Zero-party data is information a prospect or customer deliberately tells you about themselves: what they're trying to fix, when they expect to decide, which topics they want to hear about, or how they found you. Nobody inferred it from a click or bought it from a provider. The person typed it, picked it from a list or said it out loud.

Most B2B companies already collect a fair amount of it and use very little. The answer to "what brings you here?" on a demo form lands in a free-text field nobody reports on, and the richest version, what a buyer explains on a discovery call, sits in one rep's notes until that rep leaves.

It also gets oversold. Plenty of articles call zero-party data the most trusted data there is and suggest it arrives with consent built in. It's direct, and it's usually your best evidence of what someone wants. But people guess, round up and change their minds, and answering a question isn't the same as agreeing to every use you have in mind.

TL;DR

• Forrester introduced the term in 2018 and defines zero-party data as data a customer intentionally and proactively shares, such as preferences, purchase intentions and personal context. It was coined with consumer brands in mind, but B2B teams gather the same kind of answers on forms and calls.

• The line with first-party data is declared versus observed. A pricing page visit is something you watched, while "we're replacing our CRM before Q3" is something they told you, and your CRM should record which is which.

• In B2B the useful sources are one good question on the demo form, a "how did you hear about us" field, onboarding surveys, topic choices in a preference center and discovery-call notes captured in structured fields.

• Ask only what enrichment can't look up, one or two questions at a time, as a picklist, and give the person something back for answering.

• Treat an answer as strong evidence, not as fact. Timelines slip and budgets get guessed, so date every answer and read it next to behavior and intent signals.

• Under the GDPR, volunteering an answer doesn't give you blanket permission. Article 13 requires you to state the purpose when you collect it, and purpose limitation stops you reusing it for something unrelated.

• Answers only pay off when they sit in CRM properties that change routing, qualification or messaging. Anything left in a notes field is decoration.

Where the term zero-party data comes from

The phrase came out of Forrester. Analyst Fatemeh Khatibloo's report Q&A: What Marketers Need To Know About Zero-Party Data, published in October 2018, pitched it to B2C marketers as an alternative to third-party data that GDPR and data broker laws were making harder to use, and set out how it differs from first-party data.

Forrester's definition, repeated in its later blog posts, reads: "Data that a customer intentionally and proactively shares with a brand." It lists preference center data, purchase intentions, personal context and how the person wants the brand to recognize them as examples. The two words doing the work are intentionally and proactively. A form field you made mandatory to download a PDF is a bit of a stretch. An answer someone chose to give because they expected something better in return fits the definition exactly.

The consumer origin shows in the examples Forrester uses, like a skincare quiz that asks shoppers about their routine and goals. In B2B the same idea looks less playful and is often worth more, because one honest sentence about a buying process can save a rep two calls.

Declared or observed: the line with first-party data

First-party data is everything you collect on your own channels, and most of it is watched behavior: page views, email clicks, product events. Zero-party data is the part the person declared. Forrester treats it as its own category, while many B2B teams file it as the declared slice of their first-party data. Either way, the distinction you need is about how a value got into the record, not about which box the textbook puts it in.

Here's why it matters in practice. Two contacts both have "Evaluating CRM" in a field. For one, a rep typed it after the prospect said so on a call. For the other, a workflow set it because the person read three blog posts about CRMs. Those two records deserve very different follow-up, and if the CRM can't tell them apart, nobody can.

So we add a source to anything declared: a property for the answer, plus where and when it was given. The broader trade-offs between data you collect and data you buy are covered in first-party vs third-party data, and they apply here too. Zero-party answers are yours alone, fresh on the day they're given, and limited to the people who have already talked to you.

What B2B buyers volunteer, and where it ends up

Consumer brands collect zero-party data with quizzes and preference centers. B2B teams have fewer chances to ask, but each one carries more weight. These are the places it usually comes from:

• The demo or contact form. One open or picklist question such as "what are you hoping to fix?" or "when are you looking to have something in place?"

• "How did you hear about us?" A free-text answer here is the backbone of self-reported attribution, and it often names podcasts, Slack groups and peers that no tracking pixel ever sees.

• Onboarding and trial surveys. Team size, main use case and the tool being replaced, asked once someone has already signed up.

• Preference centers and subscription choices. Which topics, which frequency, which language.

• Discovery calls. Budget range, decision process, current vendor and contract end date, said out loud to a rep.

• Calculators and assessments. The inputs someone types into an ROI calculator or a readiness check describe their situation in their own numbers.

