First and foremost, my apology in advance if this hurts your feeling. It is not meant to be. My goal of this post is to highlight that there are other types of lock-in. As a result, you should not focus on vendor lock-in, but rather your business requirements. At the end of the day, IT is just an enabler in delivering business result. IT is not the business.
The term vendor lock-in is often used as if it is the only lock-in you have to consider. If customers are aware that there are other types of lock-in, they label vendor lock-in as the worst type of lock-in. Some customers go to a great length to avoid vendor lock-in, only to get deeply locked-in by other forms of lock-in. In the past 2+ decades in IT, I’ve seen many of such examples. I’m sure you have stories to tell also…
Vendor lock-in explained well by Wikipedia here, so I won’t cover it. Ed Hoppitt and Anoop Jalan have also explained it well in their blog post. It is certainly worth pondering that vendor lock-in is not what it appears on the surface. Here is a quote from their blog:
Most companies have a single vendor for their desktop operating system, but this doesn’t seem to be of great concern to most. Many companies are heavily invested, sometimes completely, in the x86 architecture. In a broader example, when leasing or renting office floor space, most companies don’t worry about all of that floor space being offered by a single vendor or in a single building
What I want to share is the other types of lock-in, as they are disguised quite well. I’m not saying they are worse. I merely say there are other types of lock in. To me, all forms of lock-in are bad. Which one is the worst? That depends on each situation.
So let’s cover the other types of lock-in. I’d give examples of each. They are just examples. I do not mean they are bad. So please don’t discuss the examples, as they are not the point I’m making.
Independant Consultant lock-in
You deal with independent consultants from an independent consulting house or multi-vendor reseller. Let’s use the abbreviation ICMV to cover all of them. They are capable of giving your products from different product vendors (a.k.a the principals). On the surface, it appears they are independent.
In here, I do not mean you get lock-in to the ICMV. You can certainly change from one ICMV to another, as you do not implement their products in production. They provide a service. What I mean here is that you do not get independent advice. You get locked-in into a particular vendor or product, intentionally or unintentionally.
There are many factors impacting their recommendation. An advice depends on many factors. I’d use an example to illustrate (again, they are just example). Say you want to invest on a storage system across your organisation. The recommendation depends on many factors, some of them may not be revealed to you. I’ll list some that I’ve personally seen in my 2+ decades in IT:
- Margin and Commission
- If vendor A is giving higher margin than vendor B, ICMV may recommended vendor A. That’s natural and makes business sense. Their top line is their margin, not the size of your PO.
- If product A from vendor A is giving higher margin than product B from the same vendor A, ICMV may recommend product A.
- Sales Rep is often directly incentivized to sell a certain product. The compensation technique varies.
- Any vendor will certainly try to up-sell. That up-sell changes the scope of discussion, and result in a different recommendation that what you originally plan to get. In my example here, your initial plan is to buy storage system, but the ICMV up-sell a converged infrastructure.
- Lack of Expertise
- The reality of IT today is it is much more complex than IT say 20 years ago. The capability of each product, the level of integration, and the functionality of each module is much more than the past. You probably experience a situation where the principal’s technical consultant or pre-sales do not know their own product very well. It is already difficult for each principal to know their product well. The situation is worse in ICMV. They have to cover many more products, none of them their own.
- If the ICMV does not have deep expertise on vendor A, it may give your the wrong advice. It may unknowingly make a mistake in recommendation.
- If the ICMV has deeper expertise on vendor A versus vendor B, ICMV may recommend vendor A. This is human nature. If they have great experience on vendor A and lousy experience on vendor B, it’s natural for them to recommend vendor A. This is especially true if they have to implement it. The problem here is they are not you, and their experience may not be your experience.
- Let me give an example. Ask an independent Java consultant whether you should develop on Java or .Net. Even though the consultant is independent of Oracle, you can be sure his recommendation is Java. That’s how he earns a living, and that’s what he knows well. Does it mean Java is for you? Not necessarily.
- New business
- If the ICMV wants to build expertise on vendor A, ICMC may recommend vendor A, even though they know vendor B product much better and has years of experience on it. There are many reasons why ICMV wants to build expertise on new vendor or product. It’s all part of business expansion.
- Strategic business relationship
- The relationship between ICMV and its principals are stronger than its relationship with 1 customer. You may be their largest customer. Your PO could be their largest PO. But you do not know how much business they do collectively over the years with their principal. I’ve seen $300m run rate between ICMV and its principal. Even if you issue a $50m PO in 2015, it’s still a fraction of their run rate. Not to mention it’s a run rate, while yours is a one time deal.
- ICMV does not typically upset their long term relationship for 1 customer.
- Vendor alignment. You probably encounter this yourself. The ICMV would politely inform you that your company is aligned to vendor A. There is territory assignment and account mapping done, without your consent.
