How to think about investments & ROI for Intelligent Automation
- The 5th of a series of 10 articles - How To Successfully Scale Automation and AI in an Enterprise
As many as 70% of digital transformations fall short of their objectives (BCG,2020). Among these digital transformation initiatives, intelligent automation is one of the larger investment areas for companies today and seems to continue to be so in the future (HFS research, 2020). How can one ensure they succeed with their automation & AI strategy and become one of 30% percent that fulfills their digital transformation objectives? Based on my experience as head of automation & AI at a large Enterprise - driving it from scratch to 1000+ deployed automation & AI solutions, giving millions of hours back to the business, and transforming hundreds of core business processes. I have seen the complexities in managing scaled-up transformations based on multiple technology platforms hands-on.
This is the fifth in a series of 10 articles where I want to share what I learned along the way and what I would do differently if I did it again today. This fifth article takes place after having discussed the strategy & vision with your automation & AI initiatives and after you have set your operating model. In this article, I will dig deeper into securing funding for your projects and how to nail the financial model.
Thoughts about Investments, ROI, and your financial model
From a CEO perspective, investments in transforming an organization with automation & AI are not remarkably different from other potential significant investments, such as opening a new factory, investing in R&D to create a new product line, or growing the salesforce. It will boil down to the benefits and cost of making the investment and the risk of not doing it. The difficulty, in this case, I assume, comes from a significant gap of knowledge and understanding of the topic, and it can be hard to measure the actual benefits.
Often, it seems, this topic becomes delegated to a CIO. Pressure is often put on the CIO to deliver IT to the organization at an ever-decreasing cost. Hence investing properly in automation & AI is challenging. It has to be squeezed into the existing IT budget that needs to be used for other major IT transformations such as cloud transformations, cyber-security enhancements, and core-system replacements. Hence, a good rule of thumb is to ensure that investments in building automation & AI capabilities at scale are funded outside of the legacy IT budget, or at least on top of the same budget. Investments in automation & AI should not only be compared against critical IT investments in the areas mentioned previously but also weighted against other business transformation investments funded from outside the IT budget. That does not mean that the IT organization cannot host the vast majority of the automation & AI capabilities; that is another question.
Two questions regarding How you can think about your investments;
- How many of you know what percentage of your IT and digital transformation budgets is spent on automation and AI?
- What percentage of IT & Digital staff is dedicated to automation & AI?
I, unfortunately, meet many companies that treat automation & AI as something non-strategic, with minor, non-transformational teams or individuals driving the agenda from several management layers below the CIO. That will not succeed nor scale. In those cases, it's "easy" to blame the technology or the people driving it for not succeeding, but is that organization really willing to grow in this area if the topic is treated with such minor attention?
2. Return on investment
I believe that there are two aspects of ROI. The first one is for an individual solution. Automation is often sold as something easy that can quickly delivers business value – that is true, but only for individual solutions, not for large-scale end-to-end transformation programs. You need to find quick wins early on since you need something to create excitement. The payback period could be around six months where you could see numbers on ROI flourish by a couple of hundred percent. These are great investments, but only if you consider them individually. McKinsey (2019) states that one fallacy when scaling automation is "the pursuit of quick small potato wins while overlooking the larger opportunities." If you consider all the capability build-up and the platform costs, you realize that you need many use cases to be automated to achieve a high ROI.
This leads to the second aspect of ROI, the structured transformation program, where you need to end up. Here you instead might have a five-year-long payback period. You need to ensure that the C-suite shares your mindset to achieve this. They need to look ten years ahead, and they need to continuously invest in this area to succeed. That will, in turn, mean adding new platforms, adding new capabilities, and exploring new use cases. You will fail some, of course, but as you learn, you will also succeed on more things, and the long-term business cases begin to look good.
For individual solutions, it is acceptable to compare the business benefit of that solution with the cost of building and maintaining the solution, not considering some of the overhead expenses on the discovery, management, and platform. However, keep these two items in the back of the head;
- Don't only look at benefits relating to FTE (Full-time employees) savings. Productivity gains, lead time reductions, and quality improvements are often more valuable than saving one or a few FTEs.
- Be aware that you need a higher volume of projects to cover the wider platform investments. As that volume grows, the delivery cost of one additional project will likely be considerably lower than in an initial scenario. Hence, the earlier journey with each technology allows for less beneficial individual business cases to allow for the growth of the volume of projects.
The large-scale transformation program will most likely require a multi-year large-scale investment where the capabilities and resources grow with their demand. To enable more value-adding end-to-end transformation, you will have to shift from pure RPA investments into a multi-platform approach where you combine technologies to automate broader and more valuable use cases. Adding several platforms in an Enterprise IT environment requires time and investment. Hence, the entire end-to-end transformation program will not have a payback period of 6 months or a year, like an individual solution can have, rather maybe five years where both investments and business benefits gradually increase. After a while, when you have significant capabilities to deliver on a multi-platform "hyperautomation" strategy, you will reap significant business benefits on an ongoing basis with a decreasing investment in new capabilities. When you reach that stage, you can also optimize the "installed base" of live solutions you have built over the initial few years.
