This is basically a polite way to fire a customer.
You don't want to sell to everyone - you need to make sure the cost of retention+support is significantly less than what they are paying.
If a customer is too troublesome or spending too little (for a company like VMWare, a $8m contract is on the smaller end) this is the politest way to fire them.
It sounds like this customer is a legal firm, which makes sense - they tend to have a much smaller IT footprint.
Every dollar isn't the same.
Edit: to elaborate - it's all about margins. Is that $8m worth it if in aggregate you spending an equal amount of time on that account as you are on a $20-30m account? It ain't
It is the most shitty (and stupid) way possible to act though. Old price x 12 is not in good faith. Contracts are based on good faith. Give them a rationale, give them a deadline 24 months hence and a fair price increase. Dare to say no to individual customers. Any customer should now start planning for the end of their Broadcom VMWare-contract. If they do not act in good faith to a random customer, why would they to you? Large company contract management is based on trust. This might cost Broadcom a lot more than a few individual customers that they were planning on 'letting go'.
I work at a company that got sold by another company with an indefinite LTA (long term agreement) in place. Former owner did not honor the LTA, but at least was transparant and timely about it. We got two years to sever several key processes. We will end up severing all ties, since the business risk of leaving open ends without the LTA is too big for my risk appetite. Fool me once. And that's our business consequence of not honoring an LTA even when the former owner severed the ties in good faith. If the buy-in is not there for both parties to the contract, then there is no ongoing relationship.
What's not good faith about this? "Your custom won't work for us unless we get 100M"
Chances are this is not actually surprising for this customer. There's probably been a lot of talk between the sides, a lot of expectations that weren't met both ways.
What you're describing at your own company seems to be that same thing, no? Relationship wasn't going well, and it's gotta end somehow.
Now that's not to say this should be normalized, it shouldn't. But there needs to be some way to say "hmm, it doesn't make sense for us, business wise"
> The strategy you described has no upside for the vendor (in this case Broadcom VMWare).
The upside is you don't scare off other customers when they hear reports of a 12x price increase.
Right now, in response to this post, VMWare's biggest customers will be drawing up plans for if they need to walk away at the next contract renewal. And they'll want those plans to be extremely credible - it's worth spending $5 million trialling and integrating a competing system, if it stops your $50m contract turning into a $150m contract.
I mean, VMWare customers were warned about this ever since the deal got announced. This is literally the Broadcom strategy for the last decade at least: Buy vendor with lockin, cut non-core features, raise prices and squeeze the customers who are in too deep.
Broadcom goes into these deals knowing it will hemorrhage a lot of customers and they don't care, they know they can make it up on those who are forced to stay. They don't care about reputation at all.
Sounds like customers were not familiar with how Mr Tan plays ball, he's been doing this forever. He isn't trying to kill the cow, he will famish the cow to the point where the cow will be alive enough for him to milk his required rate of return and not an ounce more. Its a risky strategy, but Tan is loved by Mr Market so he gets the benefit of the doubt.
Companies like Walmart, Centurylink, AWS (via Equinix), and the DoD (via General Dynamics) use VMWare and most likely have contracts in the 9 figure range for a backbone service like VMWare
As I said in my comment - you can safely assume mean contract value is in the mid 8 or low 9 digits at a company the size of VMWare.
If this scares away customers spending less than that, good for them. It doesn't make a dent when a F1000 is spending tens of millions a year on VMware and purchase by committee and will get sweetheart deals by being strategic accounts.
That’s going to get you basically the same number unless you think most of those 400k customers do not have contracts for some reason? My revenue-based approach would undercount in-year bookings but your approach would undercount NRR which I suspect is nontrivial for VMWare (but I haven’t checked).
Also just to be clear I agree with your initial comment - there are lots of reasons even big customers can be worth repricing based on their true costs and strategic relevance. And at less than 0.1% of revenue VMWare can certainly afford to risk it.
> The strategy you described has no upside for the vendor (in this case Broadcom VMWare).
It has. Product/brand trust matters... the most obvious example being Google, who have had an extremely hard time getting their stuff to be actually used as they have shown time and time again that even if you're the biggest game publishers on this rock, you won't be able to rely on Google keeping up their end of any deal.
You do not have an individual purchaser - the deal will be vetted by committee for 1-3 quarters.
> most obvious example being Google, who have had an extremely hard time getting their stuff to be actually used
Extremely hard disagree. GCP does well in enterprise/strategic accounts where a customer in some way competes with Amazon (which is around 20-30% of the F1000) and also doesn't have a significant Windows presence (minimizing the need for Azure+MS Support)
Wait, IANAL, but... If contracts were based on good faith, why do you actually need contracts? I highly doubt what you wrote. To put too much trust/faith in a bussiness partner seems like a good way to kill your bussiness.
