The 'Milk Shed' model - how enterprise buyers can avoid being semi-skimmed

Martin Banks Profile picture for user mbanks April 2, 2024
The 'Milk Shed' model has been in play for a long time, but there ways to avoid its impact on your organization.


There has been growing dismay amongst established VMware customers about the sudden surge in prices that has occurred since the company was acquired last year by Broadcom.

Those that suffered the most have been the smaller Managed Service Providers that service the needs of small businesses that can benefit from the capabilities provided by VMware but don’t have the resources to implement and manage it. 

To meet that community’s needs the company last year introduced a Virtual Cloud Service Provider program where smaller MSPs could source VMware products such as vSphere at prices that made reselling to such customers a viable business model for all parties.

Following the Broadcom acquisition, however, price hikes of as much as 10x have become the order of the day, and not surprisingly, the MSPs are upset. Both they and their customers need to be more than upset, however. They need to be proactive and already planning their way out of the situation. 

This is because VMware has become the latest target of what is being called the Milk Shed Business model. This is a model that been around the IT vendor community for many years, but which is showing signs of moving a good bit closer to center stage right now. 

It is a model that makes a great deal of sense if the purest interpretation of capitalism is your goal, for it maximizes vendor financial rewards, while minimizing costs as quickly and cleanly as possible. Inevitably, this then begs the question as to what its value is to the user community? Indeed, its impact on the latter can best be described using euphemisms such as ‘picking over the fruit basket to discard the mouldy ones’, ‘winnowing the wheat from the chaff’, or ‘carrion crows tidying up the countryside’.

The value is rarely good unless you are one of a select few, with the suggestion that the cut comes at being inside or outside the top 2,000 customers in terms of revenue. If inside they are led regularly into the ‘milk shed’ where they are divested of as much of their liquid assets as possible. 

Those that don’t make the cut can be put in a very awkward position a good deal faster than they planned. Some find that prices are hiked to alarming levels primarily intended to make them go away. All, including those that made the cut, will find that future product or technology development fades away. I note that VMware no longer has a Chief Technology Officer (CTO), and its website points to Broadcom for more information. Broadcom does not have a CTO either, only a Chief of Engineering. In addition, users can expect to see a reduction in support staff and many other functions that one might expect in a normal business. For a business in the ‘milk shed’, staffing such functions as Marketing is now superfluous expenditure that can be eliminated.

Indeed, some four and bit years ago I had an interview with Andy Nallapan, who was Broadcom’s CIO at the time. The subject was the company’s use of an application for on-boarding staff in volume, as happens when an acquisition is made. He was, however, more interested the app’s capabilities at subsequently off-boarding large numbers of staff as fast as possible, having identified those that should stay. 

It is interesting to note that, according to LinkedIn, Nallapan is now EVP, COO and CISO of Cloud Software Group, the company specifically formed in September 2022 by two equity management companies: Elliott Investment Management, which owned Citrix and Vista, which owns TIBCO, to create a vehicle that can operate the Milk Shed business model around those two companies. 

It has to be added that not all equity management operations follow the same Milk Shed model as the pair behind Computer Software Group. Some indeed should be credited with creating businesses that are now making a significant mark on major business users of IT. 

Take, for example, Thoma Bravo and Dynatrace, which it launched on an IPO back in August 2019. According to Thoma Bravo, it spotted a niche in the emerging cloud services market and sought to find a suitable emerging business that fitted, and hit upon Dynatrace. Compuware, however, beat Thoma Bravo to the acquisition so, after some time it bought Compuware. 

After much slicing up and stitching bits of its roster of businesses it pulled Dynatrace out of Compuware, added the capabilities of another business it owned and created a business now valued at some $16 billion.

Plan B, and what it should consist of 

For all the customers of ‘Milk-Shed-ed’ vendors – and there will be more in the future, of that we can be certain - there comes the point where some important questions must be answered around the core subject , starting with Now what do we do?’.

For some customers the answer could be to stick with it. But this would only make sense if the business made the cut as one of the most valuable in straight revenue terms. It would also only make sense if the business is so mature that they cannot envisage any need for any changes in the applications they are currently running  – in other words they themselves see little prospect of their own business lasting more than a couple of years. If those companies have customers of their own that want to continue or can see their own prospects for future growth, they will be looking for ways out anyway, unless the price hikes are not passed on.

So in practice any attempt by a customer of a milk-shedded business to maintain business stasis will, most likely, only hasten its ride towards entropy, for staying with the vendor only tells its own customers that the business has given up trying. What is more, it will not be hard for them to guess that they are likely to be asked to pay more for services they can probably get elsewhere, and probably cheaper. 

