Background on IT Support Staffing

We calculate our TCO for different applications by applying seven different categories of support load, with each category having its own ratio (which is also adjusted for other factors such as whether the application is delivered as on-premises or through SaaS). These categories are as follows.

• Light Support
• Medium Support
• Heavy Support
• Planning System Support (Planning systems have uniquely heavy support needs, but the number of users for planning systems tends to be small)
• BI Light Support
• BI Heavy Support
• Super Heavy Support (generally only used for specific SAP applications)

Support Ratios

There are well-known support ratios, but there is extremely little research, which supports these ratios. They are most often driven from surveys, which ask what the current support ratio is within the company, and what it should be. These studies are often presented as if they are convincing evidence when they are nothing more than surveys. Secondly, these surveys do not differentiate between different types of applications. However, both software categories and specific applications differ significantly in terms of their support load.

Beyond the weak research into support ratios, buyers rarely staff the appropriate ratios of support personnel even to match the ratios. Often the support staffing will be roughly ½ what the respondent thinks it should be. Most buyers do not account for the support load of the different applications that they evaluate – instead, they assume – incorrectly, that all applications they are reviewing have an identical support load. They strangely make this assumption even though it is entirely evident that different applications have different support implications.

The vast majority of applications are under-supported, and this is one of the reasons that the ROI on enterprise software is so much lower than its possible potential, and so much smaller than is generally estimated during the software selection phase. In fact, under-supported applications is a significant reason for the high failure rate of enterprise software implementations. Business users do not get the support that they need, and often drift away from newly implemented applications to applications, which they know they can rely upon like Excel. This is mostly un-discussed among software vendors. Software vendors themselves play into or contribute to under-supported applications by deliberately underestimating the support expectations that buyers can expect to gain advantage vis-à-vis competitive applications. In pre-sales demonstrations, no functionality is every described as cumbersome or requiring a great deal of training to master, generally, the emphasis is on how easy things are to do in the application – and therefore executives often come away with the impression that the application that they have purchased will require a low level of support. Even the terminology provides a misimpression as to the actual ease of maintenance.

A perfect example of this is in the business intelligence software category where quite complex applications are referred to as “self-service,” implying that any business user can use the application without the help of either IT or of a superuser from the business. We know this to be untrue as we test these applications ourselves and without support, leveraging the application anywhere near its potential does require assistance. This is covered in detail in our Software Selection Package for BI (Software Selection Package – BI Heavy, Software Selection Package – BI Light).

In one of the few studies on value derived from application implementation that we were able to find, one software vendor took the estimated financial benefit from its top 10% customers that were most satisfied with the application, and then prorated this over the other 90% of customers, and released this as the total value added by the its applications. This is, of course, fake research. Research of this nature from software vendors should generally not be accepted due to financial bias.

Our Ratios

Our ratios are high. However, they assume that the buyer is attempting to leverage the application they have purchased. Support is expensive – it represents over 60% of the TCO of the typical enterprise application – however, there is also no good way around this expense. Buyers have certainly tried, they have reduced support ratios, outsourced support, etc.. The results of all of these attempts to minimize support costs result in lower support quality and lower utilisation of the functionality within applications. The trick to lowering ratios is in software selection and configuration, not in support. Many companies could easily keep all support in house and stay relatively low ratios if they placed predicted maintenance overhead in their software selections, and if they tempered the functionality, they did put into place with the long term maintenance implications. We estimate both the support load per application, as well as the support load as the complexity of the configuration and customisation increases.

Percentage Allocation

The number of resources assigned also has various per cent of allocations. It may make sense to assignment two people, one at 30% and one at 70% to make one full-time equivalent. Our estimates are for upon full-time equivalents. Therefore if we state five resources, this could be ten different resources, with an average of 50% allocation to the project.

IT/Business/Internal/External Resources

Internal resources more efficiently deliver support. IT outsourcing leads to lower-cost IT costs, but higher prices on the business side – as in effect, support costs are merely moved from one budget to another budget. We do not provide an estimate for outsourced support. All of our views are for internal support – therefore the estimates must be increased if the buyer uses outsourced support. The same issue applies to IT versus business support – our FTE estimate is for both IT and business resources, and we do not differentiate where these FTEs are sourced. Our estimates also account for the fact that business has a support load.

How Applications Differ

Applications differ significantly in their support load depending upon factors such as how easy master data is to change, the usability of the application, how seamlessly functionality works, among a host of other factors. We have found support levels so high for some applications, that if buyers knew the support load, those applications would no longer be purchased. Major consulting companies have a severe conflict of interest when it comes to advising buyers of low support applications because they make more money from high rather than low maintenance applications. Correspondingly, they are significant proponents of the most top implementation as well as maintenance applications, and why they have come out so strongly against SaaS.

Don’t Forget the Business Load

Most support ratios leave out the support load on the business. However, the time the business spends working on issues in the application costs as well and must be counted. Without counting this value, it’s easy to buy software, which results in high internal costs imposed upon the business – which both means less of the proposed functionality will be accessed, and user adoption will decline. Some executives consider the business users a sunk cost, as they will employ the same number of business users regardless of the application selected. This is only partially true. The number of business users may stay the same irrespective of the application; it does not mean that the same quality output can be expected from this set number of users. Taking this approach, of thinking that the business users can absorb any application is a significant reason why so many enterprise software implementations fail or fail to meet their potential. Business users only have so many hours of the day to dedicate to the use and learning of different applications. It is not an infinite resource. IT is fond of pining any application adoption limitations on the users, but it is the responsibility of the executive decision-makers to understand both the complexity and usability of the applications they choose, and to mate these applications to their business users.

The support ratios for many applications in business intelligence are so high that they are driving buyers to applications with lower support ratios. However, this trend has not extended outside of business intelligence. Buyers could at any time obtain more knowledgeable users who can pick up applications more quickly, but buyers do not decide to do this as it would mean paying more money.

Our TCO Estimators allow buyers to get the full picture of estimated support costs, which can allow you to differentiate between applications. Our application database shows that roughly 60% of the TCO of any application over its entire life within a buyer is from maintenance costs. These costs can be easily reduced. However, the best way to reduce maintenance costs is not by any of the typically recommended approaches, such as outsourcing support. The best way to reduce maintenance costs is to select better applications.