This massively increases risk too, right? Pinning all your revenue on a much smaller customer base means losing one or two of them has a huge impact!

Also pretty much guarantees flight of your former customer base to close compliments - who then have improved revenue to bring their products closer in alignment to yours. You'd better have a hell of a moat to follow this strategy.

Tiny nitpick: complement means a good in an adjacent part of the supply chain (e.g. batteries and EVs are complements), you might have meant 'competitors'.

From the point of view of a customer, a complementary service would be a product that fits the same niche and is reasonably substitutable.

English is a living language so it's impossible to say definitively that you're "wrong" but I don't think anyone else is able to understand what you're trying to communicate through your use of this word.

According to the Cambridge dictionary, there are a couple of variants depending on your dialect:

- "to make something else seem better or more attractive when combining with it" - e.g. "Strawberries and cream complement each other perfectly."

- "to help make something or someone more complete or effective" - e.g. "She used photographs to complement the text of the news story. "

The only way I can make any sense of your words is to assume you're trying to argue that two competing but interchangeable services "complement" each other when used simultaneously because that might allow a customer to mitigate risk in a similar fashion to a multi-cloud setup. That's a pretty big jump to make from what you've actually said though and would really need further explanation for anyone to understand.

FYI in modern microeconomics, 'substitute' and 'complement' are precisely defined terms that have opposite meanings.

This is interesting. In university, I was taught that a substitute product would be an alternative choice (eg. Nathan's hotdogs vs Ball park franks) and complimentary products were ones that typically "moved" together (eg. Nathan's hotdogs and ketchup). So not really opposite unless you are talking about price vs demand.

No, they are in fact opposed definitions if you dig in a bit.

Substitute goods have negatively correlated demand, while complementary ones have demand that positively correlate. A simple criterion, at least in theory. You will buy more ketchup when you buy more franks, but you will buy fewer hot dogs.

And it's also close enough to the MBA definition of complement (of "commoditize your complement" fame).

> Also pretty much guarantees flight of your former customer base to close compliments - who then have improved revenue to bring their products closer in alignment to yours.

Classic VM hosting is a dead end anyway. Customers are shifting to the hyperscalers wherever they can, if only to reduce the headcount of their IT departments due to acounting wizardry making it worth it for the stonk markets even if the cloud costs more in the end.

And then, there aren't that many alternatives to VMware, and none of them (bar OpenStack) as comprehensive.

As long as Broadcom manages to squeeze enough out of the large customers who are vendor-locked too hard over the next 3-5 years, it'll be worth the money for them, and chances are the gamble pays off, with a small trail of extremely large customers paying for a decade until they can get their internal chaos sorted out to migrate off.

Broadcom makes $35B a year in revenue.

VMWare made at most $13.4B.

Even with severe churn, VMWare would make around $12.8-13B.

VMWare is just a BU now, not a company, and the economics of managing "just another product line" is different from a company with a flagship product

As I've mentioned before on HN, the math is different and it makes sense to up prices and only concentrate on F1000s at that size.

> Pinning all your revenue on a much smaller customer base means losing one or two of them has a huge impact

Large customers are sticky. You can't migrate your hypervisor or cloud provider overnight. These are multi-year projects.

Also, it's better to target a smaller base of high paying customers instead of a large base of low paying customers because every sales motion and support ticket is an opportunity cost and a financial cost.

Don’t attribute this stuff to strategy. The people running “the BU” are just counting on friction to hold on to revenue. They aren’t able to even meet the commitments they make and are disrespectful to customers in other ways.

I think they overestimated their market power. I have 2 years left on my main ELA. I have budget to spend 5x to exit. My spend will go from 8 figures to under $1M.

Personally, I resisted the visceral “fuck you” emotional response. But it’s one of the rare times when the quick take is correct. The reality is VMWare is dead and is too risky to keep in place. The smart move is not to play.

