TL;DR: HiBob just raised $166 million on the thesis that AI’s advantage will come from context about your people. The thesis is right, but the map is half-drawn: workforce data only says what people can do, and nothing about how they work together. A meta-analysis of 10,632 teams shows that second half decides what teams actually produce. That missing layer is cultural intelligence, and it is measurable.
What does HiBob mean by organisational intelligence?
This week, HiBob CEO Ronni Zehavi announced a $166 million investment led by Salesforce and put a definition on the term the money is betting on. Organisational intelligence, in his words, is “the ability to use trusted people context, together with relevant business information, to understand what is happening across an organization, make better decisions, and act with greater confidence.” The thesis underneath it: as AI models commoditise, “the new scarcity is not AI. It is context.”
This is true. The next era of enterprise software will be defined by how well AI understands the organisation it operates inside. The question worth asking is what counts as people context, and whether it is enough.
What workforce data actually captures
The context is structural in scope: who works where, what skills and capabilities exist, how teams are structured, where there is capacity. Richer versions add the relationships between them: people and teams, skills and roles, managers and performance, goals and capacity.
All of it describes what people can do and where they sit. What it leaves out is how people work: how decisions get made, how disagreement is expressed, how feedback is delivered and received, how trust is built, what silence in a meeting means. These patterns differ between everyone in the workplace, and no HR system records them.
Why does one team consistently outperform another?
The question comes from Zehavi’s own essay; it is one of the things organisational intelligence is meant to answer, and structural data struggles with it. Two teams that look near-identical on paper routinely produce different results.
The research has an answer, and it is about the layer workforce data does not carry. The most influential study in the field, a meta-analysis of 108 studies covering 10,632 teams published in the Journal of International Business Studies, found that cultural diversity “leads to process losses through task conflict and decreased social integration, but to process gains through increased creativity and satisfaction.” Both effects, at the same time, in the same teams. As the authors later put it, cultural diversity in teams “can be both an asset and a liability.”
In other words: the differences between people demonstrably change what a team produces, in both directions. Which direction a given team gets depends on whether those differences are visible and worked with, or invisible and left to do damage. None of the variables carrying the effect (task conflict, social integration, creativity) appears in any workforce dataset.
Does AI have a culture of its own?
Yes, and it makes the missing layer more costly. In 2023, Harvard researchers administered the World Values Survey to GPT and compared its answers with 94,278 people across 65 nations. The model’s responses resembled Americans most, and resemblance fell with a country’s cultural distance from the US (r = -.70). On the triad task, a classic measure of analytic versus holistic thinking, the model was more analytic than every human population tested except the Netherlands. The study tested a 2023-era model and has not been replicated on current systems, but the direction is the point: an AI reading culture-free workforce data fills the gap with a Western default. The layer that is missing is also the one the AI is worst at guessing.
That layer has a name. Cultural intelligence is usually described as an individual skill: the ability to read how another person decides, disagrees, builds trust, and takes feedback, and to adjust. Held as data about a whole organisation, it is the second half of people context. Workforce data tells you what someone can do. Cultural intelligence tells you how they will do it with the people around them.
What does the cultural intelligence layer need to include?
The obvious answer is nationality: add a country field, look up the averages, and compare two nations on how they make decisions, give feedback, and handle conflict. That is the shortcut most tools take, and most of the popular writing on workplace culture takes it, too. However, it produces a stereotype, not a person.
The research on why it fails is clear. A meta-analysis of 558 studies built on Hofstede’s own framework, published in Management International Review, found that approximately 80% of the variation in cultural values sits within countries, not between them. On power distance, a person’s occupation explained more of the variation than their nationality did. The authors’ conclusion: “it may be more appropriate to talk about cultures of professions, socio-economic classes, and free versus oppressed societies, than about cultures of countries.”
Two things follow. First, a national average describes nobody in particular. Two colleagues with the same passport can sit further apart on how they handle disagreement than two colleagues from different continents, and if four fifths of the difference lives inside a country, a good share of the friction inside any team does too, including between people who share a nationality. Second, the layer has to carry more than a country. Profession and seniority are what the study measured; in practice the same logic runs through the rest of a person’s history: the industry they came up in, the generation they belong to, their gender.
That composed picture, nationality plus profession, seniority, industry, generation, and gender, is what we call a Cultural Signature. It starts with the individual, because that is where the variation lives, and it takes less than a minute to map.
What does cultural intelligence look like at work?
In a 1:1 or across a team, the things people usually guess at become knowable: who will speak first, how trust is built, how decisions get made, where the blind spots are, how feedback lands, how delegation works. Not an anecdote, and not a national stereotype, but a read on the actual people in the room.
For example:
- Meetings. A cross-border meeting has a predictable set of friction points. CultureFlow Connect gives you a research-backed cultural briefing on your counterpart before you walk in, built on peer-reviewed data covering 100+ countries.
- Interviews. A standard question like “walk me through why you made that decision” reads as curiosity in one culture and as a challenge to competence in another. A briefing before the interview stops you misreading the silence that follows, and losing the hire.
- 1:1s and feedback. The same words land differently across a team. Knowing where each person sits turns a generic management style into one calibrated to the individual.
- Customer meetings. Trust is built differently in different markets. A briefing before the customer conversation tells you whether to open with the relationship or the agenda.
The layer also reaches past your own offices. Organisational intelligence stops at the org chart; cultural intelligence carries into new markets, because marketing, sales, and customer teams run into the same differences an internal meeting does. The email that reads as confident at home reads as pushy in the next market, and the follow-up that shows persistence reads as pressure. CultureFlow COMPASS applies the same cultural intelligence to the prospect: how they build trust and make decisions, read before the message is sent.
The other half of the map
HiBob is right that context is the new scarcity, and right that it starts with people. It just stops too early. Your HR system knows what your team can do. Cultural intelligence tells you how they will actually work together, and that is the half that decides what they produce.
CultureFlow builds that layer. CultureFlow Connect gives you a research-backed briefing before any meeting, interview, or 1:1, and it is free to try. CultureFlow COMPASS does the same for the prospect you have not met yet.
Start with one meeting this week: get your free briefing.