3. Financial model
Connected to the operating model and the ROI is the financial model that answers "who is paying, and for what?". The way you design the financial model will drive certain behaviors, and factors like ownership and speed will determine the success rate.
In general, the more and higher centralized budget you have, the faster you will proceed. In turn, the more the different business units pay for themselves, the more "skin in the game," which is essential to drive long-lasting transformation, avoiding the "not invented here"-syndrome.
Connecting back to the six pillars of the operating model described in the last article, I will share some thoughts on how you can think regarding the financial model for each piece of the operating model.
- Starting with the Business Transformation & Business Value Realization. Major financial items in this part are; time spent for management and key support functions as well as the transformation drivers from the CoE. Here you also include costs related to management consultants that support the initiative and its transformation efforts. No transformation of business units will happen unless crucial management parts are sufficiently involved; hence this will predominantly be a decentralized cost item. However, senior transformation competence is needed, likely centrally funded, who ideate and support the business transformation.
- Opportunity identification/pipeline generation is the oil of the machinery in any automation & AI effort. Without a steady pipeline, your initiative will soon stall and lose traction. This is a continuous investment; it's not enough to build an initial pipeline and then believe it's all solved. I am a strong proponent of having centralized funding for this. The reason is that business units most often lack the competence to identify and qualify an opportunity. It will not happen unless you provide the right capabilities to work in a structured way with the units. The business will likely only approve a solution once a business case can be presented for the solution. The purpose of the central funding is precisely that, to create solution designs and related business cases and transformation plans. Even though the specific competence is funded centrally, the business will need to provide a good amount of SME (subject matter expert) time. Without that commitment, the central investments will not generate a proper pipeline.
- Delivery is the area, I believe, where you can ensure that the business has the skin in the game. You should let the business pay once you have a stable delivery and have somewhat accurate business cases and time plans as part of the discovery phase. This is the only way to enable a scaled-up transformation program unless you have almost unlimited central funding, which I have not heard of. If you keep delivery funded centrally, you cap yourself on how much you can do, depending on your budget or how many people you have employed; hence you will not meet the demand with sufficient supply. The delivery piece will likely be your most oversized cost item; it could account for well over 50 % of the total cost for the entire initiative. Hence, it is of utmost importance that you work in a structured way to continuously optimize the delivery process to drive down the delivery cost.
- Operations costs include monitoring, bug fixing, application operations, and licenses. One of the significant issues with automation & AI operations is that these costs tend to increase linearly with deployed solutions. I don't think that is sustainable long term for companies moving into Hyperautomation. At some point in time, management will expect similar efficiency improvement and run cost reductions from automation & AI as other IT run cost items. Throwing people at the problem will work for a while (longer if you use offshored delivery), but sooner or later, you need to fix the root cause. On the financial model side, I think you should follow your overall IT run cost model, not invent a new one for automation. If you have a well-functioning run cost chargeback model, try to use the same. If you invent a new model and, for instance, charge out a full RPA license to a business, using only pieces of it, you might have many discussions about things that are not value-adding. You also tend to push rather small run cost items to operational managers who might not have the financial freedom to approve such cost items on their budget. That could lead to relatively long lead times to get approvals for these managers to cover these kinds of cost items. Hence, you can likely accelerate faster if you centralize the cost and remove it from the individual line manager's budget.
- Platforms include costs of onboarding and maintaining software and infrastructure products supporting the automation journey. There are likely significant efforts in onboarding new software platforms in large enterprises, getting all the security and architecture approvals needed. When there are upgrades in these platforms, additional efforts are required. When there are upgrades in enterprise systems, you might have to secure that your solutions keep working correctly. The more complex environment you have (multi-vendor, multi-solution), the more effort is in keeping the platforms up and running. The platform costs need to be monitored and pushed down continuously; otherwise, they risk inflating and consume quite large parts of the budgets. The strategic part of the platform costs, i.e., The ones that add additional capabilities to your automation & AI efforts, are always worth considering. The other Platform costs, i.e., costs not adding new capabilities, are not value-adding. Instead, investments directed towards engaging senior stakeholders, identifying opportunities, and optimizing processes are preferred over platform investments.These are typical IT costs that I believe should be taken centrally. This is an unavoidable cost item that an enterprise that wants to drive automation & AI at scale needs to enable the journey.
- The last bucket, costs related to Methodology & Value tracking, are, per definition, centralized cost items. These are core efforts of a CoE, if you don't do it centrally, it will either not be done at all, or you will have multiple models floating around in your organization.
In conclusion, this article digs deeper into securing funding for your projects and how to nail the financial model. I described how it, In the beginning, can be attractive to go for the quick wins to create excitement over the initiative. However, it is vital to think more long-term to succeed with the complete transformation. Putting the automation and AI initiative on the CEO and CIO agenda as any other significant investment is essential. Lastly, In general, the more and higher the centralized budget you have, the faster you will proceed. In turn, the more the different business units pay for themselves, the more "skin in the game," which is essential to drive long-lasting transformation. In the following article, I will dig deeper into the subject of optimizing your pipeline of automation opportunities.