You can fuck up A LOT and retain customers if you're honest and truly working in the customer's favor. The absolute second a customer picks up that you're in it to simply pick their pocket, there will be nothing left but a customer shaped hole in the wall. I have been in situations where we've completely fucked a customer's business causing all kinds of financial damage and customers have stayed because we committed to making things right. I've also been in situations where hubris assumed a sticky customer isn't going to unstick themselves, and of course the customer left.

It's so much better business not to be an asshole. I really don't understand how people can't internalize this.

Exactly this. Any smart CTO will be planning their exit. It’s pretty obvious that what made VMware great is now just a zombie.

> Large customers are sticky. You can't migrate your hypervisor or cloud provider overnight. These are multi-year projects.

Yes, but that scale, everything is an multi-year effort. The contracts likely as well. That doesn't mean, it's not going to happen.

And it's not like all has to happen in one go.

So before you were all in VMware, and that vendor is practically promising to hike up the prices to make you bleed.

What are you gonna do?

I'd rather start early to have a migration path, even if it is just for negotiation purposes. And if it's someone who has the resources to that, it's large customers.

When you have a sticky existing customer who's about to churn, you end up discounting below the price of the migration, then slowly rise the cost again, then do the same thing again.

All vendors do this - you can't escape it. This is why companies began leaving for the Cloud - sure it's upfront more expensive, but the negotiations are not as protracted.

> All vendors do this - you can't escape it. This is why companies began leaving for the Cloud - sure it's upfront more expensive, but the negotiations are not as protracted.

I don't get your point here: if you mean the public cloud, it's the synonym of vendor lock-in now.

Everything is vendor lock-in. The difference is negotiations are not as protracted, the muscle to develop a multi-cloud strategy is fairly well built in our industry, AND cloud purchasing offers multiple different styles of billing (eg. subscription driven, multiyear, credit driven, etc) whereas with on-prem you have only a single type of billing model.

Sorry just to clarify, leave for the cloud or leave cloud? Looks like in this case it is indeed leave for the cloud (leave VMs) but I also heard a lot about leaving cloud for self hosting.

Leaving/left for the cloud. There is some scaling down of cloud resources for self hosting, but as a whole all CSPs have seen growth in users across all buckets.

On-prem doesn't give you flexibility, and leaves you open to getting arm twisted by vendors. Similar stuff happens in cloud ofc, but it's easier to implement a multi-cloud strategy than a multi-hypervisor strategy.

Gotcha, thanks for the insight.

Also, where are ESXI customers at the enterprise level going to go? Do we really think they're going to magically switch over to HyperV (and get Win admins) or Proxmox (and have a lack of IBM-enterprise level support)?

Funny that you say that. In my personal experience, the degree of support is one of the reasons to switch. Like any large vendor, it takes almost more time to convince them that it there is a problem, then it would take to fix it, if you have the technical skills in house. (That was enough for Google to switch to their own switches).

Haven't worked yet with IBM, but if they are of equal level, then I'd rather avoid them.

But since you say IBM, they have IBM Cloud Manager, and through Redhat also an Openstack offer, and with Openshift a K8S offer. Various vendors offer either or both.

There are also companies which operate internal cloud providers for other companies. Various public cloud providers offer you to operate your datacenter with their API in-hose.

Yes, it comes with their hardware, but guess what, in three years chances are half of your hardware is deprecated and has been replaced anyway.

Yes, all that requires effort. Considerable effort. But it is a one-time effort (i.e. fixed costs) compared to a X-fold increase of licensing costs. So, you look at the ROI, consider the risk of having that degree of exposure, and guess what...

To reiterate: It doesn't have to be all in one go, it doesn't mean it has to be all of it. Maybe some of your payload will always stay on vmware, but thinking you can ask the big companies for 20x the license costs, and expect 20x the revenue is rather odd.

You may guess, where I know that from.

Let me say it differently - from what company does the CIO procure from so he doesn't get fired because he can't call up a company to support his corporate ERP running on VMs?

