How AI Can Forecast Demand
Until now, we have primarily viewed commercial intelligence as a way to better understand businesses. AI can analyze products, identify technologies, classify manufacturing processes, link public information, and highlight changes that could be relevant to sales. The next step comes when we no longer view this information in isolation but instead monitor it over a longer period of time.
This fundamentally shifts our perspective. Individual pieces of information become trends, and trends can give rise to patterns. It is precisely these patterns that are of interest when we want to know not only which companies are a good fit for our offerings today, but also which companies may be developing a future need.
Prediction does not, however, mean that AI can reliably know who will buy which product and when. Business decisions are too complex, too individual, and dependent on too many internal factors for that. What AI can do, however, is provide a significantly better assessment of where conditions are changing in such a way that a need is becoming more likely.
From the outside, an investment often appears to be a single event. A company orders a new machine, implements new software, automates a production step, or seeks a new supplier. For the sales team, this change often only becomes apparent once an inquiry has already been received, a request for proposals has been published, or the company is actively discussing a project.
By that point, the actual need has usually long since arisen.
Let’s take the example of a medium-sized manufacturer with its own production facilities. First, order intake increases. A few months later, the company begins looking for additional production staff, but the positions remain unfilled for a long time. At the same time, the company announces an expansion of its production space and introduces a new product line designed to produce higher volumes. Eventually, this leads to the decision to further automate a process that was previously manual.
If you look only at the later investment announcement, you see the project. If you look at the months leading up to it, however, you see the development that led to this project.
This is precisely the foundation of predictive commercial intelligence.
A single signal usually has only limited significance. A new job posting does not necessarily mean that an investment is imminent. Even a new product, a facility expansion, or a large order is rarely sufficient on its own to identify a specific need.
Things get interesting when multiple changes fit into a common pattern.
For a company with its own CNC manufacturing operations, for example, a significant increase in order volume might first become apparent. Shortly thereafter, the company seeks additional machinists, expands its manufacturing space, and finally posts a job opening in the field of automation or production optimization. For a provider of automation technology, tool monitoring, or production systems, the relevance of this company changes step by step as a result.
Not because any one of these signals proves that a project will materialize, but because multiple pieces of information come together to form a coherent trend.
AI can help identify such trends across many companies. It can analyze which changes occur simultaneously, what typical sequences emerge, and whether similar situations in the past have frequently led to specific investments or projects.
As a result, company research is shifting from a static search for suitable firms to an observation of movements within a market.
A traditional Ideal Customer Profile attempts to determine which companies are fundamentally a good fit for an offering. This information remains important, because a forecast is only relevant for sales if the company is actually part of one’s own business.
Commercial Intelligence therefore expands this perspective to include a second dimension.
The Need Profile describes which products, processes, technologies, applications, or operational situations align with one’s own offering. Continuous monitoring then addresses the question of whether something is currently changing within that company that could increase the need.
For example, a manufacturer of automation technology is not automatically interested in every company that relies on manual assembly. It becomes particularly relevant when additional factors come into play, such as rising production volumes, difficulties in recruiting staff, or new requirements regarding quality and cycle time.
It is only through this combination of a fundamental fit and current change that a truly interesting situation arises.
This is also why mere intent data often falls short. A website visit or a single search query may indicate interest, but it does not explain the underlying business development. Commercial intelligence therefore seeks to understand the bigger picture.
To make a forecast, it’s not enough to simply collect signals. It’s also crucial to know when they occur and how they evolve.
Plans for a location expansion may become known months or even years before a concrete investment is made. A series of technical job postings may indicate that a company is building up capabilities before a project is even announced. A request for proposals, on the other hand, comes very late in the decision-making process and often means that requirements have already been defined and potential vendors are known.
Thus, every signal has not only substantive meaning but also a temporal context.
For the sales team, this temporal dimension is crucial. While a very late signal may be significantly more concrete, it may offer little opportunity to help shape a project. An earlier signal is less certain but allows for a much earlier entry into discussions.
The role of Commercial Intelligence is therefore not to seek only maximum certainty. It must identify the point at which a development has become sufficiently relevant to warrant attention, even though no official purchasing decision has yet been made.
At this point, it’s worth making a clear distinction. When we say that AI can forecast needs, we don’t mean a certain prediction.
A company may exhibit all the visible signs of an impending project and still choose not to invest. Perhaps the budget will be cut, perhaps the strategy will change, perhaps management will opt for an in-house solution, or perhaps an existing supplier will take over the project.
Such information is often not visible from the outside.
A reliable forecast must therefore work with probabilities. It can determine that, based on experience, a certain combination of changes more frequently leads to a specific type of demand. This results in better prioritization, but not certainty.
The difference is crucial.
The statement “This company will purchase an automation solution in the next six months” would not be reliable in many cases. The statement “Several changes are converging at this company that typically precede automation projects,” on the other hand, describes exactly what we actually know.
