Have your IT investment decisions also been stretched and delayed over the past few years? There are certainly many different reasons for investments getting stuck, but typically decision-making is complicated by current geopolitical and economic uncertainty, fluctuations in interest rates, and the development of AI.

Since the spring, I have been developing and testing a hypothesis of “AI deflation” with clients.

Economists might argue that the word deflation is misused in this context, but in my view, the IT industry is currently suffering from exactly that: IT investment decisions seem difficult to make right now because what if tomorrow, next week, or in six months, you can get the same system or project much cheaper than today?

This hesitation is understandable. Generative AI has multiplied the speed of coding work, which is expected to affect the prices of applications and systems, both in terms of development and maintenance. In addition to more efficient software production, AI also naturally changes much of the functionality and content of software products and what can be done with an application in the first place, which increases uncertainty even further.

From a buyer’s perspective, lower prices sound good at first glance, but economically, deflation is a negative thing. Economic deflation postpones consumption, which slows down the economy and easily causes a self-reinforcing negative cycle. Deflation increases the real value of debt and raises real interest rates.

Fortunately, the overall economic picture is not in deflationary territory, but regarding IT investments, the situation seems to be different. However, postponing investments carries its own risks and costs. How can one know which is better: to move forward boldly, or to wait a little longer? Here are a few thoughts on the risks caused by waiting.

Risks if IT investments are postponed due to “fear of deflation” or other reasons:

1. Competitive advantage leaks to competitors who are already investing. Competitors dare to make decisions and take the lead in developing new services or streamlining operations.
2. Technical debt increases. The fragmentation of data and processes continues and worsens, and unresolved problems become more complex as the volume of data and number of users grow.
3. Expected savings may fail to materialize. At least for now, it appears that companies’ expectations for price reductions are greater than what system providers are able to implement.
4. Postponing also has a cost. Maintaining an old system, manual work, and inefficiency cost money throughout the waiting period. The “savings” from postponement are not free.
5. Information security and compliance risks increase as systems become obsolete.
6. When many postpone decisions now and decide to proceed later at the “right moment,” the risk grows that a shortage of experts and capacity will arise. This raises prices again, and the savings may not materialize at all.
7. The organization’s readiness for change erodes while waiting. When the project finally starts, key personnel may have changed or the organization’s will to change may have faded, which increases the risk of implementation.

These risks do not mean that an investment decision should be made out of fear or haste. The decision should be approached systematically. The following six methods help with that.

Six steps to a successful investment decision in an uncertain situation

1. Remember to calculate the cost of waiting. Do not just compare the cost of a new implementation now and in the future. Compare the cost of current inefficiency and lost competitive advantage to the assumed future savings.
2. Proceed in stages. Do not wait for the perfect solution. Pilot on a small scale and measure the actual savings in your own environment.
3. Build a flexible, AI-ready architecture that can leverage future technical possibilities. Ensure modularity, open interfaces, and a sound data architecture.
4. Negotiate pricing that shares the future productivity benefit (e.g., performance-based).
5. Prioritize according to risk. Invest in the most critical systems first.
6. Keep the organization and data ready for change, even if the technical implementation is waiting.

The best remedy for situational uncertainty is not to let time pass, but to make a better plan. When prioritization, phasing, and the measurement of benefits are in order, the decision will hold even if prices were to fall further.