worked in 2024. In 2026 it reads as theater and hands the customer this exact objection.

What to tell your customer: ignore every vendor anecdote, including ours. Pull the public benchmark data, put your own renewal next to it, and decide from the numbers. We are happy to be graded on that math.

Why Calm Is the Best Selling Weather You Will Get

One more thing before the whiteboard, because it ties back to the bench conversation from late April. Panic compresses decisions; calm extends them. When a customer is staring at a renewal deadline with a 4x quote, you get a fast, small, defensive deal if you get one at all. When that same customer has settled at 1.6x and exhaled, you can finally have the architecture conversation: what the platform should look like in 2028, where AI workloads land, what the 75 percent that leaves VMware actually moves to.

The quiet you are seeing is not the demand disappearing. It is the demand maturing into projects worth designing. That is the work we are built for, and the window for it is open right now, while nobody is panicking and the 2027 floor is still comfortably far away. It will not stay comfortably far away.

What Belongs on the Whiteboard

1. Calm is phase two, not the all clear. The storm priced in. Quiet customers are negotiating, shrinking, and planning. Treat every relieved renewer as a wave 2 prospect, because that is statistically what they are.
2. Openers of 2x to 5x settle at 1.3x to 2x. The benchmark data is public. No customer should ever accept a first Broadcom quote, and their discount lives or dies on having a real exit model.
3. The average renewal keeps 25 percent of the estate. The other 75 percent is the project pipeline. Ask every customer which quarter of their estate earns the right to stay.
4. Renewal date minus 18 months is the trigger. Wave 2 customers move after they sign, not before. Count backward from their next renewal and book the planning conversation for that month.
5. April 2, 2027 is the floor under everything. vSphere 7 support ends entirely. Every settled, calm, fine for now customer still has that date in front of them. So do you.

If you are a BASA looking at your calendar for the next two weeks, here is what is actually on it. Half of your accounts are sitting on a Broadcom renewal quote that landed in the last 60 days and does not look anything like the last one. And every one of those customers is going to glance at Nutanix Q3 FY26 earnings on May 27 before they sign anything either way.

That is two questions, one customer meeting, sixteen days to handle both. The SAs who walk in fluent on both come out of May with a pipeline. The ones who only have half the conversation come out with a stalled deal.

Let me walk through what is in the renewal letter, what is in the earnings setup, and how I am coaching the BlueAlly bench to handle both in a single conversation.

What Is Actually In the Renewal Letter Now

If you have not personally read a current Broadcom VMware renewal quote in 2026, you are walking into customer meetings blind. The letter has new mechanics in it that the customer's procurement team has noticed but their IT leadership often has not.

Here is the short list. Memorize these before your next call.

One. The 168 SKU catalog has collapsed to four bundles. VMware Cloud Foundation. VMware vSphere Foundation. VMware vSphere Standard. VMware vSphere Enterprise Plus. That is it. Everything that used to be a la carte (NSX add ons, vSAN tiers, Aria components, Horizon, Tanzu) now lives inside one of those four buckets or is sold separately at the enterprise level. The customer who used to buy exactly what they needed cannot do that anymore. They buy the bundle that covers the thing they need, and they pay for the rest of the bundle whether they use it or not.

Two. The minimum purchase floor moved from 16 cores to 72 cores per CPU socket, then to a 72 core per command minimum. If your customer's environment lives on smaller hosts, they are now paying for cores they will never light up. This is the number that turns SMB and mid market accounts into Nutanix conversations overnight. The math gets ugly fast when a four host cluster with two 24 core processors per host is suddenly being licensed at 72 cores per CPU.

Three. The 20 percent late renewal penalty. If a customer lets their support agreement lapse and then comes back, they pay a 20 percent premium on top of the new pricing to be reinstated. That penalty exists specifically to prevent customers from using a lapse as a negotiating tactic or as a window to evaluate alternatives. Read that again. The renewal terms now actively work against the evaluation conversation you are trying to have.

Four. Seven year subscription TCO is roughly 63 percent higher than the equivalent perpetual license investment would have been. That number is the answer to the procurement officer who says, "subscription is just how everything works now." Yes, subscription is the model. But Broadcom's subscription model is priced at a premium that the old perpetual model never carried, and the curve compounds with every renewal cycle.

Five. Real renewal hikes documented in the wild range from 150 percent to 1,200 percent year over year. A UK university that I will not name here went from a 40,000 pound annual VMware spend to a 500,000 pound annual quote on renewal. That is not a misprint. That is the customer experience. The 1,200 percent end of the curve is rare but it exists. The 150 to 300 percent band is common. The customer who is bracing for a normal annual increase and gets a number like that is either signing under duress or starting the evaluation conversation they should have started a year ago.

If you walk into a renewal discussion without these five numbers, the customer is going to teach them to you in real time. Walk in already fluent in them.

The Renewal Letter Is Your Discovery Doc

Here is the reframe that changes everything. The BASA who treats the renewal letter as the customer's problem loses. The BASA who treats it as the discovery document for a Nutanix conversation wins. Same piece of paper, different posture.

Ask the customer to read it with you. Not metaphorically. Literally. Put it on the screen in the meeting. Then walk through these five questions in this order.

One. Which of the four bundles did Broadcom put you in, and what percentage of that bundle do you actually use? Most customers cannot answer this on the first pass. That is fine. The act of trying to answer it is the discovery.

Two. Look at your core counts per host. Multiply by 72. Is that higher than what you would have licensed under the old model? Now you have the wasted core spend number. That is the most weaponizable number on the page.

