AI FOMO: Why Companies Invest Wrongly Out of Fear of Missing Out

Companies invest wrongly out of AI FOMO because the pressure not to miss out drives decisions by visibility rather than by measurable business result. The pressure is justified, since the technology is changing real workflows. What goes wrong is the sequence: licences and pilots come first, and the question of which metric should move comes afterwards. Executives resolve this by turning pressure into a plan with an owner and a metric.

This article explains where the pressure comes from, which mis-investments it triggers and how an unsettled situation becomes a workable plan. It includes a calculation of the cost of waiting and a checklist for every approval.

What is AI FOMO and where does the pressure come from?

AI FOMO is the concern among senior leaders that their company will fall behind competitors in artificial intelligence, and it arises from a mixture of genuine change and social comparison. Social psychology has studied the concept of “fear of missing out”, among others in a 2013 study by Przybylski and colleagues, which sees in it the wish to stay continuously connected with what others are doing. The same mechanism operates in companies. Since the release of ChatGPT at the end of 2022, customers and shareholders have been talking about AI, the trade press reports continually, and every press release from a competitor becomes a benchmark.

Real evidence sits behind the pressure. A field study of more than 5,000 customer support agents at a software company found that a generative assistant raised the number of issues resolved per hour by 14 per cent on average, and by 34 per cent for novices. An experiment with 453 professionals writing occupation-specific texts found working time cut by around 40 per cent, with higher-rated quality. Executives therefore take the pressure seriously and test it, because the studies describe narrowly defined tasks with a defined way of working and a measured result, and that clarity is exactly what is missing in most corporate initiatives launched out of fear of missing out.

What does a mis-investment driven by AI FOMO look like?

Mis-investments driven by AI FOMO take four typical forms, and all of them begin with a purchase before a question. The first is the licence without a use case: the company equips many employees with a tool and hopes for individual initiative. The second is the pilot without operational ownership, which convinces in the demonstration and belongs to nobody afterwards. The third is technology looking for a problem, for instance a team introducing a chatbot because competitors have one, although the most frequent customer enquiries stall in the process and the information is seldom the problem. The fourth is selection by presentation: the most persuasive vendor wins, regardless of the company's own data and systems.

That the choice of task decides the benefit is shown by a study of 758 consultants at an international consulting firm, published in September 2023. On tasks inside the model's capability frontier, participants completed just over 12 per cent more tasks, around 25 per cent faster and at markedly higher rated quality. On a task just outside that frontier, the share of correct solutions was 19 percentage points lower than in the group without AI. The same assistant therefore produces gain or damage depending on the task, and anyone who distributes licences widely leaves that distinction to chance.

What does the fear cost? A calculation

The cost of an FOMO investment can be derived from the usage rate, and this calculation takes an hour to build. The following example is illustrative. A company books 300 licences for a chat tool at 20 US dollars per user per month, a price the provider publishes for its paid offering ChatGPT Plus. The annual cost is 300 times 20 times 12, which is 72,000 US dollars. If only 60 employees use the tool regularly after three months, the cost per active user is 100 US dollars a month, five times the list price.

The counter-calculation covers the cost of waiting. It is the value that a clearly defined use case would have created per month, multiplied by the months the decision takes. A service department of 50 people in which an assistant frees up 20 minutes per person each day gains roughly 17 working hours a day on paper. That time turns into value only if the department plans for it, for instance for faster response times or additional orders. The calculation shows that waiting costs money and that buying at random also costs money. The cost of waiting ends as soon as a use case with a metric and an owner is fixed, whereas the cost of unused licences keeps running.

Which five questions should be settled before any AI approval?

Before approving an AI initiative, executives settle five questions, and an initiative that cannot answer one of them goes back to preparation. The checklist is deliberately short so that it is actually applied in a board meeting or a management round.

1. Which metric should change, by how much, and what is today's baseline? Without a baseline, success cannot be demonstrated later.
2. Who in the business unit owns the result, and who runs the solution from day one? An initiative that belongs only to IT or only to a consulting project has no recipient.
3. Which data and systems does the application need, and are they available in regular operation? A demonstration on cleaned sample data proves little about everyday use.
4. What is the damage if the output is wrong, and what human check is foreseen? The study on the capability frontier shows that errors can sound convincing.
5. At which result after 90 days is the initiative continued, changed or ended? A stop criterion protects budget and credibility.

How do executives turn pressure into a plan?

Executives turn pressure into a plan by shifting attention from the question “How do we keep up?” to the question “Where does AI work first in our business?” This happens in four steps. In the first, business units collect use cases from processes with high time expenditure, many documents or recurring decisions. In the second, a joint team from the business and IT assesses each case for value, feasibility and risk, using written anchor points. In the third, one to three initiatives start, each with an owner and a metric, plus a stop criterion. In the fourth, the management sets a fixed review date at which findings from the running initiatives may change the ranking.

This approach takes the pressure seriously. It answers the question from shareholders and customers with a concrete sequence that can be justified, and it gives the business units a task they can complete. Three carefully chosen initiatives with an owner and a metric leave more usable experience than thirty unconnected experiments, because their results settle into operations, metrics and roles.

Niehues Consulting: from pressure to sequence

Niehues Consulting is a consultancy for AI management and organisation design from a delivery perspective, based in Düsseldorf. Founder Marin Niehues builds organisations that deliver and brings AI into operations, in projects ranging from VC-funded scale-ups to DAX 40 corporations and in more than 20 countries across six continents. In the AI management field the work runs from business analysis and requirements through the business case to the selection and sequencing of applications and portfolio steering.

The approach follows the Niehues Delivery Model with its phases Understand, Build and Deploy. Three of its checks help most against AI FOMO: Which business result should change measurably? What is the baseline, and how do we measure success? Who is responsible for operations from day one? Anyone who answers these three questions has turned pressure into a decision. Scope and terms of an engagement are discussed in the free initial consultation, which you can arrange via the contact form.