Discovery calls produce the most valuable answers and lose the most of them. A buyer tells a rep the contract with their current vendor ends in March, and that fact lives in a notebook or a call recording. Putting it into a date field on the account is the cheapest zero-party win most sales teams have.

Writing questions people will answer

Every question on a form costs you some submissions, so each one has to earn its place. Our rule is simple: never ask a prospect for something you could look up. Company size, industry, location and tech stack can be filled in by inbound enrichment the moment the email address arrives. Spend the question on what no database can know, like the problem, the timing or the trigger.

Forrester's own guidance points the same way. A 2020 post describes small, mobile-friendly experiences that ask for three to four pieces of information, and its 2025 advice is that the most successful ones are short, simple and offer a clear value exchange.

A few habits make the answers usable as well as plentiful. Offer a picklist with an "other" option instead of a blank box, so the answers can be counted and routed. Ask at the moment the question makes sense, like timing on a demo request or use case at signup, and save the rest for later touches. Then give something back that depends on the answer: a demo that skips the parts they don't care about, a benchmark for their segment, content on the topic they chose. People share more once they've seen an answer change what happened next.

Saying it doesn't make it so

A declared answer removes one kind of error, our misreading of someone's behavior, and keeps another, which is the person's own guess. In our experience the timeline field is the least reliable thing a prospect will tell you. "This quarter" on a demo form often means "I'd like to," and budget ranges picked from a dropdown are frequently a shrug in picklist form. The picklist also shapes the answers, since many people pick the first option that's close enough.

None of that makes the data weak. It makes it evidence that needs a date and a second opinion. That second opinion usually comes from behavior and intent signals, which work at the opposite end. Intent data can flag hundreds of accounts that might be researching your category, each with a probability attached. A declared answer comes from one person at one account and tells you what they say they want.

When the two agree, act fast. When they disagree, for instance someone says "just researching" and then visits the pricing page four times in a week, trust the more recent signal and ask again. Answers also go stale. A preference stated eighteen months ago, or a timeline that has already passed, belongs in your data hygiene routine like any other field, with an expiry or a prompt to re-ask.

Volunteered isn't the same as consented

The claim that zero-party data comes with consent built in doesn't hold up under European law. Someone answering your form has agreed to answer your form. Any later use of that answer is limited by what you said it was for at the moment you asked.

Article 13 of the GDPR says that where personal data are collected from the person themselves, you have to tell them at that moment the purposes of the processing and the legal basis for it. Article 5 adds purpose limitation: data collected for specified, explicit and legitimate purposes can't be processed later in a way that's incompatible with them. A budget figure given to get a tailored quote is not, on that basis alone, fair game for a retargeting audience.

That still leaves zero-party data in a better position than most. You know exactly where it came from, the person knew they were giving it, and stating a purpose up front is easy when you're asking directly. Put a line under the question saying what you'll do with the answer, keep to it, and you've handled most of what the law asks.

Making zero-party data change what happens next

An answer nobody acts on teaches the buyer that telling you things is pointless. The fix is to store answers in CRM properties that drive something.

Routing is the most direct use. A demo request that says "replacing our current tool within three months" can go straight to a senior rep, while "just exploring" goes to a lighter sequence. Answers to budget, authority, need and timing questions are the raw material of lead qualification, and having them before the first call lets a rep skip the basics and start where the buyer already is.

Messaging is the next. A stated problem is the best possible hook for personalization at scale, because you're repeating the buyer's own words back to them instead of guessing from their job title. Combined with data enrichment that fills in the firmographic data around the answer, you know both who the account is and what this person said they need.

The longest-term use is learning. Pull a year of declared answers from the deals you won and the deals you lost, and you'll see which problems your best customers named before they bought. That's evidence for your ideal customer profile, and it sharpens each buyer persona with the buyer's own wording.

Recap

Zero-party data is the cheapest data most B2B teams fail to use. It costs one well-placed question, and the answer tells you something no provider can sell and no tracking can infer: what this person says they want.

Getting value from it comes down to a few disciplines. Ask what enrichment can't find, keep the asking short, and give something back. Record where and when each answer came from, read it next to behavior, and let it expire. Say what you'll use it for, then stick to that.

Done well, declared answers are the most precise input your GTM engine gets: they decide who calls whom, what the first message says and which segments you go after next. They also make it safer to hand prospecting to automation, because the automation then acts on what buyers actually told you instead of a guess.

If your forms and call notes hold more than your CRM ever uses, we can design the questions, wire the answers into routing and scoring in your existing setup, and leave the properties and documentation with your team when we're done.

Frequently asked questions