Three. What is your renewal date and how many calendar days are between today and that date? If it is more than 120 days, you have time to do a proper Nutanix POC. If it is less than 90 days, the conversation shifts to either a short term renewal extension while you evaluate, or a parallel path where you sign the renewal and start the migration plan immediately.

Four. Have you priced what a Nutanix replacement looks like for the same compute footprint? Most customers have a vague idea. Vague does not survive the procurement conversation. They need real numbers from a real partner. That is where I come in. That is where BlueAlly comes in for our customers, with the 145 site migration we ran for a wholesale food distributor as the proof point that we have done this at scale before.

Five. Who owns the renewal signature? IT? Procurement? The CFO directly? This is the sneaky important one. Because the renewal letter is going to land somewhere, and the person it lands on is the person you need to be talking to. If IT got the letter but the CFO has to sign, you are talking to the wrong person.

Five questions. Fifteen minutes. You now know whether you have a real opportunity or whether the customer is going to sign the renewal and you should come back in eleven months.

Then the Customer Pivots and Asks the Second Question

Here is where most BASAs get caught off guard. The renewal conversation is going well. The customer is leaning in. They are looking at the math. They are starting to see the migration story. And then somewhere around minute 40 of the meeting, they say it.

"But is Nutanix safe to bet on?"

Or some version of it. "What about that JPMorgan downgrade?" Or, "I saw the stock got hit last month." Or, "Are they going to be around in five years?"

This is not a hostile question. This is a procurement question wearing a strategy question's clothing. The customer is about to sign a multi year infrastructure commitment with a vendor that they have not historically bet the data center on. They need a 90 second answer that gives them confidence to put their name on the contract.

Here is the answer in the right order.

FCF first. Nutanix is forecasting 745 to 775 million dollars in free cash flow for fiscal year 2026. That is the number a CFO understands instantly. Free cash flow is what a company can return to shareholders, reinvest in the platform, or weather a downturn with. Nutanix is generating it at scale. Start there. Always.

FY guide second. Revenue guidance for the full year sits at 2.80 to 2.84 billion dollars, unchanged from the last earnings call. The thing the customer needs to internalize is the word unchanged. The analyst noise from JPMorgan and Northland was about pace and competitive friction inside the year. The company has not had to cut the annual number. That is the actual signal.

Customer base third. Rajiv Ramaswami said at .NEXT that Nutanix has migrated approximately 30,000 customers and is targeting the majority of VMware's 165,000 customer base in waves at renewal. Roughly one in five of VMware's installed base has already moved to Nutanix. Your customer is not the pioneer. Your customer is the wave.

FCF, FY guide, 30,000 migrated. In that order, said in 90 seconds, walks the customer back from the stock price tab they had open in another window.

How to Handle the JPMorgan Downgrade Question Specifically

If the customer cites the JPMorgan downgrade to Neutral with a 44 dollar price target, here is what an SA who has done the homework says.

JPMorgan moved from Overweight to Neutral on April 9 with a price target cut from 55 to 44. Northland followed with a price target cut from 76 to 53 a week later. The bear case from both notes is the same. Broadcom is counter punching on price. Constellation Research has reported that Broadcom is offering 10 to 20 percent discounts beyond standard when customers present a written Nutanix POC. That is real and it is lengthening some deal cycles.

Here is the part that did not make the customer's news feed. Rosenblatt initiated coverage with a Strong Buy at 60 dollars on April 3. The median analyst price target across 15 to 31 covering analysts sits at roughly 65 dollars. The consensus rating is still Buy. Two analysts moved more cautious in April. That is real. It is also not the consensus view.

The honest framing for the customer is this. The Street is debating pace, not direction. The platform expansion at .NEXT (NetApp alliance, NKP Metal, Enterprise AI 2.6, Cisco partnership renewal) is the multi year thesis. The Q3 print on May 27 is the next data point in the pace debate. One quarter does not unwind a multi year platform thesis. The customer is making a five year infrastructure decision, not a quarterly trading decision.

If you cannot deliver that paragraph from memory in front of a CIO, you are not ready for the meeting. Practice it.

What I Am Not Going to Pretend About

Q3 FY26 prints on May 27 after the bell. Consensus has settled at roughly 35 cents EPS on 686 million dollars in revenue. The company could miss. The macroeconomic backdrop is choppy. Federal IT spend is uncertain. Some of the .NEXT excitement was about products that are not GA yet. NKP Metal and parts of the NetApp integration land in H2 2026. The agentic AI story is real in substance but the customer reference base is still building.

If a customer asks what happens if Nutanix misses, the right answer is this. The fiscal year guide and the free cash flow guide have not been cut. The product roadmap is the multi year story. A single quarter that comes in light does not change what Nutanix can do for the customer's data center between now and 2030. The customer's question is about infrastructure, not stock.

That said, if the May 27 print misses the FY guide or the FCF guide, that is a different conversation. The bench should be ready to have it. Read the print live. Update your talking points by May 28 morning. Customer meetings that week need a current data point.

The Three Meeting Arc, Recompressed for May

Here is the meeting cadence that actually moves a renewal account through May and June.

Meeting one. Read the letter together. Bring nothing but the five discovery questions. No slides. No TCO model. The customer is doing the work in front of you. The output of meeting one is a list of the customer's pain points by number. Wasted core spend. Bundle waste. Renewal date. Signer.

Meeting two. The TCO conversation. Now you bring numbers. The customer's wasted core spend from meeting one. The Nutanix sizing for the same workload. The 145 site BlueAlly case study as the operational reference point. The earnings context answer ready to deploy when the customer asks the second question. The output of meeting two is alignment on a POC scope.

