Pinpoint’s Psychometric AI can dramatically improve your people predictions.

Pinpoint Predictive has brought to market the first privacy-safe commercial applications of Psychometric AI-powered targeting and insights. Our Data Analytics Engine uses Machine Learning to transform surface-level behaviors into individual-level models of dozens of psychometric traits that are strongly rooted in the human constitution. These standardized traits are among the most powerful data known to science for predicting individual consumer perceptions and compelling your customers to action.

By comparison, lookalike audiences and retargeting are inherently backward looking. They don’t model the raw psychological signal or translate into better personalization. Other audience approaches on the market are based on a fundamental flaw: they assume that all individuals in a segment are exactly the same and interchangeable.

Pinpoint provides you with the industry’s only custom psychometric audiences that optimize around your ad campaign in a truly data-driven way. Our psychometric models distill 70 years of scientific research on the motivations behind consumer decisions. The efficacy of psychometric advertising has been verified by research at Cambridge, Wharton, Columbia, and Stanford Business Schools.


Pinpoint’s optimized Thinkalike™ audiences can outperform lookalike audiences by up to 320%. Why? Because ad targeting is a psychological prediction, and how you think is more predictive of what you do than what you look like.


The only targeting method that allows you to identify your campaign’s Total Addressable Market. Our percentile-based audiences reveal the probability of engagement of users across the entire targeting universe. On average, 90% of users in a Pinpoint Thinkalike™ audience are missed by an equivalent lookalike audience.


Pinpoint’s custom Thinkalike™ Audiences optimize around each of your campaigns based on who’s most likely to be interested in them. Audiences can optimize around hashed emails, webpage visitors, mobile ad IDs, and ad engagement.


For the first time, you’ll have dozens of granular psychological and demographic insights into who’s actually engaging with your brand – and why. Our Explainable AI allows your team to take personalization to the next level.


Pinpoint’s proprietary technology is the only method that generates individual-level psychometric models while completely preserving user anonymity.


Pinpoint’s custom audiences function like a psychological recommendation engine, allowing you to show greater numbers of people more diverse content that they’re interested in — before they discover your product. When consumers’ purchases match their personalities, they spend more money and they’re happier customers.

The more time you spend researching consumer personality and brand preference, the greater the likelihood you’ll want to use psychometric targeting.


What is psychometrics?

Psychometrics is the scientific field that measures individual psychological differences, including personality traits that are strongly rooted in the human constitution (e.g., quantifying a person’s precise degree of Introversion versus Extroversion).

Why use psychometrics over other data?

Empirically derived psychometrics serve as a universal standard in people data that can compare people everywhere on the same traits. Psychometrics are among the best data known to science for predicting an individual’s future perceptions, preferences, and actions. Behavioral targeting, in contrast, uses surface-level behaviors to predict the same or similar behaviors in the past.

What’s the difference between psychographics and psychometrics?

Psychographics and “personas” have been used for targeting for many years in the industry. They’re typically invented marketing terms that bucket people together based on assumptions about their interests. Psychographic personas often combine demographics and stereotypes in a way that makes sense to our intuition but delivers weak performance, because they don’t capture the underlying psychological signal. This is an enormous problem because the paramount challenge in marketing is to make psychological predictions – and as accurately as possible. Our rigorous psychometric data is different from psychographic personas in the following ways:

  1. Our psychometrics measure each individual’s personality traits, modeling real individuals rather than categorizing people into assumptive buckets where everyone is supposedly the same. Empirically derived personality dimensions are normally distributed (like height, weight, and blood pressure), so we model them continuously as well. This answers questions like, “Exactly how open- or close-minded is an individual?”
  2. Unlike psychographic personas, our data are based on 70 years of psychometric research. Pinpoint models personality traits that are strongly rooted in the human constitution. We use the statistical branch of psychometrics to clean our data and validate our personality dimensions.

