Most people in this industry don’t realize that every working day, roughly 13,600 B2B contacts change jobs. That’s just one data point that showed up in our signal data for the first half of 2026: 1,470,414 people who moved companies, plus another 194,165 who got promoted in place.
Multiply that across a year and you start to see the real shape of the problem sitting underneath every GTM stack. The ground is moving faster than the data most companies store about it. Whoever notices first gets there before the record catches up.
Starting a Business
I still smile remembering how this started. In 2016, Assaf Eisenstein, my co-founder, and I built a simple Chrome extension to help recruiters find contact details for candidates they needed to reach. It was a side project that we called Lusha, after the nickname of Assaf’s daughter. We noticed that the people who used it and then told about it to their friends were mostly sales reps so we structured it for them. We got feedback, iterated, kept improving.
Everything was so simple back then. Two years in, I was at the beach with my family on a Saturday when my co-founder called: Google wanted to buy licenses. That was the moment I knew we were onto something real.
The problem we were solving then was much smaller than the one we’re solving now. Back then, a good contact was one you could reach. That’s still true, but it’s no longer sufficient. The reason is the same one behind the 13,600-a-day number: B2B data decays constantly, and job changes are the single biggest driver of it. Every move breaks a record somewhere: a champion who goes quiet, a new decision-maker forming opinions about their new company’s stack, an email address that now belongs to someone else’s deal entirely.
What AI Changed for Us
A decade later, with 200+ Lushians in the team, 1M+ users, and AI fully in the picture, our product, our focus, and how customers consume our data are all shifting. We used to be a global database for prospecting, a place salespeople came to find accurate contacts. Now we are a data and intelligence layer for winning GTM teams and AI agents, built around relevance and prioritization instead of volume.
We ran our own data against the rest of the market to define, category by category, what good actually looks like in 2026. Start with connect rate, because the gap here is larger than most people expect. A verified direct mobile number connects at 12 to 18%. A general company line connects at 2 to 4%. Same rep, same script, same effort. The only variable is whether the number in front of them actually routes to the right person.
What Are the Benchmarks for B2B Contact Coverage, Signal Timing and CRM Data Accuracy?
Coverage tells a similar story. Our benchmark is three verified contacts per active deal, not one. Coverage is the difference between a deal that survives a champion going quiet and one that dies with them.
Signal timing is the newest category we measured, and the one I think matters most going forward. Every one of those 13,600 daily job changes opens a window: risk on one side, for whoever the person left behind, and opportunity on the other, for whoever’s about to earn their attention at the new company. Our data shows that window stays open for roughly 30 days before the new stack gets locked in. Our benchmark for a good team is acting on 80% of live signals inside that window. Most teams aren’t measuring this at all, which is itself the finding. You can’t hit a bar you’ve never checked yourself against.
Underneath all of this sits a category most teams still aren’t tracking: whether their CRM data is accurate in the first place. Some of the best revenue teams have a formal process for checking that. Most don’t know their own number. They’re guessing, and it shows.
How Do Job Changes and Data Decay Affect GTM Stacks and AI Agents?
Zoom out and the pattern across all four categories is the same: most GTM teams are running more tools than ever on a foundation nobody has verified. The average stack now runs 23 vendors deep, enrichment, scoring, signals, sequencing, AI SDRs, CRM, and most of what each one produces never reaches a rep because none of it sits on a shared, verified base.
Without a way to tell which accounts are actually worth calling, a huge share of outbound goes to accounts that were never going to buy. It’s a foundation problem, and it predates AI. AI just raised the stakes.
For years, the consumer of this data was a sales rep. Increasingly, it’s an AI agent, updating records, deciding what to work next, drafting personalized outreach, and when the data is wrong, the mistake happens at scale instead of one call at a time. Other findings include:
- 121,238 sales contacts changed companies in the first half of 2026, more than any other function. Each move creates three problems at once: a stale record at the old company, an unworked coverage gap on that account, and a narrow window at the new one before the person locks in their vendor list.
- 60,384 IT contacts changed companies in the same period, the second-highest function after sales. An IT departure doesn’t just go stale, it restarts a live vendor evaluation. The incoming leader inherits a decision they didn’t make and often just lets it quietly stall.
- 869 CIOs changed companies, more than any other C-suite role, averaging 8 moves a day. A new CRO moves the fastest of any executive once in seat: the sales stack review starts within two weeks of arrival, and vendor decisions, often including a data provider, get made within 30 to 60 days.
- For every contact who gets promoted in place, six change companies entirely.
Promotions are rarer but far easier to miss. The email address still works, so nobody notices the title and routing are wrong until the wrong rep has been calling the wrong seniority level for months.
We ran a live test on data decay itself: 23 of 25 fast-moving companies we tracked had at least one material change, in headcount, leadership, or funding, inside six months. A CRM snapshot taken in January was already wrong by summer for nearly all of them, and the most damaging changes never tripped a standard alert threshold.
What Does Good B2B Sales Data Look Like for Revenue Teams in 2026?
The teams pulling ahead right now are feeding those models with the right data and context shaped by their own business: their wins, their pipeline history, their specific buyers.
Ten years ago, good data meant a contact I could reach and hear “yes” from.
Today it means a system that tells a rep, or increasingly an agent, who to call and why they’re relevant right now, using predictive models and machine learning. That’s the bar. Most teams haven’t measured themselves against it yet.
Yoni Tserruya is the CEO and co-founder of Lusha. He co-founded Lusha in 2016 and bootstrapped the company for its first 4.5 years before raising venture funding. Ten years later, Lusha has grown to 200+ employees and 1M+ users.