Meeting three. POC scoping. A real Nutanix POC against a real workload. Documented. Written down. Because that written POC is what triggers the Broadcom counter discount of 10 to 20 percent. And once it does, your customer has leverage on Broadcom that they did not have before, which they can either use to negotiate their renewal down or to fund a faster migration with the savings.

You do not have to win the renewal cycle to win the customer. You have to be the partner who walked them through the math honestly. The customers who renewed Broadcom under duress in 2025 are coming back in 2026 with three years of higher costs and a willingness to talk that they did not have last time. The conversation is patient. Be patient with it.

The Posture Shift Nutanix Just Made

One more thing worth knowing before you walk into the next meeting. Diginomica caught a pattern at .NEXT that other outlets missed. Nutanix has shifted away from anti Broadcom rhetoric and toward letting customers do the talking. Rajiv made one sharp public comment at .NEXT (the line about Broadcom not being a company that focuses on innovation) and then handed the microphone to Western Union, Everland, JetBlue, Lockton, and Purdue.

This is the SA posture for the next 16 days. Do not bash Broadcom in the customer meeting. Bash is a weak posture and the customer hears it. Bring proof points. Names. Numbers. Migration timelines. Operational outcomes. Let the customer experience of the migrated organizations be the argument. Your job is to put those stories in the customer's hands and ask what their version looks like.

What Belongs on the Whiteboard

1. "The renewal letter is your discovery doc, not your obstacle." Read it with the customer in the room. Five questions, fifteen minutes. By the end of meeting one you have the wasted core spend number and the signer. Those two pieces of information are the rest of the deal.

2. "72 cores, 4 bundles, 20 percent late penalty. Memorize these." If the customer teaches you the mechanics of their own renewal letter, you have lost the room. Walk in already fluent in the 72 core minimum, the four bundle collapse, and the 20 percent reinstatement penalty. Add the 7 year 63 percent subscription premium and the 150 to 1,200 percent hike range for context.

3. "When they ask is Nutanix safe, the answer is FCF, FY guide, 30,000 migrated. In that order." Free cash flow 745 to 775 million for fiscal 2026. Annual revenue guide 2.80 to 2.84 billion, unchanged. 30,000 customers already migrated against a 165,000 VMware base. Ninety seconds. Practice it.

4. "One quarter does not unwind a multi year platform thesis." JPMorgan and Northland are debating pace, not direction. The Q3 print on May 27 is the next data point in that debate. The customer is making a five year infrastructure decision, not a quarterly trading decision. Help them see the difference.

5. "Change the renewal calendar and you change the conversation." Customers stuck on a one year Broadcom panic cycle have no time to plan. Customers operating on a three year Nutanix planning cycle have the time to do the math, run the POC, and migrate at their pace. Moving the customer off of someone else's renewal calendar onto your planning calendar is the actual win. Everything else follows.

I am writing this Monday afternoon. Tomorrow at 8:30 in the morning I am walking into a meeting at BlueAlly HQ Atlanta that is on my mind tonight. Our CTO David Coulter is in the room. Our Field CTO Keith Manthey is in the room. Our Partner Manager Tom Capozza is in the room. From Nutanix, Chris Banes, VP of Systems Engineering for the Americas, and Chris Ramby, the Channel SE. A few others on each side, including Reuel Murray, our Principal Solutions Architect.

The agenda on paper is simple. Introductions. Nutanix Enterprise AI. BlueAlly's AI business goals and strategy. Three topics, one room, an hour and change.

What the agenda does not say is what the meeting is actually about. That is what this post is about. Because every BASA at every channel partner has this exact meeting at some point and most of us have not been coached on what is really happening in the room.

A note on timing for readers. The blog has been quiet for two weeks. Tomorrow's meeting is the right reason to come back to it, because partner alignment moments like this are exactly where the SA-to-SA coaching matters most. So this is what I am thinking about.

What the Meeting Is Actually Evaluating

Chris Banes has been at Nutanix for nine years. Before that, he led NetApp SE teams in the Midwest. He runs the entire Americas pre-sales SE organization at Nutanix. He is not coming to BlueAlly HQ tomorrow to learn what BlueAlly does. He already knows. Our Premier Reseller status in the Nutanix Elevate program is on file. Our pipeline is on file. The 145-site VMware-to-Nutanix migration we ran for a wholesale food distributor in 18 months is in the case studies.

Banes is coming to evaluate something else. He is coming to assess whether to invest more SE bandwidth in BlueAlly. That decision is not made by looking at slides. It is made by listening to whether our SAs sound like they can run customer POCs without his SEs babysitting.

That is the unstated question every partner alignment meeting orbits, and once you see it you cannot unsee it. The agenda says "BlueAlly's AI business goals and strategy" but what Banes is actually doing is reading the room for technical credibility. Can these architects deliver on what their CTO is describing? Or is the partnership going to require Nutanix SE hand-holding on every deal?

If the bench reads strong, BlueAlly gets more SE time, more deal-reg latitude, more shared opportunities. If the bench reads weak, the partnership stays surface-level. That is the real outcome of the meeting, even though no one will say it that directly.

The Difficult Moment Makes This Meeting More Important

Three weeks after .NEXT 2026, Nutanix is in a harder spot than the conference suggested.

JPMorgan downgraded NTNX from Overweight to Neutral on April 9, with a price target cut from $55 to $44. A week later, Northland Capital Markets cut Q3 FY26 EPS estimates and dropped their PT from $76 to $53. The stock closed at $40.41 on April 20, below both downgrade targets. Two analysts now have a more cautious view than the broader Buy consensus. The bear case is contained but it is not nothing.

