Buy vs Build: What to consider before your next software development purchase

Technology is not just a necessity but a powerful tool. It can drive growth, make staff more productive and lower overheads. Making the right decision regarding technology is imperative.

As technology progresses, we learn what industries need to operate at their best. We can take this knowledge and build a product, test it well and sell it, often on a licence model, known as Software as a Service (SaaS). SaaS has been an important movement in the tech world and most of us interact with Fremium versions daily, think Google Docs, Gmail, social media platforms, DropBox etc.

Sounds pretty good, right? So, does this mean custom development is dead? Let’s have a look at the three main factors to be considered in any project; Cost, Time and Value.

Cost

Custom development is expensive, just like a tailored suit or a bespoke piece of jewellery. You’re getting something custom made to fit your business and your processes. You will never license a product that fits all your requirements.

Additionally, when you build a custom solution for your business you will own the IP. This adds significant value to your business, but only if you choose your development house wisely. If you commission a build that doesn’t fulfil your needs you will be left owning IP that will hinder your business, not help it.

A licenced product will normally be cheaper. However, there are a few factors to consider that impact the low entry price of licencing software. Will you require any configuration or customisation of the product? Do you need this system to integrate with other systems in your business? Many companies are now seeing the value of integrating systems to get better insights from their data, reducing the cost of manual processes and pruning redundant data. Integrating two systems usually requires custom development to one or both of the systems that are being connected.

The other factor which makes most development teams skin crawl is data migration. It is a constant unknown and often a minefield of duplicates and half completed data which doesn’t magically -fit into your new system.

Time

Hands down a licenced product that is prebuilt should be up and running faster, especially as it can take months to scope out and define the requirements on a new build. However, in saying that implementing a SaaS solution at enterprise level isn’t always smooth sailing and can take a similar amount of time; regardless of Custom or SaaS you need to consider the below.

Is your project ‘blue sky’, is it completely new and not replacing a current system? Or are you replacing an old system? It makes a big difference as data migration is always the most under estimated aspect of changing any system, there are too many unknowns and no doubt you will stumble across land mines that you couldn’t have possibly prepared for.*

So, do you have to migrate data? Where is this data currently stored? Data migration is a project within itself and often requires its own scoping and budget. Often the data to be migrated will be in multiple systems, including spreadsheets, paper filings, contracts, and across multiple databases that haven’t been touched in years. These large disparate datasets will need to be unified, cleansed, verified and matched, often line-by-line, before migration to your pristine new system. To make it harder this usually happens while the customer details and other data is still in use. Tricky stuff!

If you are building something custom ensure your staff are involved in the process for a higher adoption rate. If your new system is blue sky, then it should define the process and adoption and training should be meet with enthusiastic staff. If you are replacing an old system, breaking old habits, regardless of how ineffective they are, takes time.

*Whilst it is possible to know with a good degree of certainty every field, table and database that you want to migrate. And, it is possible to do some analysis of the data within those tables, it is impossible to know all the complexity and data inaccuracies until you start migrating. I would like to give some examples but these issues are always so unique to each project that, to put it in perspective, completing analysis to uncover any and all issues at the start of the project would be as complex and time consuming as the migration itself.

Quality

A benefit of SaaS can be that it helps you clearly define processes where you don’t already have them. However, this is a double-edged sword. If you have existing processes that are difficult to adapt to a SaaS solution, you may need to customise it or work around it. Both of which are best avoided.

If you do go for a custom solution make sure you get the most out of it. Look at ways you can increase productivity in your company, look to the future and develop with scaling in mind. It’s always important to consider who will be using the system – make sure your end users, be it staff or customer, test throughout the process. Take any and all feedback on-board and, if the original ideas don’t hold up, iterate! Otherwise you may be left with a solution your users loathe.

So, what is it going to be?

Build? Build for the future, it is an investment and like any investment you want to do your research first. Make sure you know all the problems you want to solve and then find the right partner to work with. Perhaps the biggest benefit of building is that you have full control of the end product. SaaS products usually have automatic updates so you need to ensure they’re not going to push an update that blows up your business. Just remember how you feel when Facebook enforces its updates and multiply that feeling by 1000. All people in I.T know how a change to one aspect can have a ripple effect to a seemingly unrelated part of the code. Or routine maintenance which forces shutdown of your system at your peak time. Yes, it is true shutdowns for maintenance are normally at times of low activity, say 3am. However, every business model is different and that isn’t to say a business couldn’t be operating then, 2am in London is noon in Sydney! Something you should mull over when deciding what solution best fits your needs.

Or buy? Buy if you’re looking for help with an identified market problem that has already been solved, particularly in industries that are slow to innovate. For example, well defined industry processes such as a learning management systems or finance systems, have mapped out the best practises and this has been tested in the market for decades. Whilst I mentioned customisation above, it is best to avoid customising SaaS software as it can make upgrades and support more difficult – both internally and from the software provider. That said, SaaS products normally get more stable with each generation due to others encountering the bugs before you buy the product.

If you want to explore ways in which technology can improve your business give Raw Ideas a call and have a chat with one of our experts. They are always willing to share their knowledge and provide free unbiased advice.