- Personal friendship
- So far, I’ve discussed ICMV the company. But you are not buying from a company. You are buying from a human being, who represents the company. It could be the Sales Rep or SE. That person has personal experience, relationship, etc. The individual may recommend a different vendor or product than what the ICMV may recommend. This is because it all depends on the context.
Certain technology is considered as open, as it does not come from a particular vendor. While there is point in the technology, the practical business reality is different.
I’d give an example (again, this is just an example. I do not mean to say UNIX is bad). In the past, a lot of customers invested in UNIX, as it was considered open systems. A lot of its components were, and are, even open source and free. Is it considered open now? For practical purpose, it is considered proprietary. It costs a lot of money, not to mention downtime, to get out of UNIX. You get locked-in to an open system. There is a high element of lock-in, on a technology that was and is open.
Even in the past, where buying UNIX was normal, there was locked-in. Solaris on SPARC was not compatible with AIX on POWER. Your application could not share the resources. The same thing happens in Linux. It is open and free, but the business reality is you get into a lock-in. It is difficult to get out of RHEL and into another distro. You have to reinstall all your applications, test them, deal with IP and Hostname changes, document all these. The technology is open, the business reality is not.
A modern day example is OpenStack. A customer told me that the various OpenStack vendors have enough differences that it is expensive to get out and migrate. Even before adopting one distro they have to replace a lot of their existing tools. The process of being locked-in is already expensive. Imagine the process of getting out! Be wary of open technology that is expensive to on board. If it is hard to get in, it is harder to get out.
Frank Buechsel reminds me that technology is realized by people. While OpenStack is open and free, you get the lock-in effect. You are paying to train your staff, you possibly even hire new people for it. Hiring is one of the most expensive processes in a company and firing all that staff after you want to suddenly change technology is both expensive and tiring.
A great example of technology lock-in is the mainframe, because the lock-in spans decade in a large enterprise. I’m borrowing this from a great VMware Rock Star Massimo Re Ferre, whose blog I found insightful.
The mainframe isn’t a lock-in per se because it’s expensive or it’s closed-source or it’s supplied by 1 vendor. It’s a lock-in because there is no way on this planet anyone is going to re-write those millions of lines of COBOL applications any time soon.
Indeed! I remember my early days in a consulting firm. The year was 1996, and I was modifying COBOL wrote that was created in 1974. That code was 22 years old! And I was enhancing it as part of a global upgrade of a mission critical system. Yes, enhancing it, not retiring it.
The employees of the company are certainly independent. Right?
I think you know that the answer is no. It’s just it’s not discussed openly. So let’s list a few reasons why a company’s employees get the company locked-in into a particular vendor, intentionally or unintentionally.
- Lack of Expertise
- If an employee does not have experience on product A, do you think he will recommend product A to you?
- Deep Expertise
- If an employee has a lot of experience on product A, and they are good experience, which product do you think he will recommend to you?
- It maybe a good decision, because he is an expert. But what about if he leaves the company? I’ve seen companies unable to upgrade their VMware infra because employees develop a lot of internal codes. They have created an integrated platform that only they can upgrade. The company get locked-in as the employees, who are still around, have other things to do.
- New skills
- If an employee wants to build expertise on vendor A, she may recommend vendor A. She wants to enhance her career. The reason for this is the company (both HR and Management are at fault here) does not look after her career. She has to take care of herself.
- Personal experience and comfort feel
- I personally know a CIO who will only buy from vendor A because he knows the country director well. The CIO has done business with the vendor in his past company, which was in a different industry. When he joined this new company, can you guess which vendor he will bring with him? The new company get locked-in even before proper evaluation is made. I’m sure you have many stories here. You can comment or email me 🙂
- It is common for employee to recommend brand name vendor. It is established brand, so if there is something wrong, you don’t get fired. In some companies, they even have policies that they have strong preferences for certain vendors.
I hope it brings light into the subject of lock-in. Vendor Lock-In is just 1 type among 4 types. They exist and you cannot avoid them. I recommend that you review this article by Massimo.
Massimo brought up a good definition of lock-in:
The measure of “lock-in” should be how much effort (and money) it is required to move away and change course.
So before you “save” your company by avoiding vendor lock-in, think of the larger picture. The best way is to focus on the business requirements, instead of focusing on locked-in. Which vendors meet the business requirements?
I’d like to close it by quoting Grant Orchard, who explains it well in his blog post.
Vendor Lock-in occurs when the cost of changing your IT strategy outweighs the value gained from a previously made strategic decision.
Vendor Lock-in occurs when your data is an intrinsic part of an application or platform and the cost of decoupling it outweighs the value of changing to another application or platform.
Vendor Lock-in occurs when the risk of change outweighs the value to be gained from changing. This idea is inherent within the previous two definitions, but it’s useful to state it outright.