The thesis underlying both downgrades is the part SAs need to understand. Broadcom is counter-punching. Constellation Research reports Broadcom is offering 10 to 20 percent discounts beyond standard when customers present a written Nutanix POC. That is the real story. It is not that Nutanix is losing the migration story. It is that the migrations are taking longer because Broadcom is opening the wallet to keep accounts they thought were leaving.

CEO Rajiv Ramaswami said the public quiet part out loud at .NEXT. Quote: "Nutanix has migrated about 30,000 customers and is targeting the majority of VMware's 165,000 customer base in waves at renewal." That is a bold statement. The analyst downgrades are the pushback on whether the math holds when Broadcom is fighting back with price.

This changes what tomorrow's meeting is actually about. Banes is not just evaluating bench credibility for partnership growth. He is evaluating which partners can win deals despite a competitor that is now discounting aggressively. SAs who walk in pretending the headwinds are not real look unprepared. SAs who walk in fluent in the analyst pushback AND with a clear point of view on how BlueAlly wins those harder deals earn the credibility.

Each Person in the Room Has a Specific Job

In a partner alignment meeting like this one, every BlueAlly attendee has a role and the roles are not interchangeable. Knowing your role is half of showing up well.

The CTO sets vision. David Coulter will articulate where BlueAlly is going on AI, what our strategic bets are, where Nutanix fits. His credibility comes from his background. He co-founded SovLabs, was CTO at CloudBolt, has deep VMware automation roots, and posts publicly about BlueAlly's partnership work with NVIDIA. He is the builder type, not a stage exec, so he will ask hard technical questions and expect substantive answers. His job is not to sell. His job is to align.

The Field CTO carries the AI narrative. Keith Manthey writes BlueAlly's bylined pieces on private GenAI, HPC at scale, and ISO 42001 responsible AI. His job tomorrow is to translate BlueAlly's AI strategy into the Nutanix-specific roadmap. Banes will read his fluency on Enterprise AI 2.6, GPU density on AHV, and how Nutanix maps to ISO 42001 as a signal about how BlueAlly will represent Nutanix in customer conversations.

The Partner Manager owns economics and co-sell. TC Capozza is Assistant VP of Cloud Partnerships and Marketplace Programs. He is going to push on Elevate program economics, NC2 on AWS Marketplace, joint pipeline mechanics, MDF, and rebates. His questions land on Banes's deputy more than on Banes himself, but they shape whether the partnership scales.

The SA is the technical credibility anchor. That is my job tomorrow. My job is the narrowest and most specific in the room. I am not there to set vision or carry the narrative or push on economics. I am there to be the answer when someone asks a specific technical question about a specific customer scenario, a specific product capability, or a specific competitive situation. I am also there to demonstrate that BlueAlly's bench has the depth to justify the partnership investment Banes is being asked to make.

The SA's job is not to talk a lot. It is to be the person in the room who, when asked, gives a specific answer that lands.

What I Bring to the Conversation

Bench credibility shows up in how an SA navigates the talk track. Here is what I have ready to drop tomorrow, and the moments where each lands.

The NetApp strategic alliance and in-place ONTAP-to-AHV conversion. Banes is ex-NetApp. This story is personal to him. The in-place VM conversion in minutes without a full data copy unlocks accounts where NetApp footprints have been blocking Nutanix conversations for years. The integration is still in early access with GA targeted for Q3 2026, so I will be honest about timing. I have specific NetApp-heavy accounts in mind where this changes the pitch.

The two analyst downgrades and Broadcom counter-punch reality. This is where SAs separate themselves. I am not pretending the downgrades did not happen. The real conversation is: how does BlueAlly win when Broadcom is discounting 10 to 20 percent beyond standard? My answer is the 145-site case study. We won that migration, and we are going to win the next one, by being the architects who run the POC and document the operational story Broadcom cannot match. The discount is one variable. Operational consistency over five years is the bigger variable. We sell the bigger one.

Service Provider Central. TC and I are mapping our ex-VCSP relationships. There is pipeline there now that did not exist 30 days ago. SP Central is still pre-GA with H2 2026 timing, so I will not over-promise. The conversation is about which SP relationships are real prospects and what the early access path looks like.

Telekom Deutschland and the NAI 2.6 customer gap. Telekom is the only named Enterprise AI 2.6 reference. That is the honest gap. I will ask Banes directly what other named NAI 2.6 deployments are coming, because BlueAlly customers want references and a single logo is not enough. Asking shows I am operating at the level where the answer matters.

NKP Metal. Bare-metal Kubernetes, early access, GA in H2 2026. For the customers who told us "we want bare metal Kubernetes, not HCI," this changes our pitch. I have two accounts in mind. I will be specific about timing because over-promising hurts both sides.

The Dartmouth case study. Ten engineers, 1,000 VMs, 1,000 containers. We use this story when a customer says they are too small for a VMware migration. It moves the conversation.

Rajiv's "landmark week" framing. The CEO published his own three-pillar recap on LinkedIn last week. Agentic AI stack, NKP Metal bare-metal Kubernetes, and the NetApp/Lenovo external storage expansion. That is the language Rajiv wants partners using with customers. Mirroring it in the meeting demonstrates I read past the press release.

Rajiv's Broadcom comment. He said publicly at .NEXT that Broadcom is "not a company that focuses on innovation, but on maximizing profit." That is sharper public language than usual, and it signals that Nutanix is comfortable going on offense in joint customer calls. Partners can mirror that tone in their own conversations.

That is a 15 to 20 minute talk track. Every item is current, specific, customer-relevant, and demonstrates the bench is doing the work. None of it requires Nutanix SE bandwidth to engage. That is the point.

