Helpful Pivotal Case Study with Microsoft

https://customers.microsoft.com/en-US/story/merrill-corporat…

This pretty much sums of the transformation problems and internal resistance faced inside of large companies. This case study is with Merrill, an old company renowned for being the intermediary in mergers and acquisitions, in the old world Wall Street way. The company realized that their history and respect would only give them a little window of time before new cloud players would come in and disrupt their business. So management, not necessarily to developers delight (as there was clear resistance as stated below) decided on Pivotal Cloud Foundry instead of open source Cloud Foundry and moving to Azure for many reasons that are quite compelling. The company also considered going Dockers, but for reasons we discussed, chose not to go the higher maintenance, less “just work” method that Dockers requires. It is a long case study, over at least 2 different websites actually, with multiple videos, and there is probably a Forrester report somewhere, but instead of going through all of that, I pulled out the on point topics relative to the points we had discussed and the below summarizes the investment issues:

In addition to fully embracing cloud technology, Merrill adopted an Agile software development methodology and a microservices architecture. It sought a modern, loosely coupled development model to facilitate modular, reusable code creation, and wanted the best cloud platform to accommodate this model.
Merrill chose Pivotal Cloud Foundry, from Pivotal Software, which gives Global 2000 companies the platform they need to continuously deliver scalable software. Developers deploy applications written in any language, and the platform handles infrastructure configuration, scaling, and more.
“We looked at the cloud foundry from the open source community, but soon realized that Pivotal Cloud Foundry mapped best to our requirements,” Fredell says. “We also looked at Docker containers and other PaaS offerings and determined that they would require a lot more integration work. We concluded that Pivotal Cloud Foundry with Spring Boot and Spring Cloud would help us best meet our objectives, largely because Pivotal has already handled the integration issues.” Developers use Spring Boot to build Java-based microservices and Spring Cloud Services to run those microservices at scale.
Merrill, with Pivotal’s help, trained its team in Agile software development methods that are best used by Pivotal Cloud Foundry. At the same time, Merrill, Pivotal, and Microsoft worked together on a four-week proof of concept that involved setting up a Pivotal Cloud Foundry environment in Azure and creating Merrill services within it. “The adjustment to microservices was challenging,” Fredell says. “Pivotal’s help was, well, pivotal. Prying developers away from a three-tier architecture was the biggest challenge. Pivotal’s approach to development agility is great, but we really had to start with a clean slate and eliminate all assumptions that the way we did things in the past was the way to do things going forward.”

Further, key components of Merrill DatasiteOne…MongoDB, which stores transactional and metadata about content; and Elasticsearch.

I put in this last snippet because I have noticed that anywhere Pivotal talks about third party software MongoDB always gets tossed in there.

Note also how Dockers was not chosen, primarily for the reasons we discussed that even the latest containers are more like an IaaS and not a PaaS and this company saw no benefit in bearing excess complexities of infrastructure management. And I am sure there may be advantages for someone, but I am not clear what they are when all you want it to do is just work.

It also seems that modernizing a company’s software entails movement away from legacy databases, and the database of choice to turn to is Mongo.

Thus why I own Mongo as well. It is always mentioned, and is everywhere.

Seems, as DreamerDad specified, the wildcard issue is how fast will new companies jump on board given internal developer resistance.

My reply is that when it comes to large companies like this, who have seen so many disrupted in the past, so that it is no longer a surprise when disruption comes from upstarts, that there is pressure to not fall behind, and when they see their peers moving forward, pressured by firms such as Accenture, Microsoft knocking on the door, Google speaking at conventions, peers talking about how they made miracle transformations, that the herd mentality will turn more into a controlled panic to get on board and not be left behind.

This is the sort of thing where executives and managers can be fired because these are business critical issues. The company cannot fall behind, and the projects can not be like the monolithic software installations of decades ago that always fell behind schedule, way over budget, enormously expensive, and go live days were like the invasion of Normandy.