Our data class serves as the only universal standard in people data that can compare people everywhere. Imagine being able to line up hundreds of millions of individuals in order from most secular to most religious. That’s exactly what we do. Now, feel free to look up “Religiosity” in Wikipedia, and you’ll find real data comparing people on this trait from all over the world. Below, you’ll also see measurements of this trait’s heritability, showing how much of the variation in religiosity comes from culture and the environment and how much comes from genetic differences between individuals. These are the ingredients that the most predictive models of human individuality are made of. In contrast, feel free to look up a persona like a “Soccer Mom”. Not much actual data there, right? Not only can we use data to quantify the spectrum of religiosity; we can do it for dozens of other traits, including compassion, gender, and optimism.

What do you mean by individual-level modeling?

Unlike any other targeting data, Pinpoint uses machine-learning algorithms to model human individuality – the psychological traits that are the root cause of why each consumer has different perceptions and behaviors. Our Engine generates hundreds of millions of models that quantify each audience member’s personality and demographic traits on continuous scales. This is done in an entirely data-driven way that involves no human assumptions. Pinpoint models each audience member in dozens of dimensions. Each user is like a combination lock with dozens of dials – totally unique. We’ve finally achieved the first true segment of one. Pinpoint’s algorithms evaluate each user on an individual basis for inclusion in your custom audience, based on interpretable traits so that you’ll understand what’s driving customer interest.

What’s wrong with conventional targeting methods?

The problem with conventional targeting methods is that none of them has a way to measure the psychology of an individual and to compare them with other audience members in a standardized, universal, and interpretable way.

Publisher-based targeting: Running ads on specific publishers makes assumptions about content and context, averaging together groups of individuals who are psychologically very diverse.



Behavioral retargeting: This type of targeting is quite effective, though there are some important drawbacks:

  • Targets only a tiny part of your full market
  • Predicts the past
  • Reaches people after they’ve already formed important opinions
  • Retargets people incessantly

Psychographic personas: These combinations of demographics and stereotypes sort psychologically diverse people into enormous buckets that may seem to make sense, but don’t deal with real individual-level psychological signal. In contrast with all of these conventional targeting segments, Pinpoint’s psychometric data do model individual psychological differences. Working directly with this signal gives you a vital advantage, because at the end of the day, ad targeting is a psychological prediction. Assumptive marketing is inefficient. And even A/B testing merely provides a slightly better way to test assumptions about group averages. Playing these guessing games while other marketers use psychometric targeting is like playing chess against a supercomputer. No wonder the world’s most successful marketers put the supercomputer on their team.

What’s the difference between a Thinkalike™ and a lookalike audience?

Lookalike models function like behavioral recommendation engines. You’ve probably experienced one of these if you’ve ever been recommended a film that’s very similar to one that you’ve just watched, after you’re no longer interested. This happens because lookalike models take in surface-level behavioral data, and then they find other people who have similar past behaviors. This type of targeting is very literal, context-specific, and backward looking. Pinpoint’s Thinkalike™ Audiences, in contrast, are psychological recommendation engines. They make predictions based on individual-level models of dozens of standardized personality and demographic traits. Psychometric targeting is forward-looking and based on real interest and psychological signal rather than surface-level behaviors. So now you can reach greater numbers of people with more diverse content that they’re highly prone to be interested in – before the interest passes. A behavioral recommendation engine makes people feel as though they’re being tracked and followed around by your brand. A psychological one, in contrast, does not, removing these privacy concerns. Unlike lookalike audiences, Pinpoint’s Thinkalike Audiences are also interpretable, so you can know exactly why people are engaging with your specific creative assets and why they’re not. You can now answer questions like:

  • “Who exactly is churning and why?”
  • “What key groups are underrepresented on our website, and how can we use Pinpoint to intelligently increase our market share amongst them?”

How can Pinpoint maintain user privacy?

Pinpoint leads the field in best practices for user privacy. Please read our information about privacy here.

Information about Privacy

Pinpoint provides the industry’s first privacy-safe psychometrics for our partners and their users:

  • Our proprietary modeling method is the only technology that can generate individual-level psychometric models of digital populations with user pseudonymization.
  • Our psychometric models are linked only to pseudonymous user IDs; psychometric data are not linked to actual personally identifiable real names.
  • Our psychometric models remain on our cloud-based platform. We do not share user-level psychometric models with partners, customers, or any other third party; rather, we use these models internally to optimize audiences and to generate aggregate insights.