What the SA Brings (And Does Not Bring)

Do not bring a deck. Internal partner alignment meetings are not customer pitches. Bring questions, two named accounts where current news is moving the pipeline, and specific objections you are hearing in customer conversations.

Do not try to dominate the conversation. The CTO is in the room. The Field CTO is in the room. Their voices lead. The SA's voice lands when invited or when a specific technical claim needs to be confirmed or gently corrected.

Do not go silent. SAs who say nothing in these meetings get read as either disengaged or unprepared. Find your moments. Speak when you have something specific.

Bring a list of things you want from the partner. Tomorrow I want clarity on the NAI 2.6 fine-tuning workflow timeline, the realistic GA date for NKP Metal, what the validated AMD GPU configurations look like for GPT-in-a-Box deployments, and what the Nutanix internal talk track is for handling Broadcom's discount counter-punch in customer FUD. Asking shows I am operating at the level where the answers matter.

What I Am Thinking About Going In

Real talk, because credibility comes from being honest about the parts that take work.

Reuel Murray might be in the room. He is our Principal Solutions Architect and the lead architect on technology-focused vendor relationships. If he attends, the SA dynamic shifts. There are now two senior architects in the room and we need to be coordinated, not redundant. Walking into a meeting with another senior architect and not having pre-coordinated who answers what is a credibility-killer. I will ping him before the meeting.

Billy Hamilton I have not placed. He is on the invite as required. I do not know his role. Asking the meeting organizer who Billy is before walking in is just basic preparation. Pretending I know who someone is when I do not is worse than asking.

The agenda has three topics in an hour. Introductions, Nutanix Enterprise AI, BlueAlly's AI strategy. That is a tight schedule. The SA's instinct to speak too long on a topic will compress someone else's time. Stay tight.

This is a Banes evaluation, not a Banes pitch. I have to stop myself from treating it like a customer meeting. Banes is not the customer. He is the partner running his own evaluation of whether to invest more in BlueAlly. Different conversation, different posture.

The Real Proof Point

Here is what I keep coming back to. BlueAlly published a case study where we migrated 145 sites from VMware to Nutanix in 18 months for a wholesale food distributor. That is a real proof point. That is what bench credibility looks like in production.

When Banes implicitly asks "can your SAs run POCs without my SEs babysitting" the honest answer is the work we have already done. 145 sites in 18 months. That migration did not happen because Nutanix SEs held our hands. It happened because BlueAlly architects ran the work. That is what the bench produces, and that is what the partnership is buying. Even when Broadcom is discounting 20 percent below the line.

What Belongs on the Whiteboard

1. "The meeting is evaluating the bench, not the company." When you walk into a partner alignment meeting, the partner already knows your tier, your revenue, and your pipeline. What they are evaluating is whether your architects can actually deliver. Be that answer.

2. "Show up clear-eyed about the headwinds." Read every analyst note and every counter-move from your partner's competitor. SAs who pretend everything is rosy when downgrades are public look unprepared. SAs who name the difficult moment and have a plan to win despite it earn trust.

3. "Know your role. The SA's job is the narrowest." The CTO sets vision. The Field CTO carries the narrative. The Partner Manager owns economics. The SA is the technical credibility anchor. Stay in your lane and own it cold.

4. "Bring questions, not decks." Internal partner alignment meetings are not customer pitches. Walk in with two named accounts where news is moving pipeline, three specific objections you are hearing, and a list of things you want from the partner. The deck stays at your desk.

5. "Real proof points beat any pitch." When the partner asks why they should invest more in your bench, point to the work the bench has already shipped. BlueAlly's 145-site VMware-to-Nutanix migration is the answer. Yours will be different. Know it cold.

Every customer meeting next week is going to touch AI somewhere. Maybe it's a formal "what's our LLM strategy" conversation, maybe it's a sidebar after the storage refresh discussion, maybe it's a CIO mentioning the line item that just showed up on the cloud bill. The conversation is coming whether you are ready or not. The question is whether you walk in credible or whether you walk in pitching.

This post is about how to walk in credible.

Start With the Bill, Not the Pitch

The biggest mistake BASAs are making right now is leading with a Nutanix Enterprise AI deck. Wrong move. Before you pitch anything, you need five answers from the customer. Get them in this order:

1. What is your current monthly spend on cloud LLM APIs? Anthropic, OpenAI, Bedrock, Azure OpenAI. Pull the actual bill if they will let you see it. If they cannot or will not tell you, the conversation is not ready.

2. What does the projected 12 month volume look like? Are queries growing 10% a month or 50%? Linear or exponential?

3. Are most queries hitting public knowledge or your own data? This separates "we use ChatGPT for general help" from "we are running RAG on our contracts."

4. What is the data classification on the inputs going to those models? Public, internal, confidential, regulated. Each one changes the math.

5. Who owns that bill today, IT or a line of business? This is the sneaky important one. If the LOB owns it, IT cannot move the conversation. You need to know who you are really selling to.

You will know within fifteen minutes whether you have a real opportunity or whether the customer is six months away from being ready. Either answer is fine. The wrong answer is leaving the meeting without knowing.

The Math, On a Napkin

Once you have the bill, you can do the math the customer probably has not done yet. Do not memorize specific per-token prices. They change every quarter and you will look stale. Memorize the shape of the curve.

Here is the napkin version, and you can do this on a whiteboard in front of a customer in two minutes:

A typical RAG query against a hosted frontier model is a couple of thousand input tokens and a few hundred output tokens. At sustained volumes around 200,000 queries per day, you are looking at five figures per month in cloud LLM spend, easily. A two L40S node from a Nutanix validated OEM, running Nutanix Enterprise AI on top of NKP, runs the same workload for a fraction of that on a fixed cost basis your CFO can forecast.

