Skip to main content
Announcements

Autobound + 5x5: The Behavior Layer for an Intent Engine That Shows Its Work

We're combining 5x5's Market Pulse behavioral data and identity graph with Autobound's signal graph and closed-won outcomes to build intent that finally shows its work.

·5 min read

Key Topics Covered

  • Why 5x5
  • Where 5x5 plugs in
  • Signals plus outcomes: the predictive engine
  • A two-way street

Article Content

B2B teams have never had more buying data, and never had a harder time acting on it. Too often, what reaches a rep is a bare score with no reason attached: a topic spiked, good luck. Turning buyer behavior into predictions people trust takes two things that rarely live together - a deep, verifiable stream of what buyers are doing, and the outcome data proving which patterns end in a deal.

Today we’re putting both under one roof. Autobound is partnering with Five by Five (5x5), the member-driven data co-op behind Market Pulse, one of the largest behavioral datasets in B2B. 5x5’s person, company, and intent data now power core layers of the Autobound signal graph, and our intent products are live in the Signal API and embedded where 5x5’s members already work.

Autobound
×
Five by Five
The behavior layer meets the answer key.

“Everyone in this market is sitting on signals. Almost nobody has the answer key. 5x5 gives us the deepest behavioral dataset we’ve ever plugged into the graph, and our closed-won co-op supplies the outcomes to learn from. That combination is how intent finally shows its work.”

Daniel Wiener
Daniel Wiener
Co-founder & CEO, Autobound

Why 5x5

5x5 is built differently. It’s a data cooperative: members across marketing, advertising, sales, HR, finance, and fraud contribute data into a shared graph, and every contribution makes the whole thing more accurate - a self-healing dataset, continuously validated by the members who use it.

Market Pulse, by the numbers
6B+
observed signals per day
43K+
intent topics, 36K+ of them B2B
500K+
domains tracked, 7× the closest competitor
310M+
Universal Person IDs

Two things sold us. First, identity: Market Pulse signals resolve to hashed emails and Universal Person IDs, so intent ties to real people and companies instead of loose account-level guesses. Second, honest scoring: every account’s composite score is benchmarked against its own 12-week baseline, with Early and Late stage labels drawn from the content buyers are actually researching - and billions of new observations land every day. Deep coverage, real identity, fresh behavior. That’s the layer we wanted.

Where 5x5 plugs in

1
The resolution graph5x5’s person and company data strengthens our general resolution graph - the identity spine that ties 1B+ buying signals to the right company and person across 50M+ companies and 250M+ contacts. It’s the layer that figures out the company domain of a business mentioned on a podcast, or who a speaker on a podcast or an earnings call actually is. Signals only matter if you know who they belong to.
2
Buyer intent, live in the Signal APIMarket Pulse powers our buyer intent endpoints today. Search companies and contacts by topic, pull account timelines, and combine intent with hiring, funding, and tech adoption in one API call.
3
The Intent InitiativeWe’re building a predictive engine that uses Market Pulse behavior in conjunction with real closed-won outcomes to learn what buying actually looks like. More on that below.

Signals plus outcomes: the predictive engine

Earlier this month we opened the Intent Initiative, our program to build a signal-native intent engine trained on the one dataset this industry has never had: real closed-won and closed-lost outcomes. It’s early, and we’re building it with design partners. Signals are the exam. Outcomes are the answer key. Without the answer key, everyone is guessing which patterns matter.

Everyone has signals. Almost no one has the answer key.

We didn’t just theorize this. We backtested it on 82.8M signals across 31 types and 14 months of history, training gradient-boosted models - auditable by design - to predict which companies would adopt a software category within 90 days.

The backtest, in three numbers
4–7.6×
more buyers in the top decile vs. random
0.97
peak AUC on category models
2–4×
lift over a generic baseline

Tech adoption was one label, and it proved the method. Now we’re building the engine around the highest-value label of all: what actually closed. As contributors share anonymized won/lost outcomes under mutual NDA, the engine learns the buying fingerprint behind real deals so it can flag the companies exhibiting it next.

Market Pulse changes the math. Our graph already knows who’s hiring, who raised, and what leadership said on the earnings call. Market Pulse adds what those same accounts are researching, resolved to the same people through the same identity spine. Using that behavior in conjunction with closed-won outcomes gives the engine more predictive surface area than any single-source product can offer - and every prediction traces back to the signals behind it. No black box. Breadcrumbs all the way down.

There’s a symmetry here we love: 5x5 built a co-op for behavior. We’re building a co-op for outcomes. Together, that’s the full loop.

Signal Data API

Turn these insights into pipeline

700+ real-time buying signals from 35+ sources. Know exactly when prospects are ready to buy.

A two-way street

The value flows both directions. Autobound’s signal data is beginning to flow into the 5x5 ecosystem too - putting hiring, funding, tech adoption, and hundreds of other signal types to work for 5x5’s members. More on that soon.

“We built 5x5 for one purpose: powering data-driven product companies. Autobound is exactly the kind of member the co-op was designed for. They’re pairing Market Pulse behavior with real outcomes, and as their signal data flows into the co-op, every member gets stronger. That’s the model working exactly as it should.”

Nick Weldon
Nick Weldon
Founder & CEO, 5x5

What you can do today

  • Use the intent product now. Buyer intent - company and contact level - is live in the Signal API, alongside 700+ signal types.
  • Join the Intent Initiative. If you have closed-won data, you have the answer key. Contribute it and get a map of who’s about to buy in your market, plus design-partner access as we build.
Help build the intent engine that shows its work
Contribute anonymized closed-won data. Get a map of who is about to buy in your market.
Explore the Intent Initiative →
Two minutes to apply. Mutual NDA before any data moves.

The takeaway

Intent doesn’t have to be a score you take on faith. Pair the deepest behavioral dataset in B2B with a signal graph that spans everything else a buyer does, train it on deals that actually happened, and every prediction can show its work.

Signals, outcomes, identity. That’s the engine we’re building. 5x5 just made it a lot stronger.

Signal Data API

Turn these insights into pipeline

700+ real-time buying signals from 35+ sources. Know exactly when prospects are ready to buy.

Frequently Asked Questions

What is 5x5's Market Pulse?

Market Pulse is the behavioral intent dataset from 5x5, a member-driven data cooperative. It observes 6B+ signals per day across 43K+ intent topics, resolves them to hashed emails and Universal Person IDs, and scores each account's topic interest against its own 12-week baseline with Early and Late buying-stage labels. Autobound uses Market Pulse to power buyer intent in the Signal API.

What does the Autobound + 5x5 partnership include?

Three layers: 5x5's person and company data strengthens Autobound's resolution graph (tying signals like a podcast mention or an earnings-call speaker to the right company and person), Market Pulse powers the buyer intent endpoints in the Signal API, and both feed the Intent Initiative - a predictive engine being built with real closed-won outcomes. Autobound signal data is also beginning to flow into the 5x5 ecosystem for their members.

What is the Intent Initiative?

The Intent Initiative is Autobound's program to build a signal-native intent engine trained on real closed-won and closed-lost outcomes contributed by design partners. In backtests on 82.8M signals, category-specific models surfaced 4-7.6x more actual buyers in the top decile of predictions than random targeting, with peak accuracy of 0.97 AUC. Teams with closed-won data can apply to contribute and get design-partner access while it's built.