B2B Intent Data: How to Act Within 48 Hours

Key takeaways

• Intent data is a timing tool. A signal that sits in a dashboard for a week is just a more expensive list.
• Pick two or three signals tied to your ICP, score them, and decide in advance who contacts whom and with what message.
• Act within 24 to 48 hours, keep the first touch relevant to the signal, and review which signals actually produce meetings.

Your marketing team buys an intent data subscription. A dashboard fills up with accounts “surging” on your topics. Two weeks later, nobody on the sales team has opened it, and the renewal conversation gets awkward.

That’s the usual failure with B2B intent data. The data is fine. The workflow around it doesn’t exist. This post lays out a simple signal-to-action process you can run inside 48 hours of a signal appearing, without a big tooling project.

What intent data actually tells you

Intent data says an account is more likely to be in a buying window than a random account on your list. That’s all. It isn’t a lead, and it isn’t a promise.

There are three kinds worth knowing:

• First-party: your own website visits, pricing page views, content downloads, product usage.
• Third-party: research activity across other sites, grouped by topic and account.
• Event signals: hiring for a relevant role, a funding round, a leadership change, a new tool appearing in their stack.

First-party is the most accurate and the cheapest. Most teams skip straight to buying third-party data while ignoring what’s already in their analytics.

Why speed matters more than the data source

Vendor research from Landbase says 97% of marketers value intent data, yet only about 25% actually use it, and it recommends acting within 24 to 48 hours. Treat that as vendor research, not gospel, but the gap matches what we see: plenty of people own the data, few have a process.

Speed works the same way for inbound leads. The same logic is behind our post on speed to lead: interest cools fast, and the first relevant vendor in the inbox gets the conversation.

The signal-to-action workflow

Here’s the loop we’d set up for a 30-person sales team. Nothing exotic.

Each step needs an owner and a rule, otherwise it stalls.

1. Choose signals you can defend

Start with two or three. For a team selling a sales engagement tool, that might be: three or more visits to your pricing page from one account, a new VP of Sales hire, and a job post for an SDR manager. Each one points at a real reason to talk.

More signals means more noise. Add them only after the first ones prove they convert.

2. Score the account, not just the signal

A signal from an account outside your ICP is a distraction. Check fit first, then add the signal as a multiplier. If you don’t have a scoring model yet, our B2B lead scoring model post covers how to build one.

3. Enrich the right people

The signal tells you the account. You still need the person: the buyer, the likely champion, maybe one more. Use verified contact data and pull from more than one source if coverage is thin. We cover that in waterfall enrichment.

What the first touch should say

Reference the reason, not the surveillance. Nobody wants to hear “I saw you read our pricing page four times.”

Here’s an illustration (not a real message). The signal is a new VP of Sales hire at a 200-person SaaS company:

“Congrats on the new role. Teams in the first 90 days usually inherit a pipeline they didn’t build. Happy to share how a couple of peers audit theirs in the first month, if useful.”

It’s tied to the event, offers something, and doesn’t pretend to know more than it does.

Autonomous blasts don’t do this well. Vendor research from Laxis puts replies from autonomous AI SDR outreach at 1 to 3%, against 15 to 25% for supervised, signal-triggered outreach. Treat the exact numbers cautiously, but the direction is clear: a human-reviewed, signal-based message beats volume.

Set the 48-hour rules in advance

The reason signals go stale is that every decision gets made live. Decide these once, in writing:

If the owner is on leave, the account should route to someone else. Otherwise the 48 hours quietly becomes five days.

Where AI helps, and where it doesn’t

AI is good at the grunt work: watching for signals, pulling account context, drafting a first version of the message. A research agent can assemble that context in minutes instead of a rep spending 20 minutes per account.

It’s weaker at judgement. Someone still needs to decide whether the signal is real, whether the timing is awkward, and whether the draft sounds like a human wrote it. Keep a person in that seat.

Measure the right thing

Don’t judge the workflow by signals detected or emails sent. Track:

1. Meetings booked per 100 signal-triggered accounts.
2. Time from signal to first touch.
3. Which signal types produce meetings, so you can drop the weak ones.

After a month you’ll know which signals earn their place. Cut the rest.

The bottom line

Intent data is only useful if it changes what your team does this week. Pick a few signals you trust, set the rules before the signals arrive, and make the first touch within 48 hours. If you’d like help designing that workflow, a free 30-minute consultation is the place to start.

Sources

• Landbase, intent data tools for RevOps teams (vendor research): https://www.landbase.com/blog/intent-data-tools-revops-teams
• Laxis, state of AI SDR 2026 (vendor research): https://laxis.com/blog/state-of-ai-sdr-2026