You do not need to be an accountant in front of the customer. You need to make their accountant want to do their own version of the math. Once they put a real number on the right side of the whiteboard, your job gets easier.

The other number that matters: the published academic break-even is roughly $20,000 per month in API spend, or 3 to 5 million tokens per day sustained. Below that line, cloud is cheaper. Above that line, on-prem starts winning on TCO inside 18 to 24 months. Memorize that threshold. It is your qualifier.

Qualify Hard. Do Not Pitch Every Customer.

Real talk: half of the customers you are meeting with should not run LLMs on Nutanix. Telling them that is the most credible thing you can do.

Here is when on-prem actually wins:

• Daily query volume is high and predictable. 200,000+ queries per day. Claims processing, document extraction, internal RAG, code assist running all day every day.
• The workload repeats against the same data. RAG over a stable corpus is the textbook win.
• The data has a sensitivity classification cloud cannot touch. Legal, HR, M&A, clinical, regulated industries.
• The customer needs a deterministic cost basis. CFOs hate variable cloud bills. Fixed cost is a feature.

Here is when you should slow the customer down or steer them elsewhere:

• Volume is under 50,000 queries per day. The cloud is genuinely cheaper. Do not pitch Nutanix here.
• The customer needs frontier reasoning capability. GPT-5 class, Claude Opus 4.6 class. Open weight models have closed the gap to about 5 to 10% on most benchmarks, but the hardest reasoning tasks still go to the frontier APIs. Be honest about that.
• The workload is spiky and unpredictable. You will over-provision GPUs and waste money.
• The customer has zero platform team. Nutanix Enterprise AI helps a lot. It does not eliminate the need for someone who knows Kubernetes and basic MLOps.

A BASA who walks into a meeting and says "you should not do this on Nutanix yet" earns more trust than one who pitches at every customer. That trust is what gets you back in the room when the customer's volume crosses the threshold.

And here is the uncomfortable truth most vendors will not say out loud. Gartner is forecasting a 90% drop in LLM inference cost per million tokens by 2030. Cloud API prices are falling fast. That means the on-prem ROI window is narrowing, not widening. Stop selling 5 year payback stories to customers. Build your math on 12 to 24 month payback windows or do not bring it up. If a customer's CFO finds the Gartner forecast in their own analysis after your meeting, and you did not address it, your credibility is gone. Address it first.

The Gotchas Nobody is Telling Customers

Three things customers do not hear from vendors that they will hit in production:

GPU lead times are better than 2024 but they are still not Amazon-instant. L40S nodes from Nutanix-validated OEMs typically run four to eight weeks. Plan for it. If a customer needs a model serving a production workload by next month, the cloud is the bridge. On-prem is the destination.

Model lifecycle is a real change management exercise. You will swap models every six to twelve months as new open weight releases land. Nutanix Enterprise AI makes this easier (the AI Gateway hides the model behind a stable OpenAI compatible endpoint), but someone still has to evaluate the new model, validate it on the customer's workload, and roll it forward. That someone is a person, or a fraction of a person. Pretend otherwise and the project fails nine months in.

MLOps does not disappear because you bought a platform. NAI ships with a catalog: vector databases, notebooks, agent frameworks. NKP handles the Kubernetes layer. But model evaluation, fine-tuning decisions, and prompt engineering are still human work. If the customer has a strong platform team, they will love this. If they have nobody, point them at Lenovo or Cisco delivered GPT-in-a-Box turnkey configurations and set realistic expectations.

The Conversation Arc: Three Meetings, Not One

Stop trying to scope a Nutanix Enterprise AI POC in the first meeting. The arc that actually works:

Meeting 1: Discovery only. Pull the bill. Map the use cases. Earn the right to come back. Do not bring a deck. Bring questions.

Meeting 2: The math conversation. Bring napkin numbers and a real ROI story. Reference the published academic break-even. Talk through Nutanix Enterprise AI 2.6 and the AI Gateway. Show the customer how a hybrid setup (frontier queries to cloud, high volume to on-prem) gives them the best of both. This is also where you tell them what you would NOT recommend, so they know you are being straight with them.

Meeting 3: Pilot scoping. One use case. One application. One node configuration. Do not boil the ocean. RAG over a confidential corpus is the cleanest beachhead. Win that pilot, expand from there.

If you try to compress this into one meeting, you will lose the deal. If you spread it across three, you will close more than you would have.

What Belongs on the Whiteboard

1. "What is your current monthly cloud LLM spend?" This is the first question in every meeting. Do not skip it. Without this number, you cannot have a credible cost conversation. Pull the bill if the customer will let you see it.

2. "Below twenty thousand a month, the cloud is cheaper. Tell the customer." The published break-even is around $20K per month or 3 to 5 million tokens per day sustained. Below that, do not pitch on-prem. Above it, the math flips fast. Customers respect BASAs who tell them when NOT to buy.

3. "Two L40S in a Nutanix-validated node is the sweet spot." You do not need an H100 farm to start. L40S delivers most of the inference value at a fraction of the cost, runs on validated configurations from Dell, HPE, Lenovo, and Cisco, and Nutanix Enterprise AI 2.6 lights the whole thing up. Start there.

4. "Cloud prices are dropping fast. Use 12 to 24 month payback windows." Gartner is forecasting a 90% drop in LLM inference cost per million tokens by 2030. Address this before the customer's analyst does. The on-prem ROI window is narrowing. Honesty about the trend is what makes the rest of your math credible.

5. "Start with RAG on confidential data. Win the beachhead, expand from there." The cleanest first deployment is internal RAG over a stable corpus where data sovereignty matters more than cost. Legal, HR, M&A, clinical research. Win that pilot. Use it to earn the next conversation about scaling.

