If a company isn’t already employing artificial intelligence (AI) to its benefit, it will soon. Both the immediate utility and long-term potential of the tool are only beginning to be realized, and yet the majority of C-suite executives already rank AI as their number-one tech priority, according to a recent BCG report.
CEOs across the globe are enthusiastically greenlighting its adoption, docking the upfront costs from relevant departmental budgets, and sitting back to watch the magic happen. Unfortunately, however, there is much more to taking advantage of AI than selecting solution providers. A full 66% of executives from that BCG report were dissatisfied with their AI programs due to a lack of an AI strategy, roadmap, and skills.
The only way to ensure an optimal return on investment (ROI) is to think long-term. Setup, training, maintenance, and more will eventually be required whether they are budgeted or not. Planning for those costs ahead of time will ensure foundations are laid before the walls begin to crack. Precisely how an AI budget should be allocated depends a great deal on how far a business has already progressed on the AI maturity curve. So each of the seven budget items listed below are applied to both new and established users of artificial intelligence.
1. Initial investment and setup
The most obvious costs are those of attaining and implementing the AI technology. This means purchasing or subscribing to the relevant hardware and software, installing said hardware and software, and integrating them with existing systems. Hardware in particular can be very expensive, so businesses should consider employing cloud computing resources like Google Cloud Platform to onsite and incorporate AI models. These base costs should be allocated roughly 30% of the AI budget to begin with—a number that reflects the importance of a firm foundation upon which all future progress will be built—before scaling down over time toward a steady 10%.
2. Research and development
Every business has its own unique set of challenges to overcome. While AI offers just the kind of fluidity needed to solve situational problems, there are many different algorithms, data sets, and models to test and configure before applying any technology to the problem itself. Services like IBM Watson can be used to develop and train AI models that are tailored to, for instance, the predictive maintenance of a particular set of machinery used by a manufacturer. Companies starting afresh should save around 15% of the budget for ensuring that the systems they start with are optimal, and gradually reduce it to 10% as time progresses.
3. Talent acquisition and training
AI is a complex system and requires skilled staff to develop, maintain, and optimize. A portion of the budget should thus be reserved both for hiring AI specialists, data scientists, or other forms of technical staff and for training existing employees. Setting up an ‘AI academy’ that integrates both external professionals and internal staff can be accomplished with an online platform like LinkedIn Learning. Such a Learning Management System—or an equivalent—can equip the workforce and minimize the need for new hires. Training is the last ‘foundational’ budget item, and should accordingly be allocated a full 20% at the outset and 10% down the road.
4. Data management
An AI system is only as good as the data it is provided with. Businesses must therefore establish processes that ensure regular collection, cleaning, and storage of data to maintain both quantity and quality. Training AI on this kind of data—doable via platforms like Snowflake—will increase the effectiveness of the technology and, more importantly, help to guarantee ethical use. Budget allocations for data management can start around 15% to make room for the foundational items above, but should eventually receive closer to 20%.
5. Maintenance and scaling
Once the AI models are in place and working, businesses should begin to think of maintenance—a necessity for any system—and scaling. Maintenance consists of regular updates and performance checks, which can be monitored by software like Amazon Web Services (AWS). AWS can also help to expand effective AI models across departments, using modular designs that scale seamlessly and efficiently. Maintenance and scaling also become more important after the initial stages of adoption, so their allocation should climb from an initial 5% toward 20% as the organization progresses in scaling AI throughout all workflows.
6. Risk management and compliance
Implementing AI comes with certain risks, particularly concerning ethical questions like data privacy and decision-making bias. Mitigating that risk involves creating guidelines and audit procedures that reinforce and monitor for fairness, transparency, and compliance with legal regulations. Products like KPMG’s Ignite AI can aid businesses in this effort by automating the governing and monitoring processes. Similar to maintenance, risk management should begin at a low budget allocation of around 5% and eventually scale up to 20%.
7. Evaluation and ROI measurement
Finally, the ROI made in the previous six budget items must be measured and evaluated. This means developing a thorough analytics structure that can indicate tangible value generated for the business—or lack thereof. Constantly evaluating existing data, together with the exponential growth of new data, provides insight as to whether or not the investments are aligned with overall company goals. Software like Tableau is useful for such visualization. Measuring and managing data is of paramount importance from the first day forward, and warrants an unwavering 10% allocation.
Final thoughts
Today’s businesses find themselves in the early years of a technological revolution. Over half of companies with 5,000 or more employees have already adopted AI, and the reasons behind the trend are evident. AI’s potential to increase revenue, reduce overhead, save time, and enhance customer experience is nothing short of groundbreaking, and the technology is improving by the day.
Even so, this promising tool—like any other tool—can only deliver on its potential when it is used correctly. CEOs will do well to educate themselves and their teams about every aspect of the technologies they adopt. With these seven necessary budget items in mind, they can then jumpstart and maintain an AI engine that consistently delivers genuine business impact.
Ed Valdez is a Partner and CMO with Chief Outsiders, the nation’s leading fractional C-Suite firm with Fortune 500 experience.
© YFS Magazine. All Rights Reserved. Copying prohibited. All material is protected by U.S. and international copyright laws. Unauthorized reproduction or distribution of this material is prohibited. Sharing of this material under Attribution-NonCommercial-NoDerivatives 4.0 International terms, listed here, is permitted.
In this article