Negotiations should be based on good faith from both parties. If a party attempts to negotiate in bad faith, it becomes very clear and ruins the relationship.
And people enter into unfair contracts regularly without realizing it or understanding it.
Is $8m small for a VMware contract? Surely if you’re big enough at some point it becomes cheaper to put something together in house - surely this point is far before spending $100m a year?
Direct Sales means there is personnel (inbound or outbound) selling directly to the customer on behalf of the Vendor.
If you sign up to AWS on your own credit card, there was no inbound or outbound sales motion at all.
AWS's self service motion is extremely efficient for lower tier customers (as the upside is basically infinite as you basically spend $0 to generate N dollars). <- This is also why YC's Request for Startups is mentioning Open Core startups, as the open source offering acts as a loss leader (eg. Kong, Hashicorp, Pulumi eventually)
Traditionally, companies like VMWare didn't support such a model and your only way to purchase was to jump on a call with their account team, but companies are transitioning to that to target low tier sales.
This strategy works... until it doesn't. You may well be correct for the next 3-5 years; but after that those customers will leave and they won't come back. Ever.
Tech is littered with the twittering ghosts of companies that had something "essential" and "irreplaceable". The strategy is "a retreat to the high end" and a self-serving re-definition of "who are customers are" and "what we do".
Yes. Mind share is important.
Does VMWare have the best product or are they so popular because "that's what I know"?
As smaller shops switch away (the Proxmox folks must be in a pretty good mood), the job market will be full of people who know the alternatives, but not VMWare products. And when "the real thing" is so incredibly expensive, why not switch to something comparable that your people already know?
That's what I don't get either. If they can shove off one of their top 1500 customers (probably much higher) and, as someone put it, "everyone is using them", that means unless you're one of the top few hundred customers, this can absolutely happen to you. Which is what people are upset about.
And the risk they're most concerned with is the business ceasing to function.
I worked at a F500 that used VMWare; they'd probably pay a 12x subscription fee. Not because they want to, but because they started trying to get rid of their mainframe 30 years ago and 70 something percent of their business still relied on the mainframe last I was there a decade ago.
VMWare will likely go the same. It'll take them a year or two to PoC an alternative and get contracts signed, another 5 to move the low hanging fruit, and a decade to finish off the long tail of stuff that won't migrate easily.
We're talking migrating stuff running versions of Java that were EoLed a decade ago, using an app container (like Tomcat) that was also EoLed a decade ago, and trying to get that all working on a new infra stack.
And it's not just one of them, there are thousands of apps, many of which are outdated or have bizarre vendor requirements you'd likely never accept (like "we will not support you if you run this bog-standard Django app on a VM").
Stuff like this moves slooooowwww, especially at companies who only use tech to enhance their primary business (as opposed to tech being their primary business).
You seem to be assuming the customer takes up man-weeks or months of support and legal, not particularly typical for a repeat customer in that game - not sure how else you could account for the margin on $8m being completely eaten up
I work in enterprise saas as well and I can't imagine how bad this customer must be that they can break pricing unless VMware runs really low margins vs the industry. There must be other things as well
8m is 0.6% of their revenue, and they grew in revenue 4% from 2022 to 2023[1]. Losing 0.5% kinda hurts when they are growing that slowly. That 8m can't be that easily magicked out of existing customers or the freed up resources supporting that customer they fired. I know nothing about their business or financials however
My hunch is an Account Team gave a heavily discounted deal to land an upper market customer in order to meet quarterly commitments or get a nice commission.
The customer ended up accepting the $100M price-point as well (https://twitter.com/cioontherun/status/1761882689886433742), so they were clearly mis-tiered. If you can afford (what I assume is) a multi-year contract with a 9 figure TCV, you are generating 9 or 10 figures in revenue.
I've heard of similar shenanigans down the grapevine at w/ Cisco and Dell EMC and absolutely wouldn't be surprised if a similar thing happened at VMWare.
I can understand VMWare doesn't want to keep doing this, but with $8M, assuming a 100k annual salary, you can dedicate 80 full time employees to this customer yearly.
I understand 8M is nothing next to 13.5B, but that can't be unprofitable, can it? Or the contract is too low.
1. "assuming a 100k annual salary" - the cost of salary is 1.75x stated wage, so in reality it's $175k
2. "you can dedicate 80 full time employees to this customer yearly" - you mean you need almost a hundred bodies per customers?!? In fact, this is a major reason to fire a customer. It means they are demanding and trying to outsource their entire org to you. Either pay up or piss off.