Three things landed from .NEXT 2026 that you need for customer conversations this week. Here's the context on each one and how to use it in your next meeting.

NetApp Alliance: How to Use This in a Storage Conversation

NetApp and Nutanix announced a strategic collaboration to integrate ONTAP with the Nutanix Cloud Platform on AHV. Here's why you need to know this and how to bring it up.

The objection you're going to hear: "We love the Nutanix operational model, but we have a massive NetApp footprint and we're not ripping that out." You've probably heard some version of this already. It's a completely reasonable position. Enterprise storage investments are measured in petabytes and millions of dollars.

How to respond now: The ONTAP integration means you can reframe the conversation. Instead of "Nutanix OR your existing storage," tell them it's "Nutanix AND your existing storage." Same data, same protocols, unified operations. That changes the sizing conversation entirely. You're no longer asking them to forklift their storage investment. You're offering a compute and management layer that works with what they already have.

The caveat to be upfront about: This is a conference announcement, not a GA product yet. Don't oversell it. The right framing is: "This is on the roadmap and the strategic intent is clear. Let's start planning what an integrated architecture looks like in your environment so you're ready when it lands." The timing alongside Dell PowerStore and Everpure announcements tells us Nutanix is serious about being storage-agnostic at the platform level.

What to tell your customer: "Get on the early access list. Don't wait for GA to start planning the architecture. Let's scope what this looks like for you now."

Dartmouth Migration: Your New Proof Point for Hesitant Prospects

SiliconANGLE published a piece you need in your back pocket. Dartmouth College migrated approximately 1,000 VMs off VMware to Nutanix. Their entire IT infrastructure team? Eleven people. They came out the other side running Kubernetes workloads and positioned for AI.

When to use this story: The next time a prospect says "we're too small to do this" or "we don't have the team for a migration." Eleven people. Not a massive migration factory. Not a two year timeline. A small, skilled team that just did it.

Combined with the Western Union migration from the first post (enterprise scale, global financial infrastructure), you now have proof points at both ends of the spectrum. Big enterprise and mid-sized institution, same outcome.

The honest follow-up you need to give: SDxCentral reported this week that some VMware migrations are taking longer than expected, especially for organizations deeply embedded in the VMware ecosystem. Don't hide this. Your credibility depends on it. The right question to ask the customer is: "How deep does your VMware integration go? Are you running NSX, vSAN, Horizon?" If yes, plan for complexity and longer timelines. If they're running straightforward VM workloads, Dartmouth is their blueprint.

What to tell your customer: "The complexity of this migration isn't about your team size. It's about how deep the VMware integration goes. Let's assess that together and build a realistic timeline."

AI Gateway: A Door Opener Most Customers Haven't Thought About

Nutanix introduced an AI Gateway that provides visibility into agent behavior, enforces cost and security governance, and manages AI agent sprawl. Here's why this is a conversation starter you should be using.

The question to ask in every meeting: "Who owns AI governance in your organization?" If the answer is "nobody" or "we haven't thought about it," you just opened a door. Every enterprise is either running AI experiments or planning to. The excitement is real. But so is the fear of what happens when 50 AI agents are running across the infrastructure and nobody knows what they're doing, what they're costing, or what data they're accessing.

How to position this: The AI Gateway isn't about making AI faster or smarter. It's about making AI manageable. Visibility, cost controls, security guardrails. Frame it as the boring infrastructure work that prevents the expensive AI disaster. Customers understand that framing because it's the same argument we've made about every other enterprise technology: the governance layer is what makes production possible.

Palo Alto Networks was named Nutanix's 2026 Global Security Partner of the Year and is launching Prisma AIRS integration for model security scanning and autonomous red teaming. When you're in the AI governance conversation, bring up the security angle too. Security plus governance plus the platform layer is a real enterprise AI stack, not a marketing slide.

What to tell your customer: "Start the AI governance conversation now, before your experiments become production workloads that nobody can audit. The framework exists. Let's talk about what it looks like in your environment."

The Ecosystem Play: How to Handle the "Is Nutanix Big Enough?" Question

You're going to get asked: "Is Nutanix a real enterprise platform or still an HCI appliance?" Here's your answer this week.

Point to the partner announcements. NetApp for storage. Cisco for networking and compute. Dell for hardware flexibility. Palo Alto for security. Dynatrace for observability. That's not a niche product. That's an ecosystem. When a customer can run all of that under a unified management layer, the "why would I switch?" question gets very hard for competitors to answer.

How to use this in competitive situations: If you're up against a competitor trying to unseat Nutanix in an account, the ecosystem is your moat. The "rip and replace" math just got a lot worse for anyone trying to displace a Nutanix environment that's integrated with NetApp storage and Cisco networking. And for new accounts, the ecosystem story is what turns a one-time infrastructure deal into a long-term relationship.

What Belongs on the Whiteboard

1. "How deep does your VMware go?" That's the first question in any migration conversation. NSX, vSAN, Horizon mean more complexity. Straightforward VMs? Point to Dartmouth: 11 people, 1,000 VMs, done. Tailor the timeline to the depth, not the team size.

2. "Who owns AI governance in your org?" If nobody does, you just found your opening. AI Gateway gives you the framework: visibility, cost controls, security guardrails. Get ahead of this before production AI forces the conversation.

3. "Nutanix AND your storage, not OR." For NetApp shops evaluating Nutanix, the alliance means they don't have to choose. Get them on the early access list and start scoping the architecture now.

