A list of people who were researching your category three weeks ago is not a list of buyers. It’s a list of people who already picked someone else.
I am George, founder of Leadpipe. We refresh every intent record every 24 hours. That’s not a marketing choice. It’s a mechanical consequence of how buying decisions actually unfold, and it is the single biggest reason our intent data converts when stale intent data doesn’t.
The shelf life of an intent signal
A B2B buying cycle has two phases: the active evaluation window and everything else. The active window is when the buyer is researching, comparing, and shortlisting vendors. It’s when a cold email gets opened and a demo gets booked. Miss the window and the same buyer is dead to you for six to eighteen months.
The active window for most B2B categories is short. It depends on the price point and complexity, but the pattern is consistent:
| Category | Typical active window | Implication |
|---|---|---|
| SMB SaaS ($50-$500/mo) | 1-2 weeks | Weekly refresh already misses half the window |
| Mid-market SaaS ($5K-$50K/yr) | 2-4 weeks | Weekly refresh catches maybe one touchpoint |
| Enterprise platforms ($50K+/yr) | 6-12 weeks | Monthly refresh starts to work |
| Services and consulting | 2-6 weeks | Weekly is the floor |
If your intent data is refreshed quarterly, you are systematically selling to buyers who already bought somewhere else. You’re the vendor who shows up after the RFP is closed. The deal is done and you didn’t know.
This is why we built Leadpipe around daily truth. The manifesto has the long version, but the thesis is simple: the company that sees intent first, wins.
What “refresh” actually means
Refresh is a word vendors use loosely. When you read “refreshed weekly” or “refreshed monthly” in a data sheet, it could mean any of these:
| What they say | What they might mean |
|---|---|
| ”Refreshed weekly” | New signals ingested weekly, scores recomputed in the same window |
| ”Refreshed weekly” | Signals ingested continuously, scores published as a weekly export |
| ”Refreshed weekly” | The list of accounts is pulled from a 90-day rolling window once a week |
| ”Refreshed monthly” | The vendor runs one batch job a month against stored signals |
Only the first one is honest. The rest mean your “fresh” intent is built on signals that could be weeks or months old, and the vendor just rebuilt the CSV on a weekly or monthly schedule.
Leadpipe’s daily refresh means two things at once:
- Signals are ingested continuously. Every page view on the 5M-site Orbit pixel network is processed as it arrives.
- Scores are recomputed every 24 hours. Every person’s topic affinity, every audience, every watchlist is rebuilt from the most recent signal history.
Yesterday’s record is not today’s record. That’s the whole point.
What stale data looks like in practice
Four failure modes show up when intent data is not refreshed often enough.
The already-closed deal
A vendor exports a list of accounts researching CRM software. The list is two months old. Half the accounts have already signed with a competitor. Your reps call anyway, get told the deal is done, and blame the SDR team.
The stale topic
A company was researching “Salesforce alternatives” in March. By April, they’ve moved on to implementation and now need integration tools. Stale intent keeps selling them Salesforce alternatives. Fresh intent catches the topic shift.
The departed champion
The person showing intent at the target account left the company. They took their research behavior with them to a new employer. Your outbound to the old contact bounces; your rep never knows the champion is now at a different account shopping for the same thing.
The spike you missed
A cybersecurity incident drives a sector-wide surge in vendor research. Accounts that were cold last month are hot this week. Monthly refresh catches the tail of the spike. Daily refresh catches it the day it starts.
Each of these wastes your outbound capacity in a different way. The common thread: buyer behavior moves daily, and any intent cadence slower than daily is leaving signal on the floor.
Why daily, not hourly
A fair question: if daily beats weekly, does hourly beat daily? Sometimes, but not usually.
Intent is a noisy signal. A single visit to a category page is not enough to label someone as “in market.” It takes a pattern of repeat visits across multiple sites over a short window to separate real intent from accidental clicks. That pattern needs time to form.
Daily is the cadence at which:
- The pattern stabilizes. Single-visit noise washes out.