VMware isn't just VMs. They are deep, deep, deep in the stack at some places. There isn't a realistic migration path away from VMware for the customers that Broadcom is interested in keeping. Or, put another way, the time it will take for those customers to move away from VMware (which they won't) will take Broadcom to make their money back on this acquisition a few times over.

> Large customers are sticky. You can't migrate your hypervisor or cloud provider overnight. These are multi-year projects.

I can assure you first hand than not only F1000 but also F100 consider migrating away from companies that introduce sudden price hikes counting on inertia. It's not a question of "if", only "when".

> Even with severe churn, VMWare would make around $12.8-13B.

Down from the $13.4B -- that's not 'severe churn' you're describing there, it implies only a few percent drop in revenue. We've yet to see if that's the likely outcome here, but touchy-feely sentiment suggests that it'll be worse than that.

> Large customers are sticky.

Isn't TFA a precise counter-point to that assumption?

> that's not 'severe churn'

Companies do not churn 100% customers.

Most cases "severe churn" is counted as 80-100% NRR as customers are on multiyear contracts that are much more expensive to break than they are to wait out.

> Isn't TFA a precise counter-point to that assumption?

The customer was already a Nutanix customer, so the hard work was already done.

Basically, this customer was using BOTH Nutanix and VMWare internally (I am VERY surprised how the previous CFO did not get fired for something like that), and because they already had the Nutanix knowhow and licenses, migrated fully to it.

For reference, this article was written by the Register journo who was at Nutanix .NEXT (Nutanix's corporate conference).

> Basically, this customer was using BOTH Nutanix and VMWare internally (I am VERY surprised how the previous CFO did not get fired for something like that), and because they already had the Nutanix knowhow and licenses, migrated fully to it.

Why do you think that? Not putting all of your eggs into a single vendor basket seems like a solid plan - if anything the Broadcom/VMware disaster supports it.

> Why do you think that

Have you ever seen Nutanix's pricebook as well as VMWare's?

Spending 2x on hypervisors is dumb because now you need 2x the headcount on SMEs because you'll need both a Nutanix and VMware SME, as well as 2x the contract negotiations, and the money you are spending on both could have been better spend improving your product or hiring more people to sell your product.

It's a bad use of capital. At the end of the day, Infra is a cost center. It's something used to keep the lights on, but doesn't expand your TAM.

> It's a bad use of capital. At the end of the day, Infra is a cost center. It's something used to keep the lights on, but doesn't expand your TAM.

Indeed it is and while CFOs may not completely understand the technology they do understand risk and if the CTO has flagged "single vendor" as a risk then the CFO will go with that.

I also wouldn't be surprised if pre-Broadcom VMWare APJ AEs gave these guys a sweetheart deal just to make logo/ACV quota in their region. Shenanigans like that were VERY common at VMWare before the acquisition. Their SalesOps was horrid.

And yet, every single large company I worked for had AWS, Azure and Google Cloud, without exception. Many also have extensive on-prem resources.

Also, the contract negotiations you mentioned work much better if the competitor is already well present in your company.

> every single large company I worked for had AWS, Azure and Google Cloud,

Multi-cloud is different from on-prem related stuff like multi-hypervisors, because there are multiple billing methods, the muscle to migrate is much better built in the industry, and your cloud costs can be placed within R&D (which traditionally gets way more leeway due to tax benefits) whereas any IT Infra spend will inevitably fall under the Finance&IT budget.

Of course it's very different, but it's the same in how buying the same functionality from a competitor is not a reason to get you fired unless there are other specific conditions.

> buying the same functionality from a competitor is not a reason to get you fired unless there are other specific conditions

There are different expectations depending on the kind of bucket you are spending from.

The amount of capital you have to spend in the R&D bucket is much larger than the Finance&IT bucket and who you report to (CTO vs CFO) is different.

> (I am VERY surprised how the previous CFO did not get fired for something like that)

It always make me smile how fast people on the Internet would fire CFOs without knowing the context of the situation. (Not to mention that this case proves the guy was right.)