> 1. "assuming a 100k annual salary" - the cost of salary is 1.75x stated wage, so in reality it's $175k
I know, this was ballpark thinking, it's not a fundamental flaw in my thinking. 100k with taxes is still a 60k salary and that's about what I was thinking. Not great everywhere, but still good in many locations. I also don't know how much support people are usually paid, and 200k is still 40 people.
Now, indeed, I wasn't imagining hundreds of people per customer required. I'm willing to believe it's realistic for some kinds of businesses, for very big customers.
60k is likely very low for this kind of support. The people calling in are generally IT professionals that have issues a script won't solve.
When I worked with VMWare, it also wasn't uncommon for us to have support on site. If we were doing a big migration or had a significant issue, they'd send someone out for a week+.
They likely paid a premium to have support willing to travel for significant spans of time, plus now they're paying for hotels, daily stipends, etc.
Then you need labs for engineers to try to replicate problems, office space, general overhead like taxes and payroll providers, etc.
Then there's costs to the platform itself from some customers. E.g. small customers tend to be less homogenous. Every Fortune 500 has a SAN, many small companies don't. Now you need an "integrates with EMC" storage option for big companies, and an "integrates with everything else" option for small companies. Big companies want a service mesh, small companies want a billion different networking configurations, etc.
Basically big companies tend to resemble each other and view their differences as a competitive advantage they will pay their vendors to support so they can keep. Small customers diverge a lot, don't want to and/or can't afford to pay for the changes they need, and are generally more price-sensitive due to their ability and ability to switch vendors.
So simultaneously the small clients are profitable, but at a lower margin than bigger customers, and more willing to migrate away. Rather than taking a risk that they can do enough development to keep the small customers, they'd rather double down on the customers that can't switch, raise their margins, and cut their risk.
> With the level of revenue VMWare had, it's realistic to assume mean deal size is in the 8-9 digit range.
8 million is quite close to the low end of 8 digits. And at the high end of 8 digits, your assumption woud already mean they have only a few hundred customers in total, which is about as far from "realistic" as can be.
Do you have information not in the link? I don't see anything related to your comment.
> This is basically a polite way to fire a customer.
I'll throw out that raising the price like this is also a strategy to increase profit. In the absence of a source for your claim, I'll go with that instead.
> Is that $8m worth it if in aggregate you spending an equal amount of time on that account as you are on a $20-30m account?
If that's how these decisions are made, the person making them needs to be fired, and fast. They should have considered that they could handle both accounts. Increasing the number of customers is called expansion.
I do this for a living, first as a PM and now as a VC. These multiples are fairly common knowledge for anyone in Enterprise SaaS, and have have worked with accounts teams that have done this and have recommended similar strategies (at a smaller scale) to portfolio companies.
> the person making them needs to be fired, and fast. They should have considered that they could handle both accounts. Increasing the number of customers is called expansion.
I can expand the # of customers by selling 1,000 $1 contracts - that doesn't materially impact my company's financials.
You want to increase your (Revenue-COGS).
If you need to drop a needy customer to increase COGS because you have mistiered accounts caused by an account team targeting a hefty quarterly bonus (happens everywhere, and caused a big shitshow at a major cybersecurity startup recently that brought a federal investigation) so be it.
> I do this for a living, first as a PM and now as a VC. These multiples are fairly common knowledge for anyone in Enterprise SaaS, and have have worked with accounts teams that have done this and have recommended similar strategies (at a smaller scale) to portfolio companies.
So the answer is no, you don't have any information not found in the link, and everything you wrote was speculation.
> You want to increase your (Revenue-COGS).
Exactly, and without knowing anything about the financials of this deal, there's no way to know if your claims are correct. It might have been an attempt to take advantage of someone that built on their product, expecting that they'll pay more.
I've worked on these kinds of accounts at a VMWare sized company before.
If you don't trust me, go chat with CROs, Product Leadership, or Execs are similarly sized companies.
I agree this is tribal knowledge, but that's how most pricing and renewal negotiatons occur, with maybe a bit of RevOps black magic to justify a decision one way or the other.
>Is that $8m worth it if in aggregate you spending an equal amount of time on that account as you are on a $20-30m account?
I think it depends? Let's assume that both accounts are profitable overall, but one is massively more profitable. The logic to cut off the $8m account makes sense if you can move those resources towards more $20m accounts. But that isn't something that is necessarily possible, is it? There's only so many of those $20m customers out there.
This obviously changes if the $8m customer is not profitable or is just _that_ unpleasant to work with.