It is an interesting side issue of the Milk Shed business model that it can appear a bit like a Ponzy scheme, in that it needs new cattle to keep coming along before the old ones end up in the knackers yard, yet the Milk Shed business process itself hastens their demise.

Part of customers’ plans has to be to look around at what alternative products or services are now available. The bare minimum they must provide is the same level of capability at a lower price – and lower than already being paid before massive price hikes are applied. But it is also an opportunity to consider what additional capabilities and services are now available as part of the deal with any alternative vendor.  A user may not have considered `Business Transformation’ as a necessary direction, but it maybe should reconsider such options if circumstances force the possibility on them.

So surveying what alternatives are available is the first imperative. One such when it comes to VMware is SoftIron, which has just introduced VM², a virtualization platform aiming to combine simplicity with a modern cloud experience. Developed from the company’s existing Private Cloud  capabilities it is said to provide a simple virtualization platform that installs in less than 30 minutes.

As current Citrix users may be aware Workspot is pitching itself as a replacement, with added benefits and capabilities, in the VDI sector. I spoke to President and Chief Operating Officer, Brad Tompkins,  to run through some of the questions customers of milk-shedded vendors need to ask themselves and possible new suppliers to help get a sense of what they should do next. 

Tompkins looked first at the level of importance facing the customers:

VDI is one workload that runs in the data center. That's Citrix. But the vSphere stack is really the modern data centre. The modern data centre was built on VMware, and it runs everything else. It runs SAP, it runs databases, it runs all your applications. Now it's exiting the modern data center that it helped build. I think customers have a lot of massive choices to make in the next three years.

It will even have an in indirect impact on the decisions Citrix users now have to make, as It has been one of those applications running on VMware in on-prem data centers. So does it accelerate a switch to the cloud? Does it actually accelerate what companies like Nutanix do? How might it affect the existing Citrix workload and would a change of provider  get a better result, by exploiting cloud services? 

Tompkins sees the Milk Shed model as having a serious impact on businesses of all sizes, from young and fast-growing through to established and very mature, the questions they must ask themselves, however, are significantly different. For example, more mature businesses usually have a strong focus on cost, so having suppliers that suddenly hike prices significantly could have a major impact on the bottom line: 

The other thing is changing the support arrangements. You don't get the same level of support anymore so it's going to cost more money, you're going to get less support, and you’re going to get very little, or no, innovation. I think if you're growing slowly, you're even more worried about those things because you really need to optimize every single asset you have.  And then if you're growing fast, then you need innovation. And how do you get innovation from a company that really is not focused on that anymore?

The businesses that are likely to be hit hardest, and therefore prime providers of the ‘milk’ will be those that have been running a software stack for 15 or 20 years and no longer have a clear insight what tweaks have been made, and who made them, all those years ago. They will now be facing the biggest challenge. 

There is always a risk for those customers that then consider making the jump to new alternative solutions; but that should end up being less of a risk than staying with the old platform, which is not a risk but a racing certainty. Tompkins compares the situation of running old Citrix or VMware environments to how users might feel about running old Palm or Blackberry phones from the same era, versus current Apple or Android-based smartphones. It really is a chalk and cheese level of difference, he suggests: 

That's probably the level of change in technology from what Citrix and VMware built 15 years ago to what is possible now. When you start afresh, you can start from new technology components, so you can build it differently. It does all the old things, but it does more new things also. The old things are now all Citrix can do, so do the users fully understand what they need from an alternate solution? And then does an alternate solution give them all the things they actually need? Not all the things, not all the features, but the things they need?

And if we can provide them with exactly the same interface to those tasks they can carry on as though nothing had happened.  And if they do need `this one thing’, that's something we would go build for them.

One final point Tompkins made addressed something that some of those customers might not put high up their list of self-assessment questions, but maybe should. Basically, it is the question of talent. It is always at a premium but never more so than now. Today, talent wants to be where the excitement and challenge can be most readily found. That is not on 10 year-old or more technology stacks, it is out in the hybrid cloud of serverless software-defined everything. So unless businesses that are using technology from ‘Milk-Shed-ed' suppliers are happy to drift towards entropy and a quiet grave they now really only have one option.

My take

The 'Milk Shed' business model has been part of the IT industry for a long time, but the world in which it has survived is changing fast. The long tail of established users committed to an old technology and content to be `milked’ for years is fading fast: the pace of change in IT means the long tails are getting rapidly docked. But recent activity by companies practicing the milk shed model has raised the question: what do the customers of ‘Milk-Shed-ed' suppliers need to do to avoid their own consequential entropy as the length of the tail gets ever shorter?

A grey colored placeholder image