Intelligent Automation is often sold as a quick, easy, and cheap software solution that can automate all your business processes with a few clicks on a ‘low-code’ interface.
Unfortunately, the promises of RPA software sales teams, are often full of marketing spin and obscure some key factors you should consider when embarking on an Automation solution. In this article, I’m going to explore one of the most important factors when selecting the correct automation vendor and solution for your business, the total cost of ownership.
To begin, I’m going to share what happens when you go beyond simple RPA automation and what it can typically cost?
Costings seems like an obvious place to start for many businesses exploring RPA for the first time
Automation tools come with various license models, often resource-based or commonly referred to as ‘Bot licenses. These licenses are linked to a resource that can be used to execute multiple processes, but it’s important to remember that each resource or bot can only do one thing at a time. Most vendors have a minimum commitment in terms of the number of licenses and a minimum license term, so you may have already committed to a higher bot capacity than you really need when starting out with your automation programme. Then, once your bots are fully utilised you will need to buy more, further increasing your total cost of ownership.
Other more flexible license models do exist in the market, such as consumption or transactional licensing, where you only pay for the work being completed. If you scale the solution, the cost will go up but in a more manageable way. There are pros and cons to each model, but the key takeaway is that you should make sure you maximise your utilisation to avoid having expensive bots sat around idle.
Like any technology, Intelligent Automation needs to be delivered in a tightly controlled way to ensure standards are met and outcomes are achieved successfully. Developers, Project Managers, Business Analysts and IT staff are all involved in an Automation project. Many software vendors will claim that their tools can be deployed with no specialist skills in a matter of days. This simply isn’t true.
Deciding to work with a Professional Services partner or consulting firm can de-risk the resourcing of your project for you, but this option has its own associated costs.
Taking things further, if you want to staff your own CoE (Centre of Excellence), you will also need to consider operational roles for the governance and management of your Automation capability. This will extend to Project Managers, Testers, Business Analysts and change management resources. For example, experienced RPA developer salaries can be as much as £60k per year. The smallest automation COE will need at least 5 unique roles to function properly, so building your own in-house team can quickly escalate into a significant operational cost. Citizen developers are often sold as the solution to your automation delivery needs, but they too come with a whole host of challenges. The promise of ‘A Bot for Every Person’ is not the future of work, more the future of shadow IT and compliance issues.
Your new software robots will need somewhere to work. This comes in the form of servers, databases, and virtual desktops (VDI’s) where they will interact with your organisations applications.
For most RPA and Intelligent Automation vendors, each robot needs its own VDI to execute business processes.
Servers can be provisioned on premise by your IT team or hosted in a virtual private cloud with one of the many cloud hosting services. Either way, servers cost money and the larger your implementation, the more these servers will cost to provision and run overtime. You will also need to consider the cost of any operating systems and application licenses needed to run on these servers to match your production environment. One basic server with an instance of Windows server, hosted on Microsoft Azure could cost as much as $1200 a month so it’s important that you ask your IT team to provide you with accurate costs for your organisation. Running your automations on physical desktops is a bad move, even for the pilot phase of your project and should be avoided. You may have spare desktops in your office, but they present challenges and risks to your automation including security and cost.
Resource costs can be considered as ‘upfront’ as the delivery of your initial automation ramps down, but the ongoing costs can also be significant.
Once you have your automations up and running, they will need maintenance.
Depending on the technology you have deployed, you may even need to manually schedule your automations to run every month, depending on your business needs.
This requires additional resources, beyond your development and testing teams, to keep your bots working properly. Any changes in underlying applications will need to be monitored in case they introduce bugs or defects, and your maintenance team will need to work closely with business operations to manage things like business continuity planning and change requests over time.
Simplified Total Cost of RPA Example:
Let’s say we have 5 processes we need to automate, which will require 10 software robots to run on simplified architecture, hosted in the cloud. What would that roughly cost?
*These costs are indicative high-level costs, that do not take into account things like volume discounts, corporate pricing and introductory offers that may be available from vendors.
Software costs are only one aspect of the total cost of ownership for RPA and, 75% of the cost of your Automation programme will come from infrastructure, professional services, and maintenance costs. When selecting an Automation partner, you must ask for details of what any solution costs to implement, run and maintain across its entire lifecycle.
Alternatively, you can take advantage of a fully managed service, like NexBotix, that focuses on reducing the barrier to entry so you can get started with Automation right away.