- Sales teams can act. Reps can’t react to hourly flickers. A daily watchlist fits the morning routine.
- The cost-to-value ratio breaks even. Sub-daily refresh doubles the compute cost without doubling the outcome.
Certain categories do benefit from faster-than-daily refresh: breaking events, launches, outages, security incidents. Those are edge cases. For steady-state outbound, daily is the right cadence. How Orbit works covers the aggregation pipeline that makes daily reliable.
The engineering tradeoff
Daily refresh sounds simple. It is not cheap. Every 24 hours, the system has to:
- Ingest the day’s new signals.
- Reclassify against the topic taxonomy.
- Re-cluster against the identity graph.
- Recompute every person’s topic score.
- Rebuild every saved audience.
- Republish to the API, the MCP server, and the dashboard.
Running this pipeline at 60B signals and 280M profiles, on a 24-hour cycle, without breaking the customer-facing SLA, is the infrastructure bet we chose to make. The cheaper alternative, which most competitors took, is to ingest continuously and publish monthly. That’s how you end up with “intent data” that’s built on old signal by the time anyone sees it.
For the category-level comparison with Bombora’s weekly cadence on account-level intent, see Orbit vs Bombora. For why person-level intent is a different animal from account-level, see person-level intent data explained.
How to evaluate a vendor’s refresh claim
If you’re shopping for intent data, ask these four questions. Vendors that refresh honestly can answer all four; vendors that don’t will get vague.
| Question | Honest answer looks like | Vague answer to watch for |
|---|---|---|
| When was the most recent signal in today’s export from? | ”Yesterday, up to 11pm UTC." | "Recent” or “in the past few weeks.” |
| How often is the score recomputed per person per topic? | ”Every 24 hours." | "Our scores are continuously updated.” |
| How long does a signal stay in the score? | ”We decay signals with a half-life of N days per topic." | "We use all available data.” |
| What’s the oldest signal that can influence a published score? | A number in days, like “30” or “60." | "Historical data is included.” |
The right answer to the first question, in particular, is a timestamp. If the vendor can’t point at a timestamp, the data is not as fresh as the brochure suggests.
The compounding effect of freshness
Freshness isn’t just about catching individual buyers. It compounds across the whole outbound motion.
A sales team with weekly intent data runs the same play every Monday: pull the new accounts, assign them to reps, send the sequences. By Friday, the intent is five days old and the conversion rate is falling. The team ends the week with stale targets and blames the messaging.
A sales team with daily intent data runs the play every morning. Today’s watchlist is based on yesterday’s behavior. The conversion rate doesn’t decay through the week because the list is refreshed through the week. Same rep, same messaging, higher yield.
The difference isn’t the data model. It’s the cadence. Daily refresh lets you run a daily motion. Weekly refresh forces you into a weekly motion, whether you want one or not.
For the sales-team angle on daily watchlists, see Orbit for sales teams and sales pipeline without SDRs.
Why we can offer daily refresh at self-serve prices
An honest question: if daily refresh is so much better, why doesn’t everyone do it?
The answer is cost. Running the full refresh pipeline every 24 hours is expensive. Signals have to be ingested continuously. The topic classifier has to run on new URLs daily. The identity graph has to re-verify records. Every saved audience has to be re-materialized. At the scale we run (60B signals, 280M profiles, 5M sites), that’s real compute.
The tradeoff most vendors make is to refresh less often and charge less. That works for them and breaks for you. What we chose instead: invest in the infrastructure to make daily refresh affordable at scale, then pass the cadence through to every plan tier.
The result is daily refresh on the $147/mo Pro plan. Same cadence on the $1,279/mo Agency plan. Same cadence on enterprise. The graph is a shared asset, which is the only way this math works.
What this means for customers
Leadpipe refreshes 280M profiles and 60B behavioral signals every 24 hours. Your audiences, watchlists, and intent scores are rebuilt every night. When you export a list today, it reflects what your buyers were doing yesterday, not last quarter.
That’s the whole point. Intent is a perishable asset. Build your outbound on fresh signal or accept that you’re running yesterday’s race.