So far, since 2015 growth has been tremendous starting basically at 0 for subscription services. Will it snowball? Or will it get bogged down from long sales cycles and tentative commitments due to internal resistance? The latter scenario is of course possible, but the entire purpose of Pivotal is to stop companies that are already bogged down and fearful of falling behind.

Therefore, there should be a good number of forward looking companies who will not accept either their company to continue to be bogged down by legacy software and fall behind, nor allow the excuse that being bogged down, as something that is internally enjoyed, to allow the company to continue to be bogged down to its own detriment. Any company who cannot overcome this resistance, or have enough will to try, in this day and age of software magic and upstarts everywhere, deserves to die a slow death anyways.

We shall see.

Tinker

27 Likes

Prying developers away from a three-tier architecture was the biggest challenge. why I pay only limited attention to many “insiders” and “experts”
Resistance to change is not limited to non tech types. It is the natural human reaction to having old skills devalued and being forced to put in a lot of hard work to learn new ones. With the fear that maybe you can’t learn fast enough.

1 Like

Nice post Tinker. I squirted some tea out of my nose after reading this:

Prying developers away from a three-tier architecture was the biggest challenge.

3 tier architecture designs are at least 20 years old! It’s possible that it is accurate, but I think it is more likely that Thomas Fredell, who I see is the Chief Product Officer and not a technologist, was coming up with some technology terms from his past. But you never know - Cobol is alive and strong in the financial world.

You are pulling up some interesting case studies. I’ve been talking to another “top 5 global bank” (this one based in Asia) and may be jumping ship over to them. They have started the cloud transition with 2 separate proofs of concept, one on AWS and the other on Kubernetes. I’ve talked to the separate directors of each initiative, and they never thought about Pivotal. If I join, I’ll consider making the Pivotal business case (if something in the home country isn’t preventing it), and I may need to go back and collect some of your case studies.

7 Likes

Nice find Tinker. I’ve been trying to make sense of what Pivotal does exactly, and this has helped.

O.k., I own 10 (I think) companies that I put in the data thesis part of my port. At this point I still can’t do a GouchoChris-like one sentence summary of what they, how they fit together, overlap, compete, … I read, I find empty words (to me) describing ephemeral (to me) concepts. I’m trying to compare these functions to petrochemical processing, with which I have a nodding acquaintance. From reservoirs of crude through pipelines to refining to using…

Anyway, without making P-Chem analogies for IaaS, PaaS, DBaaS, GPU’s, various ‘software disguised as hardware’ … here, I just want to give an example of the verbiage swamp I need to drain:

MongoDB, which stores transactional and metadata about content;. That damn “transactional” is just floating there on the swamp water, looking for a noun. Is it transactional data??? I mean, doesn’t “metadata about content” translate to “data about data about content”???

I mean in my 10: Talend provides an integration platform, Nutanix provides an enterprise cloud platform, Pivotal provides a cloud-native platform, and Mongo is a modern, general purpose database platform. So I have four companies that are at least partially, or mostly, PaaS, and I’m trying to figure out if these stack or compete or function parallel or series…

Or, I can say to Hell with it. I can just throw them all against the wall and see what sticks and keep the ones that grow, sort of Rule Breakers style.

Am I alone here or are others wandering the same wilderness?

KC

36 Likes

I was in the same boat… but Tinker did all the heavy lifting to explain it. There are threads here and on NPI that explain each company in great detail. Also SteppenWulf has a knack for explaining it in English. I’d suggest you start with their posts.

Mark

5 Likes

That’s a fantastic example Tinker! It gives me hope for the world of work. I’m retired now from the coporate world but I have to say, the business I was in (nuclear power plant engineer for a major utility) was the most inefficient business I’ve ever worked in (and I’ve worked in a dozen or so different lines of business). Of course most contractors love that slowness because of the long term job stability it gave them but I could not stand it! It’s one of the many reasons I retired early. I could go on and on about all the ways it could have been better but I won’t bore you with that. From my view point, I can see this change to a faster way of doing things being the death, yes I mean death, of those industries who choose not to change. We are living in some interesting times. The cream is going to rise to the top.

