Importance of cost during software development

One of the fundamental concepts in any business decision is the cost that a company is willing to pay for a service. The cost of the services we buy is given by the traditional law of supply and demand. The more the professionals that provide a service, the lower the cost. There are usually more features included in the service to be able to compete in a saturated market.

When we dive in the world of technology and mainly into software construction and design we can observe that some principles and concepts act in unexpected ways and if that behavior is ignored it might impact the quality of the final product. When costs are way below a certain limit it is usually a sign that some of the software expectations are not going to be fulfilled. There must be a careful selection of budget and how it is invested in a software product.

Scope

There is much work behind a software development project. There are many steps after stating what the software should do. There is a definition phase where a limit is set regarding functionality and time. This is very important since this limit is what relates all of the resources available including budget. Then the software is designed, coded, tested and installed in a production environment. If the whole process is unknown then it is easy to think that software will operate exactly as we need with a few generic parameters when in reality it is all the opposite. A lot of information is needed to define good functional requirements that set goals and limit resources.

The real programming effort or code writing usually represents around 10% to 20% of the whole process under most traditional methodologies. That means at around 80% of the time is spent in other activities. If they take that much of the development process then it means they should not be ignored. There is only one reasonable way to sustain a drop in budget to lower the cost and involves leaving out functionality so that less features are expected. If less work is done then less resources are used and the limits are tighter. In a desperate effort to compete with even lower prices some might be tempted (and in fact they do) sacrifice some of those valuable activities that make up that other 80%. There is still enough time, budget, and a final version will be delivered. What is it exactly that is lost then? Quality? Performance? Security?

Professional and business expertise

There are more professionals willing to lead development software projects every single day. An important factor that is considered for calculating how much time a project will require is how familiar the development team is with the programming tools, methodologies, and functional requirements that the software will implement. The more familiar they are with the development tools the lesser time required to setup an environment, automated tests, compilation rules, and packaging for release. When the methodology is new even for a few members of the team then more time is required for communication, efficient organization, and documentation. Expertise in the tasks that the software will perform is always a big plus as the team will know what the user is expecting, workflows will make a lot more sense, and there is a higher chance that the overall manipulation and input required by the software will be better optimized and friendlier.

It is possible to have a successful implementation even when no one in the development team is familiar with the business activities. It requires much more detailed information, more experience defining functional requirements to limit the scope, and a lot more time dedicated to the end user experience assessment. The methodology in use would need to be flexible to changes and longer delivery times. As you might imagine by now lowering cost is not a reasonable option. If the budget for such conditions is pushed too low something must be sacrificed. The only possible outcome is software that does not perform well, is unreliable, and has a many unsatisfied users. It will quickly become an obstacle for daily activities, waste time, and produce even higher costs due to rewrites on core functions or even another full implementation.

Expertise is a magnificent asset in software development. The more time someone has spent developing software means there is greater control when things do not go as planned. The final product is usually more polished and has solid recovery plans with proven strategies. Additionally, There is a greater range of values that are not quantified such as commitment, honesty, and that special desire to see a happy end user. We expect those values in all situations but lets face the fact that more reliability comes at a price.

Risk assessment and scalability

There are many successful projects and there are those that some would prefer not to talk about. Projects fail for many reasons and it is perfectly normal as most cannot be blamed on a single cause but rather a series of events or a series of conditions that made way for those events to happen in the first place. After things fail and all variables are analyzed expertise is tuned to foresee and steer away from unfavorable conditions. Most seasoned developers for example favor a democracy where functionality, reliability and general acceptance is more important while most inexperienced developers tend to try newer approaches to prove themselves. Using newer and possibly better skills is not bad. Risking a project when constrained to a team is. All daily activities that involved the software would need to depend on those specific pieces of code. It could pass all QA cases, perform well in test environments, and even reach production. It needs to be said that QA is also defined during the development phases and it is then also limited by scope. What happens if the newer approach in code does not scale well? What if there is a hidden security flaw that might only be exposed through pen testing? All user data might be compromised, all operations could halt. This is usually when a high price paid.

As part of the software development process there should always be a thorough analysis to determine the best way to interact with other interfaces, handle a certain number of users at the same time, and manage whatever hardware resources are needed. A more experienced development team will always have a much better overview of the whole solution and will use more stable and proven approaches.

Software development methodology

Things are done in a certain way using a defined set of steps in order to achieve an objective. That is how a process is defined. In very simple terms methodologies are proven guidelines to reach a set of goals. The emphasis can be on quality, faster iterative cycles for rapid software delivery, more traditional fixed requirements in an effort to foresee results, or simply a better strategy that aims at getting the best results when there are quite different skill sets in the development team. The main point here is that methodologies are good and if followed correctly will keep a project away from failure.

Dropping functionality does not look too good when lowering prices to compete in a saturated market. It can only be guessed that methodologies are not going to be followed in certain cases or might even be absent. There is actually a shorter delivery time but documentation, QA, and overall design will suffer. We are basically going back in time throwing away the proven benefits of newer development models and introducing human error again. Bugs will most likely be discovered once the software is in production due to inefficient QA. Lack of proper planning will probably produce unexpected software behavior and in most cases there will be little chance to even match the final product to the initial functional requirements.

If there is no way to match the initial requirements with the final product, does it even matter to discuss when the responsibility of corrective maintenance is enforced? If the cost is already too low and the project is poorly handled can we even trust there will be good corrective plans? A fair budget allows enough resources for a project so that it is properly defined, deployed, and corrected if necessary.

Only after we gather all necessary information we are able to make an educated decision. Is it worth to risk the custom software implementation of your daily business operations?

--A small fragment to show a few of the many ideas that inspire us every day at the office. Erick J. / IT Express