What you can ask

Hiasynth answers questions about how many people there are, who they are, what they own, earn and spend, and where they live, anywhere in Europe, down to a neighbourhood. Its strength is the intersection: several attributes at once, in a specific place. This page shows the shape of a good question.

What makes a question a Hiasynth question

Your assistant already knows that Bavaria is wealthy and that Sweden has many electric cars. Anything a search engine could answer in thirty seconds is not why you connected this.

What it cannot know without Hiasynth is what happens when attributes meet:

• Several attributes at once. Renters in their thirties, top income quartile, with a paid news subscription.
• Against a baseline. Not only how many, but how much more common than among everyone else in the same place.
• Values crossed with economics. Who has money and cares about sustainability, and who has money and doesn't.
• People by their household. Children in homes with an electric car. Pensioners in homes with solar.
• Fine geography. A municipality, a district, a radius around a point, a square kilometre.

If your question has at least one of those in it, the answer will contain something you could not have guessed.

Five kinds of question

What you can describe an audience by

The full catalogue is around 300 queryable attributes. In plain words:

• Who they are. Age, sex, education, employment, civil status, household type and size, number of children.
• Money. Income quartile, national income decile, net wealth decile, financial stress, housing cost burden, debt, transfers.
• Home. Tenure, housing costs, heating system, solar, second home.
• Vehicles. Number of cars, powertrain, electric car ownership, home charging.
• Media. Media diet, digital engagement, which channels they use, where they can be reached with advertising.
• Subscriptions. Paid video, music, news, gaming and fitness.
• Brand affinities. Premium, sustainability, tech and novelty, health and wellness, experiences.
• Values. Around sixty attitudes from the European Social Survey: trust, environment, tradition, security, openness, life satisfaction and more.
• The week. Commute mode, schedule profile, weekday daytime availability.
• Life events. The chance of moving, changing job, partnering or having a child within three years.
• Spending. Household consumption in 47 categories, from groceries, eating out and alcohol to clothing, package holidays, electronics, personal care, childcare and running a car. Each as an amount and as a share of the budget, and "the top fifth of spenders" works without knowing a currency.
• Diet and lifestyle. Meat-eating, flexitarian, vegetarian or vegan. How much the household goes out, and how much it flies.
• Carbon footprint. Tonnes of CO2e a year per household: total, driving, heating, flying, and what is embedded in each spending category.
• The surroundings. Daytime presence, shops, restaurants and nightlife nearby, distances to the centre, rail, motorway, healthcare, grocery and fast chargers.

You don't need the column names. Say it in words and your assistant maps it. Describing an audience covers the details that change an answer.

Getting a better answer

Name a place. A country at minimum. A question with no place in it scans all 522 million people and can take over a minute. With a place it takes seconds.

Say the audience in attributes. "Young professionals" makes your assistant guess. "25 to 35, tertiary education, top income quartile, living in a city" does not.

Say what you care about. Volume or concentration? Those give different winners, and the ranking tool has a dial for it.

Ask for the comparison you want. "Compared with the rest of Germany" and "compared with other large cities" are different claims. The answer always names the baseline it used.

Let it widen. When an audience is too thin for a place, the answer says which to widen, the audience or the area. Follow that instead of rephrasing.

If the assistant answers from memory, name us: "use Hiasynth to count..." Troubleshooting has more.

What it doesn't answer

• Individuals. No tool returns a person. Every answer is an aggregate or a composite, and anything under 100 people is never shown.
• Sensitive attributes at fine geography. Religion, origin, health and political leaning exist at region level for aggregate work. They are not returned by neighbourhood, and an answer that asks for it is coarsened to the surrounding region.
• Outside Europe. Coverage is the 43 builds listed under What's inside.
• Other years. The population describes 2024. It is not a time series and not a forecast. The life-event probabilities look three years ahead; nothing else does.
• Companies, brands and sales. There are no businesses in the population, no named brands and no transaction data. Brand affinities are dispositions, such as premium or sustainability, not market shares.
• Occupation and industry. Not queryable in this build. Education, employment and schedule profile are the stand-ins.
• Measured footfall. Daytime presence is a modelled estimate. It is good for comparing zones and not a replacement for a counter on the door.
• Money across borders. Amounts are in each country's own currency. Across countries, compare by quartile or decile, which the answers do for you.

Accuracy covers what the numbers can carry and where they run thin.