During the sales process, we’re often called upon to not only set expectations of what something will cost to implement, but also, in many cases, to determine what a three- or five- year total cost of ownership will look like. A big variable in this analysis is how much money you, the client, should expect to pay after we launch your site.

In the Software world, a general rule of thumb is that when you custom-build software, you should expect to set aside roughly 20% of that initial effort to support and maintain that software on an annual basis. That means, five years of maintenance will cost the same as building the software. Most people who knock on our door don’t have that expectation.

Luckily, at Blue Acorn, we have data to work from that helps us ground those expectations in reality. And while that data may vary from the sites we build to those we inherit, we can still crunch the numbers and identify what’s realistic and practical.

The question we want to answer for you is this: “How much should I plan on spending after launch?” As you can expect, this is a loaded question with a lot of variables. To improve our findings, we broke them up into two parts. First, we have “custom software maintenance and support”  as an annual Percentage of the original project. Then we have “enhancement and optimization,” which is decided by our clients, and depends on how aggressively they want to continue to pursue improvements to the user experience after launch. Because this part is much more irregular than the first one and requires information gathered during the sales process, we’ll focus on the first component. So our new question then becomes, “What is the percentage I should anticipate for post-launch support and maintenance?”

The Data

We looked at data from launches in 2014 and 2015, and used actual hours from implementation projects. Here are a few real examples to show you the range we saw:

• Client 1 – 6%
• Client 2 – 12%
• Client 3 – ~15-20%*
• Client 4 – 23%
• Client 5 – 40%

*The site was built by another party; we estimated approximately how many hours were used to build the original site.

The average of the above is 19.6%, almost exactly the common rule of thumb. But clearly, the average is a terrible way to estimate cost. The broad range can be explained by a few factors, which will help you estimate the right number. We recommend starting with 20% and adjusting according to how you fit these variables:

• If the client’s roadmap is more strategic, the percentage is likely to be lower on the overall scale (closer to 10-15%)
• If the roadmap fluctuates consistently, then the percentage is going to be higher on overall scale (20-40%)
• The number of third-party integrations will increase the total percentage
• More internal resources account for lowered costs
• The number will be lower for more aggressive implementations
• The number will be higher for more reactive companies
• Cutting timeline usually translates to cutting scope, which leads to higher year-one costs.

In addition to looking at our own data, we also took a look at reports from Gartner and Forrester, whose researchers have looked at gross merchandise volume and the costs of an implementation as related to the costs of software in year one. Gartner stands by the idea that an implementation normally costs three times the software costs in the following year. Forrester pegs that number at 4.4x, and foresees that rising to 4.7x by the end of the decade. This year’s Forrester’s State of Retailing Online reports that, “In 2013, merchants reported that they spent approximately 5% of their web revenue on IT. That figure is now 9%.”

We hope you find this data useful in determining what your future costs will entail. Questions regarding our analysis? We’d love to hear them. We’re also interested to know if our findings aligned with your expectations.