> I am VERY surprised how the previous CFO did not get fired for something like that

Could be they acquired another company that ran the other one. We see this with our customers all the time. They acquire a competitor, slap their name all over it, but we still have to treat them like effectively two separate customers for ages.

Good point! UBS-Credit Suisse is dealing with that right now, and heads are about to roll over that.

Forgive me, but it feels like a lot of contradictions, assumptions (not in evidence), and hand-waving with this and your other comments in this thread.

You've said 'severe churn is 80-100% NRR' in one message, but in another that 'severe churn' would involve only a difference from 13.4b down to 12.8 - 13.0b (vmware / broadcom revenue).

You've said that it's foolish to have two hypervisors in play in an org, because then you're doubling up on SMEs for hypervisor infrastructure.

Computershare's revenue last FY was 3.2b -- and as per TFA they were running 24,000 VMware VMs (one can only speculate on the Nutanix VM count) -- so it seems reasonable that they'd have sufficient baseload of SMEs to split across two technologies without sending the company under. Given their YoY revenue increase from the previous year, it evidently wasn't a constraint.

You've said that multi-cloud is different (more acceptable / forgivable) to multi-hypervisor for three reasons:

   a) different billing mechanisms
   b) migration is baked into cloud services
   c) cloud budgets come under R&D rather than finance / IT

I don't know if (a) and (c) are the same thing worded differently, but I'd vigorously dispute that on-prem has only one billing method, that migration from one SaaS provider to another is 'better built into the industry' (all the players make it monumentally difficult to migrate off their platform), and with (c) I'd once again murmur 'facts not in evidence', especially in the context of TFA (Computershare)

You have not addressed that multi-cloud has the same two problems you accuse multi-hypervisor of suffering - a requirement for multiple sets of differently skilled SMEs, and 2x contract negotiations.

Large customers also can have resources to simultaneously negotiate you on price while starting 3 initiatives to "remove vmware ASAP".

F1000 can have sticker price shocks and sensitivity too, especially if you happen to raise your prices soon after other events that might have made them look into cost savings...

I was an infrastructure architect at a F50, and I'm sure they'd have been quite sensitive to the changes. Like, once we got a confirmation of new pricing, the day after we're kicking off a Discovery process to find and cost out some alternatives.

They were, and I'm guessing still are, also very resistant to SaaS / IaaS plays, for a lot of (arguably) good reasons, and I'm not sure what they'd go with as an alternative. OpenShift? Raw dog some DIY docker clusters?

Probably various integrators? Oxide misses some critical features in networking right now but I suspect it would a very tempting option for many once they added support for the typical unholy mess that are vlans connected to vmware networks.

I don't know where exactly in spending bracket my current $DAYJOB fits, but I did hear both about negotiating prices with Broadcom and grumbles of looking for replacement including accelerating movement to AWS where possible.

> Large customers are sticky. You can't migrate your hypervisor or cloud provider overnight. These are multi-year projects.

In VM's case it is not that long really. It really depends how much priority you put into it.

> In VM's case it is not that long really

It is. You have teams critical apps running and you do not want to cause downtime for customers (internal or external).

I've seen this saga happen dozens of times and it is a very delicate and intricate process that requires a lot of planning.

A failed migration can become news, for example - https://www.bloomberg.com/news/articles/2023-11-15/ubs-says-...

It only is because companies usually don't staff to be ready for large migrations. They should.

They'll be sitting around twiddling their thumbs most of the time. Seems like something to be contracted out when necessary.

Why should I hire an additional 50 Infra Engs who are underutilized when I can use that to hire 50 additional SWEs or Salespeople to expand my TAM?

Because migrations should happen all the time. If things are done right requires the same effort to patch and have downtime of all your systems than it is to migrate from one hypervisor to another. I've migrated tons of systems, sometimes p2v and v2v makes it simpler/faster but more often than not it is more useful and as fast to reinstall system then redeploy app (you can often use that time to migrate to major OS release) and it is dead easy to do a rollback if you encounter issues in migration, stop the new one, restart the old one.

A company that is not staffed to migrate all its servers in less than 3 months is not staffed to have the most minimum level of IT security.