4. When they ask "is Nutanix big enough?" this week is your answer. NetApp, Cisco, Dell, Palo Alto, Dynatrace. That's not an HCI appliance. That's a platform with an ecosystem. Use it to build confidence that this isn't going anywhere.

If you weren't paying attention to .NEXT 2026 this week, it's time to catch up. Because between the KubeVirt announcement, NKP Metal, and some very public VMware defections landing in the news simultaneously, today feels like one of those days where the infrastructure landscape actually shifted a little.

Let me break down what matters and what I'm going to be talking about in customer meetings for the next several weeks.

KubeVirt: The Edge Play Nobody Saw Coming (But Makes Total Sense)

The Register had an exclusive this morning: Nutanix is planning to support KubeVirt to let customers run virtual machines on top of Kubernetes. specifically targeted at edge deployments. ARM support is on the roadmap too, because, as The Register correctly notes, AI is going to run on everything.

Here's why this matters more than it might look on the surface. The edge is a mess right now. You've got customers who need VMs for legacy workloads and containers for modern ones, and they're being forced to choose platforms or run two separate stacks. KubeVirt changes that calculus. If Nutanix can deliver a unified control plane where you're running Kubernetes natively and VMs as first-class citizens on the same cluster, that's a legitimately compelling edge story.

What I'm telling BlueAlly customers who are evaluating edge deployments: this is worth putting a pause on any commitments right now. If you're designing a 3-node edge cluster for a retail location or a manufacturing floor, the KubeVirt support changes what Nutanix can offer you in 12-18 months. The ARM angle also can't be ignored. NVIDIA's Vera Rubin NVL72 rack was on display at GTC just weeks ago, and inference at the edge is coming fast. Nutanix is clearly trying to be the OS layer for all of it.

NKP Metal: Bare-Metal Kubernetes Is Now a Nutanix Story

This one is huge for a specific segment of the market that I run into more than you'd think: customers who want Nutanix's operational model. the simplicity, the unified management, the lifecycle automation. but don't want to run AOS underneath it. They want bare metal. Maybe it's a performance requirement, maybe it's cost sensitivity at scale, maybe it's a regulatory thing.

Nutanix Kubernetes Platform Metal extends NKP to support Kubernetes deployments directly on bare-metal infrastructure. Same operating model. Same governance. No hypervisor tax.

This is a direct shot at the Pure Storage, DDN, and WEKA-type plays that have been positioning themselves as the "Kubernetes-native" infrastructure layer. WEKA was just in the news today claiming Nvidia CMX support plays to their strengths. and they're not wrong, they've got a strong story for GPU-dense AI infrastructure. But NKP Metal means Nutanix can now walk into that same conversation and say: we can give you Kubernetes on bare metal and it still looks like Nutanix to your ops team. That's a different value prop than pure performance, but for enterprise shops with existing Nutanix investments and teams? It's compelling.

The Agentic AI Platform Positioning

The overarching theme from .NEXT 2026 is Nutanix positioning the entire platform around "Agentic AI". optimizing, governing, and accelerating agentic AI workloads across a growing ecosystem of infrastructure, cloud, and software partners.

I'll be honest: I have some healthy skepticism when any vendor slaps "agentic AI" on everything and calls it a platform. But in Nutanix's case, the underlying substance is real. The combination of NKP Metal for bare-metal K8s, KubeVirt for VM/container convergence at the edge, and the existing cloud-native tooling gives them a genuine story about running AI inference workloads at every tier. edge, on-prem, cloud. That's not just marketing.

What enterprises actually need for agentic AI isn't just raw compute. it's the data infrastructure and governance layer underneath it. AWS announced S3 Files this week, making S3 buckets accessible as file systems, and it's a reminder that the hyperscalers are constantly closing the gap on unified data access. Nutanix needs to be the on-prem answer to that story, and .NEXT 2026 suggests they're thinking about it the right way.

Western Union Walks Away From VMware. And They Won't Be the Last

Meanwhile, in "the VMware migration wave is very much still real" news: Western Union has commenced a migration from VMware to Nutanix. South Korea's largest theme park is apparently on the same journey.

I don't want to beat the dead horse here. we've been talking about the Broadcom/VMware pricing fallout for over a year now. But stories like Western Union landing in The Register matter because they validate the conversation I'm having in virtually every enterprise account at BlueAlly right now. These aren't small shops looking to cut costs. Western Union is a global financial infrastructure company. When they decide the VMware licensing math no longer works, that's a signal.

What's interesting about the timing is that these migration announcements are landing the same week Nutanix is unveiling its most ambitious platform expansion in years. It's a one-two punch: the pull of a maturing, genuinely capable platform and the push of a vendor relationship that a lot of organizations are ready to exit.

The Bigger Picture

If I zoom out from today's headlines, here's what I see: Nutanix is methodically closing every gap that competitors have used to poke holes in their story. "You can't run Kubernetes on bare metal". here's NKP Metal. "You can't unify VMs and containers at the edge". here's KubeVirt. "Your AI story isn't mature". here's a full agentic AI platform announcement at .NEXT.

None of these things are magic. The KubeVirt support is on the roadmap, not GA today. NKP Metal will take time to mature. The agentic AI positioning will need to be proven in production deployments before it shows up in customer references. But the direction is clear, and in my work at BlueAlly, direction matters as much as today's feature set when you're building a 3-5 year infrastructure strategy.

The customers who locked in Nutanix commitments during the post-Broadcom chaos are going to look smart in 18 months. And the ones still sitting on the fence waiting for Nutanix to "prove" its platform story just got a very full week of evidence to work with.

More from .NEXT as the announcements continue to roll out. there's a lot to unpack and I'll have more to say about the partner ecosystem pieces tomorrow.