Chyral Blog

Oft-Ignored Aspects of Re-platforming

or: How to deal with a bad choice and still be a hero

Your CMS: With any luck you’ll outgrow it and need to replace it. But there are some companies who have to replace theirs long before that point. In fact, they’ll never outgrow it because they never got it to work to begin with.

Deciding to pull the plug on a moribund CMS may seem simple, but for most businesses deciding to stop fixing and start replacing is a difficult and painful process. And consultants are not always comfortable with telling a client their baby is ugly, or even able to early enough to head off disaster. But after a year or two of struggling to make a failed system work even the most irrational businesses are usually ready to cut their losses. And when they are there are some often overlooked concepts that make the bitter pill go down a bit more easily. The first of these is the concept of sunk costs, or specifically, avoiding the sunk cost fallacy.

In both economics and business decision-making, sunk costs are retrospective (past) costs that have already been incurred and cannot be recovered. Once a decision-maker has irreversibly committed resources, sunk costs become an unavoidable cost and should not be included in any decision-making processes. For example, if a business is considering purchasing a new CMS system, but has not actually purchased it yet, the cost remains avoidable. But in the case of a failed CMS, a business must choose between the following two end results:

  1. Having paid the price of the system and continuing to suffer with a system that does not address their needs
  2. Having paid the price of the system and having changed course to buy and build something more suitable

In either case, the business has paid the price of the system so that part of the decision no longer affects the future, meaning the current decision should be based on whether it can be made to work at all, regardless of the price, just as if it were a free system. Since the second option involves suffering in only one way (spent money), while the first involves suffering in two (spent money plus wasted time), option two is obviously preferable. But complicating this is the fact that managers and CTOs have a strong aversion to “wasting” resources, called “loss aversion”. In this example, many managers would feel obliged to continue with the project despite not really expecting any success because doing otherwise would be wasting the cost of the CMS and the time and cost of the teams implementing it; they feel they passed a point of no return. This is the sunk cost fallacy, and it’s flawed reasoning. Because the first option misallocates resources by depending on information irrelevant to the decision, it is in truth more wasteful than the second while superficially appearing to be less so.  The more quickly decision-makers decide to go with the second option the better it will be for both them and the business as the costs associated with time are reduced or avoided completely. Understanding why people ‘throw good money after bad’ helps decision-makers avoid this flawed thinking and make effective decisions faster.

Managers often make a subjective decision that by purchasing a CMS they are making a public and professional commitment to launching and running it successfully. To abandon the chosen system is to make a lapse of judgment obvious to the business, an appearance they may rationally choose to avoid. But a more effective line of thinking and approach is to stress the positive aspects: Though they may have comprehended the flaws of the first choice later than ideal, they still recognized the opportunity cost early enough to change course. Obviously the sooner the decision to re-platform is made the more effective this approach.

Which brings us to another useful notion from economics: Opportunity cost. Opportunity cost is the next-best choice available to someone who has picked between several mutually exclusive choices. Opportunity cost is assessed in terms of anything which is of value such as money, time, and material. For example, a business which desires to launch a CMS and a DAM simultaneously and does not have the resources to execute both can launch only one of the desired systems. Therefore, the opportunity cost of launching the CMS could be launching the DAM. But if a business delays one program while launching the other, the opportunity cost will be the time that that business spends launching one program versus the other. One case I was involved in was a project that decided to devote a period of time to conducting more integration testing rather than spending it on UI design. The opportunity cost of having a more stable system was therefore a prettier and friendlier UI, a decision that made sense for this business. In the situation of deciding what to do about a refractory CMS, the opportunity cost to a business of struggling with fixing CMS A would be implementing CMS B. For the team making that decision, the opportunity cost of staying with the wrong system could be twofold – the extra costs of struggling to fix the first system, and their reputation within the company and their field. This has a lot of explanatory power when trying to understand why companies press forward regardless trying to make the wrong system work.

Fundamental to assessing the true cost of any business decision or course of action is assessing opportunity costs. Ignoring opportunity costs results in the illusion that benefits cost nothing at all where there is no explicit cost attached to the course of action, or the cost is negligible. The unseen opportunity costs then become the implicit hidden costs of that decision. Keep in mind that opportunity cost is not the sum of the available alternatives when those alternatives are mutually exclusive to each other. The opportunity cost of a business’ decision to dedicate a server to a CMS is the loss of the server for a DAM instance, or the inability to use the server for a backup, or the money which could have been made from renting space and time on the server to a partner, as any one of these would preclude the possibility of the others. Most opportunities are difficult to compare but it remains a crucial exercise in the end.

And there are another set of commonly ignored costs that impact re-platforming decisions: Switching costs. Switching costs are any impediment to a customer’s changing of suppliers. Simply put, switching costs are barriers  to making a switch to another vendor, platform, framework, anything, really.

Types of switching costs include: search costs, learning costs, cognitive effort, emotional costs, equipment costs, development and operations costs, financial risk, psychological risk, and social risk. Some of these costs are easy to estimate. The more tangible of these to a business, search costs and learning costs; the effort and expense required to find an alternative vendor and learn how to use the new product, are usually expected in these situations, as are the costs of developing and integrating the new system, training users, and supporting it in production. But chronically underestimated or overlooked by businesses are the psychological, emotional, and social costs of switching. These switching costs go a long way to explain the reluctance of some managers to cut their losses when it becomes obvious a system is never going to meet the business’ needs.

Conversely, you can use switching costs to your benefit when soliciting a new project by employing the 10x rule business strategy. It states that order-of-magnitude improvements in costs, efficiencies, and benefits to the business will overcome most reluctance arising from switching costs. If you can show that a new system delivers improvements in most aspects on the order of 10x, you’re much more likely to secure approval, funding and have the project be deemed a success — assuming you launch it successfully.

Originally published here



  1. Rajesh Sreenivasan  February 2, 2012

    I’m sceptical of the utility of switching costs for justifying new project spend exactly because of the 10X rule: Coming up with believable numbers for costs and efficiencies on the order of X 10 is very difficult. The rest of what you’ve said makes sense and is uncontroversial. Good insight and useful to keep in mind . . . subscribed.


Add a Comment