The reality is that in many organizations, nobody wants to update or reboot anything unless absolutely necessary. I once took a contract job at a company that wanted me to migrate and upgrade some Linux VMs running on a 5+ year old distro. This was a decent sized, established company, not a startup (1000's of employees, 100 million+ revenue)

The existing VMs had not been updated or rebooted in years. Several had 1200+ day uptimes! These systems were accessible from the Internet and regularly used by customers for an important, highly specialized product. The only reason they were even bothering was they had failed some security scan and nobody there could even understand how to use SSH.

The minimum level of IT security is apparently barely anything.

> Because migrations should happen all the time

Hypervisor migrations? Complete DC migrations? If you are doing that every single quarter you are doing something VERY wrong (or you should migrate to the cloud).

You don't have to do these migrations, but you should be able to do them because basic security maintenance of systems require regular patching, reboots, downtimes, etc. Setting up a new VM and deploy the apps/services takes minutes, not days.

The infra has to be setup to support downtime of systems (not all at once obviously).

> Large customers are sticky.

Many vendors confuse inertia for stickiness. Large customers are slow. But like super tanker they are very hard to course correct once you send them one. Course you don’t like.

The revenue drop is more significant. Broadcom bought CA in 2018, CA had roughly $4b in annual revenue. Bought Symantec in 2019. Which had roughly $2B in annual revenue. The consumer version, Norton Locklife, now also has roughly $4b in revenue without the slash and burn experienced at Symantec. In roughly 5 years, the combined franchise has managed roughly $1.5b in annual revenue growth. VMware in the same period managed around $6b in growth. Roughly 50% growth in the top line, without the slash and burn approach. With the trust of the ecosystem. So, yes, in the short run, they will have some growth in revenue, but it slows down, quickly. Last year their infrastructure software revenue grew by 3% - the average inflation rate was 4.1%, they shrunk in the previous 12 months. So I'm not sure how this will all work out but given that they have a target of $94b to return to shareholders (dilution + the cash outlay). In Broadcom's Q12024 (which was historically VMware's Q4), VMware's contribution to Broadcom was down by $1B YOY. I am waiting for earnings reporting next month, but I'm not sure they will hit their target. Customers left. Their reputation preceded them.

In this case, if you increase your price 10X and they don't leave you, there must be some deep technical reason why they cannot move even if they want to.

You no longer have customers, you have hostages.

I remember my father doing this one. He was an SME support company. With two clients. He cranked the price up 4x and said ”we’ve got them by the balls”. They both opened a tender process and shot him and he was bankrupt 6 months later.

4x and in one go is too much for many to bear. If you do it at 25%-50% a year or so you can boil the frog with much more reliability.

Everyone and everything is replaceable. The question is if it’s worth it.

A shakedown price increase means that your vendor isn't your partner. Because major vendors are business partners.

It's like blackmail. If you pay up, how do you know another one isn't coming the next year?

But... Oracle stays in business, so what do I know.

Exactly. I do not know how what is happening with VMware didn't happen to Oracle years ago, apart from them being a legal firm with a tech product attached.

Because Oracle does have an actual product and team that is actually skilled and delivers what they promise. Sure, there are competitors that might do similar, but they also cost a pretty penny.

> They both opened a tender process and shot him

What does that mean?

They approached multiple vendors for quotations who undercut him vastly. So he didn’t get the business.

I read as https://dictionary.cambridge.org/dictionary/english/get-be-s...

I interpreted it as "fired him"

They went shopping for a new vendor by sending out a request for proposal to multiple companies.

> They both opened a tender process and shot him

> What does that mean?

Self defense. Gun rights.

I'm guessing those examples are legion. Always treat your costumers and partners with respect. Another famous example is BeOS thinking they have Apple by the balls to be the basis for OSX. Demanded an outrageous price, Apple just used a BSD.

