In 2023, AI quickly moved from a novel and futuristic idea to a core component of enterprise strategies everywhere. While ChatGPT is one of the most popular shadow IT software applications, IT leaders are already working to formally adopt AI tools. While the average use of provisioned ChatGPT licenses is still fairly low (under 30% across departments), data from Productiv shows that IT departments are using the application the most at 27%, signaling that companies are testing AI tools, like ChatGPT, to see how to best bring them under policy.
ChatGPT isn’t the only new AI software IT leaders are testing out before formally adopting the tool — which means business leaders face a pressing challenge: mastering the art of AI budgeting. The cost of implementing generative AI in business can range from a few hundred dollars per month to $190,000 (and counting) for a bespoke generative AI solution based on a fine-tuned open-source model. Whether you’re investing in AI projects directly, or buying software that has AI built-in, the costs of AI are real, growing, and being passed down to the end user. Businesses simply can’t afford to be blindsided by AI-related costs that can skyrocket without warning.
The reality is, most of the vendors you use will adopt — or have adopted — some sort of AI strategy, and many will need to adjust their pricing structures to account for the real costs of running these AI models.For their end users, these AI-assisted functionalities represent real added value. To get ahead of this, here are three ways to manage AI expenditures and maximize their return on investment in the coming year:
Brace for a change in how software is sold
Bundling is par for the course in software deals. Unfortunately, the cost of AI tools might be too high to be bundled into your existing package (even Microsoft’s Copilot costs roughly $30 per user, when a typical enterprise bundle with Microsoft sets companies back between $25-40 per user). We’re on the brink of a massive shift in how vendors package features, and that will impact your overall spend. Your legacy deals are safe, but expect new contracts to reflect the fact that AI processing costs more than your average SaaS.
Part of the shift is because AI tools often adopt usage-based pricing; AI notetakers like Fireflies AI, for example, bill per minute of usage. This model can quickly escalate costs, especially for businesses that fail to monitor their usage closely. Pay careful attention to these pricing structures –– and start building out new models to help evaluate cost and value to reflect this shifting paradigm.
When it comes to AI ROI, work to understand outcomes
Nearly every incumbent software tool is launching new AI features –– if they haven’t already, they likely will soon! We’re looking ahead to a new wave of SaaS sprawl, driven by AI, which will likely prompt a reinvigorated quest to evaluate and try new vendors. There will be plenty of new tools and features to evaluate, but if you’re not paying close attention to quantifying the outcomes for these tools, you’ll be at risk of spending a lot, while getting very little in return.
It’s imperative to go beyond cost when considering if an AI tool is worth the spend –– it will be increasingly important to understand which AI tools in your stack are driving actual outcomes. Who is being more productive, and in what way? What systems or processes are being meaningfully optimized thanks to a specific adoption? Working to understand the value of outcomes being delivered will help you make the case for spending extra and empower you to decline an upsell if a feature doesn’t prove truly useful.
Calculate Total Cost of Ownership (TCO)
Assessing the TCO of AI tools requires a deeper analysis than traditional software. Assess the unique setup and integration requirements of the AI software before making the leap to a new tool. If the software requires you to train its models on your own data, costs are likely to creep up. Meanwhile, continual data and safeguard management can be more complex and resource-intensive than maintaining traditional software. For example, a healthcare provider implementing an AI-based patient management system has to factor in additional expenses for integrating with their existing electronic health records system and for continuous data management to ensure the AI’s effectiveness.
It’s clear that AI will inevitably transform business operations, but these shifts demand a new approach to budgeting and cost management . Business leaders must stay ahead of the curve by understanding the unique pricing models, evaluating AI’s role in their software ecosystem, conducting thorough value assessments, and calculating the TCO. By adopting these strategies, businesses can harness the power of AI without falling victim to unanticipated expenses, ensuring a smart, sustainable integration of AI into their enterprise fabric.
Credit: Source link