For the sales team, this information is often entirely sufficient. They don’t need a perfect prediction. They need a good reason to focus their attention on specific companies.
This approach becomes particularly powerful when public information is combined with a team’s own sales experience.
Experienced salespeople often have a very good sense of when something is developing within a company. A salesperson may recognize from many projects the pattern that rising production volumes, combined with staff shortages and quality issues, often lead to investments in automation.
Until now, this knowledge has often remained tied to specific individuals.
AI offers the opportunity to leverage such experiences more systematically. It is possible to analyze which characteristics were actually visible in advance in successful opportunities, which signals were largely insignificant, and which combinations occurred particularly frequently in conjunction with projects.
This creates a learning cycle. The sales team provides experiential knowledge and actual results, while AI verifies whether these patterns also hold true for larger sets of companies.
From a need profile, an opportunity profile can thus gradually emerge—one that not only describes which customers are fundamentally a good fit but also identifies which developments occur particularly frequently prior to relevant projects.
The economic benefit of this development lies less in predicting a future purchase as accurately as possible. Far more important is the ability to prioritize the market more intelligently.
Let’s imagine that, in principle, 8,000 companies are eligible for a specific offering. After applying the Need Profile, perhaps 900 high-fit accounts remain. Among these, 150 companies are currently undergoing relevant changes, while several signals point to a particularly interesting development in 30 accounts.
For a sales team, this completely changes the starting point.
It no longer has to decide which of the 8,000 companies to call today. The research has already significantly narrowed down the relevant pool and also provides a rationale for why certain companies deserve special attention right now.
This is precisely the value of a sales forecast. It is not intended to replace the salesperson’s decision, but rather to help them allocate their limited time to where the best conditions for a meaningful conversation currently exist.
This also changes the way we think about how new customers are found.
Traditional lead generation is often a recurring project. A sales team needs new leads, compiles a list, works through it, and starts over a few months later.
A commercial intelligence system, on the other hand, can continuously monitor a market. Companies remain visible in the commercial space even if no concrete need is currently apparent. Only when their situation changes does their priority increase.
A company that seems completely uninteresting today can thus become a relevant opportunity six months later. Perhaps a new product line is launched, perhaps the production process changes, perhaps the company wins a major customer, or perhaps it reaches the limits of its current capacity.
The key is to recognize this change without having to manually re-research the entire market over and over again.
This transforms the search into a continuous monitoring of developments.
If we take this idea further, a different picture of modern sales emerges.
A sales representative no longer starts their day with a large, static list and the question of which company to target next. Instead, they receive alerts about which accounts have changed and why these changes might be relevant to their offering.
One company has announced a new production facility, another is building out additional technical capabilities, while a third is expanding its product portfolio into an area that, for the first time, aligns with its own need profile.
The sales team’s task is then to evaluate these situations, gather additional information, and decide whether they warrant a follow-up conversation.
Commercial Intelligence thus becomes a kind of early-warning system for business opportunities. It doesn’t simply monitor companies, but rather tracks changes within a market and highlights developments that could lead to future needs.
The step from search to prediction is smaller than it initially appears. As soon as companies are continuously monitored, a temporal perspective automatically emerges. Products change, technologies are introduced, jobs are created, locations expand, and production processes face new pressures.
AI can connect these changes and look for patterns that indicate future needs. It cannot predict certain purchasing decisions and should not claim to do so. Its value lies in better structuring uncertainty and alerting the sales team earlier to companies where a relevant development is emerging.
As a result, commercial intelligence is shifting once again. The question of which companies are fundamentally a good fit is giving way to the question of which of these companies are just beginning to become particularly interesting.
And it is precisely this development that leads to a fundamental conflict with one of today’s most widely used tools for acquiring new customers.
Lead databases have made B2B sales enormously efficient for many years. They help filter companies by industry, size, region, technology, or contact persons and identify large numbers of potential leads in a short amount of time.
The problem isn’t that this data is incorrect or useless.
The problem runs deeper.
As modern sales work increasingly depends on recognizing changes, needs, and emerging business situations, a traditional lead database may very reliably answer a question that is no longer the most important one for sales.
It shows us which companies exist and could, in principle, be a good fit.
Commercial intelligence, on the other hand, seeks to determine which of these companies currently present a reason for a conversation.
It is precisely this difference that is the focus of the next article:
CI-028: Why Traditional Lead Databases Solve the Wrong Problem.
In addition to manual research, we’ve developed automated tools for you:
bloo.research— Find the right B2B companies in minutes
bloo.radar - Find out what your competition—
—is going to do next before they do it.
If you want to not only understand AI but also implement it in a structured way within your company, then:
👉 Learn more about our AI training program:
https://bloo.school
👉 Learn more about our Smart Market Fit offerings:
https://bloola.com/smf - The Smart Market Fit Course
https://bloola.com/smf-system - The Smart Market Fit System for Businesses
👉 Or learn more about our consulting and automation solutions:
https://bloola.com