That is not a good comparison. NeXT was far more than "a BSD". It was a whole bunch of frameworks, development tools, applications, and a hardened team that had already ported to several other architectures. BeOS was still mostly vision at that point.

wasn't Next actually developed by the former CEO of Apple as well?

Yes, Steve Jobs was CEO of Next when they got bought

that seems like it had way more to do with the decision that anything technical. What the team at Next built is quite impressive as well, but the goal was to get Steve Jobs back.

Yes, but I guess you get to know the customers better and make sure the product fits them very well, or atleast better than the competitors.

Another more cynical thought is that this is a way of increasing revenue and profit quickly so the share price rises quickly. You get your options and bonus and leave before the rest of the customers leave.

Your second cynical answer is the correct one. It's always about the shareholder and never about the customer at Broadcom. Remember Computer Associates (CA)? Yeah, they're now Broadcom. That pattern of acquire - consolidate - turn the screws on the customer DNA / culture is very strong in Broadcom.

I used to work for an enterprise software vendor that had LOTS of customers, and losing a single customer was painful but didn’t really affect that much. I’m now at a much smaller company, and I feel like it’s noticeable in the stock price when we gain or lose a customer. Kinda scary, but exciting at the same time. No idea if it’s just my brain inventing patterns.

Modern business seems to not care about resilience or having a plan B at all.

Acquiring new customers would become quite difficult, though. If you increase prices to a point where everyone who can leave will leave, who in their right mind would choose your product and risk another price hike in the future?

All prospects are not the same.

You have 5 types of accounts:

- Strategic: F500s or very well known startups

- Enterprise: F1000s

- Mid-Market: Companies below $1B a year in revenue

- Federal: Federal Government

- Channel/Reseller: For companies that are too small, you have a MSP sell for you because you can't be bothered to sell to them.

Strategics, Enterprises, and Fed will always get good deals and will always have protracted conversations, as these are accounts spending 7-9 figures.

Mid-market and below will always have a bad time because they ain't spending enough.

We're businesses, not your friends. We'll try to help, but at the end of the day, we want to get paid as well.

Makes sense. Does that mean there is a bigger market for mid-market and below now? Maybe it is too annoying for the big companies to care, but could be lucrative for smaller ones.

> Does that mean there is a bigger market for mid-market and below now

Nope. Tech forward mid-markets (eg. every single tech startup) are all cloud first, and the legacy mid-markets will go thru MSPs and Resellers to buy these products.

No one likes IT spend. It's a cost center.

Gotcha, thank you. Yeah, no one likes a cost center...and IT is usually the first to layoff.

Problem is, you are opening up the market for open source competitors by not servicing the mid market.

Customer service is becoming truly extinct.

> Problem is, you are opening up the market for open source competitors by not servicing the mid market.

You aren't. They go to resellers or other large vendors (eg. this example of going from VMWare to Nutanix)

Open Source competitors (NOT open core companies) do not provide the SLAs on support needed if SHTF.

Sounds like a plan to invest into competition

Similar to pmx's point, an implicit assumption in this (which is unlikely to be true in many scenario's) is that the there is a negligible spread in the distribution of revenue per customer. Imagine if you have one whale customer and a bunch of small fish, if you lose the whale in the change you won't maintain the same revenue.

Some scenario's this assumption is reasonable though (say Netflix subscribers).

A lot less sales and cuatomer support people

except that customers that pay 10x have different expectations. They will have a dedicated account manager, SLA's with very short response times, etc

^^^ This

For a while.

This is a massive invite for competing businesses to offer the same service 20% cheaper, and still make lots of money.

But that's the way to become a niche product and not one of the standard supplier of VM solutions.

They will lose reputation and reference customers.

Broadcom absolutely does not care about this.

But we're talking subscriptions so the drop-off is recurring. You will have customers who will burden the cost while they migrate and ditch you as soon as they're done.

If you start celebrating your new pricing structure with just a 90% customer loss in Y1, you're in for a nasty shock in Ys 2-5.

That's before you factor in the shift of mindshare to alternatives. Pointing this thing at ultra-large customers means everyone else is using and training on something else.