Peace,
Dana

6 Likes

I feel it’s the wild wild west. (And yes, I’m old enough to remember that show.) And, it reminds a lot of the internet explosion in 1999 except with some key differences. First, these are real companies with real revenues doing real useful things. Second, instead of being mostly just the US, this time, it’s world wide. So I think it’s worth the time to read the boards and/or visit the companies websites to see what they do. Then pick a couple or few that you understand and take a small bite.

Peace,
Dana

2 Likes

@0ForFive

I mean in my 10: Talend provides an integration platform, Nutanix provides an enterprise cloud platform, Pivotal provides a cloud-native platform, and Mongo is a modern, general purpose database platform. So I have four companies that are at least partially, or mostly, PaaS, and I’m trying to figure out if these stack or compete or function parallel or series…

I can try to help here where I have some knowledge - but short descriptions won’t capture all the flavour, and I’m no expert on everything cloud. You have two cloud companies, and two data companies, and each is in completely the opposite end of the market, so there is no overlap in these companies at all. In both segments, you have a company that provides the raw materials, like groceries, and a company that provides the tools, like an oven - and you need both to make a meal.

Nutanix and Pivotal are in the cloud - Nutanix provides raw materials and Pivotal provides tools. Even though they both use the word “cloud”, they mean different things by it.

When Nutanix says “cloud” they mean they will take the servers in your closet or your data centre, chop them up into virtual slices, and then you can put you applications on those slices without having to worry about which server you are running on. If a server goes down, all the virtual slices on it disappear, but the Nutanix/VMWare software sees that and just moves your applications to different server slices. Their Xi software is just going to make AWS or Azure look like the server slices in your data centre, so you can move your applications to those also - mostly (right now) to support backup and disaster recovery functions.

When Pivotal says “cloud”, it means it will take your application code, and will follow rules you set up to automatically move it wherever it needs to go, build it, test it, verify everything, and then move it from environment to environment as you need. It will make sure you always have enough instances running to support all your users, from second to second, as long as you give it access to enough resources to run it on. The destination can be AWS, Azure, Nutanix slices, VMWare virtual machines, even your local laptop.

So Nutanix provides the raw material in a form that looks just like the Amazon and Microsoft cloud - basically they do for your data centre what Amazon and Microsoft already implemented with painstaking work to create the cloud from their data centres. Pivotal uses any cloud to automatically run your applications. They are both leaders in their field. Nutanix is having a difficulty implementing Xi because this is a completely different type of software from what they have experience with, but I’m not even sure it is important to their thesis. My opinion - both are leaders and are good investment.

For MongoDB and Talend, both are in data which is also exploding. Here MongoDB is the raw materials and Talend is the tools.

MongoDB is the story of 30 year old technology (SQL) being replaced by more modern technology. This story is only tangentially about the cloud - with or without the cloud, the amount of data is exploding and old technology can’t handle it.

MongoDB is the top general purpose No-Sql database. It is implemented with an engine that allows it to run in different types of No-Sql models, so it is like a swiss army knife. It is superb with enterprise data and good as a middle tier for AI data lakes. They are coming up with ACID compliant multi-document transaction support in 4.0 - if that works they will also be great for transactional data - but this will take time to prove out. So it is trying to take over the entire database world. It doesn’t have to succeed - Oracle didn’t, but it still made lots of money for investors, and got close enough. Databases are the type of technology where once an enterprise company makes decisions about one, it is hard to change, so once you are winning, you keep winning.

Talend is in the exact opposite part of the world as Mongo. It is in the extremely crowded field of companies providing data tools, integration, and data dashboards. I had a company that was an IBM partner in California in the past, and have lots of experience with both the technology and enterprise sales behind ETL, especially Informatica which is a Talend competitor.

The thing about this space is that it is extremely easy for enterprise companies to change their minds about their data dashboards and data tools and there is very little downside to changing your dashboard on your data to move to something better (data integration is different, but there is just enormous competition in this space from enormous companies). Updating a data dashboard is also a great way for someone to get a promotion - put in a new dashboard, impress the CEO, and up you go.