> But that isn't something that is necessarily possible, is it
At VMware size it absolutely is.
I've worked on accounts at similarly sized vendors and even AV or Endpoint spend will be in the $1-5M range alone even if per unit price is in the $20-30 range.
With a backbone product like VMware, you can safely assume spend is 4-6x as you will be bundling network security along with IT infra.
If you spend 75 cents to earn a dollar on one customer and can spend 25 cents to earn a dollar on another customer, you'd be dumb not to take customer 2.
It's a SAAS product why not take both customers. Don't leave breadcrumbs to feed your competitors.
It seems nobody can believe your assertion that a company would deliberately fire a profitable customer by quoting an x12 contract increase. There's no point asking me to talk to people in my business. You're going to have to point to publicly accessible resources.
Tl; dr thing you should keep in mind(and this will happen to every SaaS): 1.5 mio in revenue per employee.
Your SaaS vendor will hike prices or fire people until they get to this number. Why ? Because VC/Wall Street told them that if they want a Meta/Apple/Microsoft like valuation or a good upround they need to hit this key metric.
Over the next 2-3 years every SaaS that is priced by some sort of market(so not passion projects/ projects run by guys who don't look at comparables) will at least try to hit the above metric. Whether that is massively laying off people(ring any bells?) or massively hiking pricing depends on the product in consideration.
The era of "profitable customer" is over. Now it's all about how profitable he is, and 1.5 mio in revenue per employee is the benchmark.
While i agree in broad strokes, this isn't exactly related to upping revenue per employee.
This sounds like a mis-tiering of an upper market or strategic account by an account team.
That said, Margins and RPE are critical now.
A lot of this is also because there are too many damn SaaSes now, and enterprise products bill monthly now as well.
In 2013 the average was 8 SaaS products per company and in 2023 around 80 (I can't remember the blog link I got this from but it was a fellow VC or CRO who pointed this out).
There isn't as much money floating around anymore as everyone is at their limit, and it's getting ridiculous to spend a couple hundred thousand on some random SaaS that's unneccesary.
I think in this particular case you are probably right(you see much more experienced in actual enterprise sales), but I think my more general point is that these sorts of price hikes are what is priced into this business now.
VMware was bought at about 400k rev/employee, Broadcom was running at about 1.7mio per prior to that. Tan is essentially running a PE buyout shop that is listed, and he uses a lot of leverage. He might not get it to 1.5 mio but he wouldn't have bought it if he thought it could churn out anything less than like 1.2(over 6-7 years).
> these sorts of price hikes are what is priced into this business now
Pretty much.
> Tan is essentially running a PE buyout shop that is listed
I'd disagree on that. A PE play would severely underpay engineering and look for a short (2-3 year) turnaround by re-IPOing. Meanwhile, Broadcom tries to hold portfolio companies and pay AWS level salaries to hires.
Instead, Broadcom is using the same strategy that PANW and ZS are mimicking to own the F100 market.
There is severe vendor fatigue in the market, as various different apps are viewed as a massive cost inefficency and there is a very real chance of different apps becoming an attack vector (eg. Slack at Uber).
The only solution for this is vendor consolidation, as this
1. Limits the amount of RFPs you need to vet significantly
2. Narrow down any spending inefficiencies
3. Limits your attack vectors with a vendor breach or compliance overheads
This is why Niklesh Arora announced "Platformization" at PANW's earnings call a couple days ago, and is the same strategy Broadcom has been using for almost a decade, and what ZS+CRWD have started using as well.
This is a demand from F1000s now as well as institutional investors, and a massive reason why Broadcom's stock has gone from $40 to $1300 in a decade.
They're not volume providers so much as focused on the top revenue generating accounts. This is targeted and specific to remove the low revenue accounts from the business. It's a normal tactic.
> Just emailed the CEO of Broadcom and got his OoO: “Thank you for your email. The exorbitant price increases are due to inflation and macroeconomic conditions. Also, the taxes on my mansion in Cabo are sky high!”
Not that VMWare customers aren't faced with challenging circumstances...
Their strategy is to go after a few of the biggest organizations, and not a bunch of small mom-and-pop shops. Just like IBM with mainframes, which still seems to be profitable for them.
IBM play works because they have 60 years of legacy and path dependent development. You can't simply turn into IBM. And that strategy also made sure that IBM was the one delivering hardware to hypergrowth companies like google and friends. And it made IBM itself not into a large cloud company either.
Is there that much negativity towards VMWare? I got the impression they were generally liked by users? Still enterprise software, but good stuff, and not awful about their licencing like some other companies?