With Talend’s growth and taking market share from Informatica, I have no question that it provides lots of value to users and it is going to do well in this heated market. If you were to ask me if Talend is going to succeed as a company, do well, be a good takeover target, I would say it looks good. If you were to ask me if I think it is a good investment at cloud-type valuation, I would have to say no.

They are currently winning in a market that is growing like crazy, but I think there is just too much risk that their advantage can’t hold. There are too many competitors all providing new ideas in this old ETL technology, and this is an easy place for a startup to sell to the enterprise. But they don’t just have startup competition, there is also Informatica - a thousand pound IBM gorilla that has lots of money to throw at sales, and for which there are thousands of consultants who make their living on it.

I think there is just too much risk that the competition will cause their growth to slow and/or their margins to contract - and this stock market is not kind to high flyers whose growth rate starts slowing.

Now just to repeat - my views on Talend are really based on my views on this part of the industry - I don’t know enough about Talend technology and the company, and I could be completely wrong. I’m just explaining what I think Talend is about, and why I’m not interested in investing in or even looking closer at this company.

44 Likes

Nutanix and Pivotal are in the cloud - Nutanix provides raw materials and Pivotal provides tools. Even though they both use the word “cloud”, they mean different things by it.

When Nutanix says “cloud” they mean they will take the servers in your closet or your data centre, chop them up into virtual slices, and then you can put you applications on those slices without having to worry about which server you are running on. If a server goes down, all the virtual slices on it disappear, but the Nutanix/VMWare software sees that and just moves your applications to different server slices. Their Xi software is just going to make AWS or Azure look like the server slices in your data centre, so you can move your applications to those also - mostly (right now) to support backup and disaster recovery functions.

When Pivotal says “cloud”, it means it will take your application code, and will follow rules you set up to automatically move it wherever it needs to go, build it, test it, verify everything, and then move it from environment to environment as you need. It will make sure you always have enough instances running to support all your users, from second to second, as long as you give it access to enough resources to run it on. The destination can be AWS, Azure, Nutanix slices, VMWare virtual machines, even your local laptop.

So Nutanix provides the raw material in a form that looks just like the Amazon and Microsoft cloud - basically they do for your data centre what Amazon and Microsoft already implemented with painstaking work to create the cloud from their data centres. Pivotal uses any cloud to automatically run your applications. They are both leaders in their field. Nutanix is having a difficulty implementing Xi because this is a completely different type of software from what they have experience with, but I’m not even sure it is important to their thesis. My opinion - both are leaders and are good investment.

God, SteppenWulf, you have a way of putting things that makes them really clear and easy (for me) to understand. Thanks so much for all your contributions to the board!
Saul

25 Likes

God, SteppenWulf, you have a way of putting things that makes them really clear and easy (for me) to understand. Thanks so much for all your contributions to the board!

Let me just echo that sentiment. Your contributions have been invaluable lately, Steppenwulf. Thanks so much. Same to Tinker, GauchoChris, DreamerDad, et al…Great stuff lately!

Matt

13 Likes

God, SteppenWulf, you have a way of putting things that makes them really clear and easy (for me) to understand. Thanks so much for all your contributions to the board!

Thanks a lot Saul, Matt.

I get enormous value, Saul, from your clear investment theses on stock suggestions on this board.

It’s a great community where people from different backgrounds and perspectives can contribute their expertise, and everyone gets smarter and becomes a better investor as a result. Thank you all.

18 Likes

I’m very much there with you Oforfive, and I also appreciate the helpful explanations from others. What I would really appreciate would be any book recommendations for those not steeped in the field but that can give a helpful background and overview of the big picture. Maybe this industry is just moving too fast for anything useful to be out there, but here is an example from a much slower-moving industry.

When I was looking at investing in pipeline companies, I found it very helpful to read a book called The Domino Effect. It just did a great job of explaining how the various market pieces of oil, natural gas, production, transit, etc fit together and affected one another. It got into some technical detail, but definitely nothing requiring a chemical engineering degree.

I’ve been searching for something like that in the cloud/data world, and would really appreciate if some of the more knowledgeable people in this area have any recommendations.

2 Likes