Not really. On the smaller scale there was no reason outside of familiarity to run VMware over hyper-v.
This will be a shift for on-prem deployments. For cloud solutions running VMware in AWS/Azure, unless those renewals go up significantly, most of VMwares mrr is already locked in and the people running these setups generally refuse to learn how the cloud works so it’s a pure lift and shift mentality.
But the pattern begins, broadcom buys VMware does nothing increase prices. Why would any sane client stay around to get VC squeezed.
On prem deployments at my sites swapping to hyper v prior to renewal with VMware. We saw the writing on the wall after what broadcom did to us when they bought Symantec.
The software always was fairly horrible behind the scenes, albeit with a nice user interface. However I've spent years and years dealing with the sheer awfulness which is VDDK (https://libguestfs.org/nbdkit-vddk-plugin.1.html) which leaves a bitter taste.
Does VMWare software do much more then VM handling? Is there something extra they provide to make it worth paying that much? I have only ever worked a little with the free product.
Also, what are the alternatives proprietary and open? And what are those solutions missing? Hopefully this is good news for the open project, getting more commercial support and more users.
Also sounds like good news for people like Oxide. If costumers are already considering leaving VMWare and need to do a lot of software changes, they might be willing to make a larger step for new hardware as well.
Depends what you mean by "more than VM handling". VMware handles everything you could possibly think to do with VMs and many you probably never thought about. For example: Scripted live migration. Single click to deploy a test network of VMs. Virtual desktops. Wrap up a VM in a single file that can be "played" remotely. Scan all your VMs for vulnerabilities.
Beyond VMs they attempted to branch out into containers with a version of Kubernetes (VMware Tanzu) which I believe didn't get many sales.
The competition are roughly:
- All the cloud providers
- oVirt (formerly Red Hat's RHV, probably the closest thing to vSphere in terms of UI and deployment)
- Kubenetes + Kubevirt
- Openstack, probably
- Virtuozzo
- Virtual Box
- Proxmox
- Nutanix
- Single node installations using things like virt-install
- Rolling your own management on top of KVM or Xen (surprisingly popular but I wouldn't recommend it myself)
I get that this is a deliberate strategy by Broadcom, but can someone explain how the strategy makes any sense? Broadcom apparently spent $69 billion to acquire VMware, so how does setting it on fire make sense financially?
So if you have a customer (A) whose current business is worth $80M to you, and you have a customer (B) whose business is worth $8M to you, you might look to consolidate some of your operations to try to get $88M of value out of customer A with less expenditure than you need to lay out to get the $8M from customer B.
Interesting bet.
From now on no one in their right mind will BEGIN using VMware.
Whatever new companies will be founded, none of them will choose VMware, neither will growing companies switch to VMware.
Nothing because you can already use KVM + oVirt to get something which is reasonably close to vSphere's user interface for all but the highest end users.
You are comparing what a single windows host to business is compared to whole management environment: Active Directory/Entra and gazillion of other things that help manage the business.
You can't really "build" VMware though. It's a proprietary hypervisor and ecosystem of interoperable products with tons of hardware vendor drivers. Building on KVM or even a new hypervisor would not get you 'VMware"
For 100M $ you can buy Proxmox, start advertising, scoop up all the lower end VMware customers and fill in the feature gaps according to customer demand over time.
The whole Oxide computer company raised less then $100 million $ as far as I know. With that money and open source they build not only service that can do live migration, but also the computer to run the whole thing on top.
Sure that's not a UI and doesn't include a lot of other stuff that VMWare does, but it does illustrate what a gigantic amount of money 100 million $ actually.
The underlying open source tools are so good, if you have the money to have real engineering team and combine them in interesting way and to costumize them for your needs, 100 million $ seems like it would go a very, very long way.
Specially when we are talking 100 M$ every year or something like that.
Live migration is a pretty hair-raising feature if you know how it is implemented internally, but nevertheless qemu/libvirt put in a vast effort to make it work and many Red Hat customers are using it routinely.
Or has someone done some really weird math on this... First drive the prices up, exploits as much money as you can from whatever customers stick with you and then after those left, write off the entire thing somehow...
You don't want to sell to everyone - you need to make sure the cost of retention+support is significantly less than what they are paying.
If a customer is too troublesome or spending too little (for a company like VMWare, a $8m contract is on the smaller end) this is the politest way to fire them.
It sounds like this customer is a legal firm, which makes sense - they tend to have a much smaller IT footprint.
Every dollar isn't the same.
Edit: to elaborate - it's all about margins. Is that $8m worth it if in aggregate you spending an equal amount of time on that account as you are on a $20-30m account? It ain't