Every intent vendor ships a score. Most hide the inputs. You get a 0 to 100 number with no definition, no band documentation, and no way to tell whether 70 means “casually curious” or “requesting a demo next week.” Buyers learn to distrust the number, so they ignore it.
I am George, founder of Leadpipe. This post is the documentation the vendors never write. It explains exactly what an Orbit score of 80 represents, what the bands mean, where the thresholds break, and how to set your floor so the audience surfaces buyers instead of browsers.
The short answer
An Orbit score of 80 on a topic means the person has shown multi-session, multi-page, cross-site research behavior on that topic within the last 30 days, weighted toward recent and high-intent page types. In practice, that looks like 6 to 15 pageviews across 2 to 5 distinct sites, with at least one pricing page or head-to-head comparison page, and at least two separate sessions. Below 70 is ambient noise. 70 to 84 is meaningful research. 85 to 94 is late-stage. 95 to 100 is a buyer who is almost certainly about to decide. The default audience floor is 70. Raise it for precision, lower it for volume.
Now the details.
What goes into the score
The Orbit score is built from five inputs, each weighted and decayed.
| Input | What it measures | Weight |
|---|---|---|
| Page count | How many on-topic pages the person has viewed | Moderate |
| Dwell time | Cumulative time spent on on-topic content | Moderate |
| Site diversity | Number of distinct sites where the person showed on-topic behavior | High |
| Page type | Pricing, comparison, and “alternatives” pages weighted heavier than top-of-funnel | High |
| Recency | Last 7 days weighted heaviest, 30+ days decayed | High |
Page count alone is cheap. A person who loops on a single blog post all afternoon is not a buyer. Cross-site, cross-page, cross-day behavior with recent pricing-page activity is a buyer. The scoring math is built around that distinction.
Every topic is scored independently. The same person can be a 90 on “CRM Software” and a 32 on “Marketing Automation” on the same day. You score the person against the topic, not against some global intent measure.
The five bands
Here is what each band actually looks like.
0 to 49: ambient
One pageview, maybe two. No cross-site behavior. Probably an incidental visit from search or social. Not intent. Orbit keeps the signal so it can upgrade later if behavior continues, but the person does not appear in any default audience.
What to do: Ignore. These never populate a watchlist unless you drop thresholds to zero (which you should not).
50 to 69: mild
Two to four pageviews, usually one site, typically one or two sessions. Someone reading a couple of articles in a category. This is early Phase 1 buyer behavior, but it is indistinguishable from casual curiosity. Content readers, students, prospects who abandon, competitor researchers, and real buyers all live here.
What to do: Useful for brand-level retargeting. Not useful for direct outbound. If you lower an audience floor to 65, expect a 2x to 3x volume increase with proportional noise.
70 to 84: meaningful research
This is the default audience floor and where most of the value lives. Four to ten on-topic pageviews, two or more sites, typically two or more sessions over several days. Pricing or comparison page activity usually appears somewhere in this band.
A score of 80 specifically is a person in the middle of active evaluation. They have visited more than one vendor, spent real time on on-topic content, and have come back at least once. Most B2B buyers live in this band for 5 to 15 days of their cycle.
What to do: Outbound. Slack alerts. Personalized sequences. SDR priority queue. LinkedIn retargeting. Direct AE touches for named accounts.
85 to 94: late-stage
Cross-site behavior on at least three distinct vendor sites, repeated pricing-page visits, comparison content inside the last week. This is a person who is building or has built a shortlist.
What to do: Personal founder-level outreach for enterprise deals. Same-day response from SDRs. If the person is at a top-tier target account and you are not already in the conversation, you are behind.
95 to 100: late shortlist / ready to buy
Repeated high-weight page activity (pricing, demo, “alternatives”, head-to-head) over the last 3 to 7 days, often on multiple competitor sites including yours. The person is probably in a buying process right now, including one you are not in.
What to do: Treat as a live deal signal. Page the account owner. If you are not in the evaluation, this is your last intercept window.
Visualizing the distribution
For a typical B2B category, the daily active audience on a well-chosen topic distributes roughly like this:
Score 0-49 ██████████████████████████████████████████ ~65% of touched people
Score 50-69 ██████████████████████ ~22%
Score 70-84 ██████████ ~10%
Score 85-94 ██ ~2.5%
Score 95-100 █ ~0.5%The top 13 percent of the distribution (anything 70 and above) is where almost all real buying behavior lives. Below 70 is noise for most use cases.
Why 70 is the default floor
We picked 70 as the default because it is the point where a second confirming behavior almost always exists. One pageview and five minutes of dwell rarely cross 70. Two sessions across two sites almost always does.
Most false positives live below 70. Most real buyers live above 70. The band is not arbitrary; it is where the signal-to-noise ratio flips.
If your business needs volume (advertising audiences, broad awareness campaigns), you can drop to 60. If your business needs precision (named-account outbound, founder-led outreach), raise to 80.
The score against time
Scores decay. A person who hit 88 two weeks ago and has had zero on-topic activity since will be at 62 today and 40 in another two weeks. This is a feature, not a bug. Stale intent is not intent.
Every time we rematerialize an audience (every 24 hours), each person’s score is recomputed against the latest 30-day window with recency weighting. That is why daily refresh matters. We cover this at length in why Orbit refreshes intent daily.
The upside of decay: you never have to “clear” a stale list. The output is always fresh by construction.
How this compares to other vendors’ scores
Most intent platforms have a scoring model. They rarely document the inputs. Here is a generalized view of the landscape.
| Vendor | Score shape | What it represents | Refresh |
|---|---|---|---|
| Bombora Company Surge | 0 to 100 | Company-level surge above baseline | Weekly |
| 6sense account intent | A/B/C/D or 1-100 | Multi-source blended account score | Varies |
| G2 Buyer Intent | Category and product views | Review-site-only behavior | Daily |
| Clearbit Reveal / HubSpot Breeze | Fit + intent score | Firmographic fit plus on-site activity | Continuous on site |
| Orbit | 1-100 per person per topic | Cross-site, page-type-weighted, decayed | Daily |
Bombora’s surge score is a legitimately well-defined signal at the company level. The limitation is that it is company-level. You know Acme is surging; you do not know which of Acme’s 500 employees actually cares.
6sense and Demandbase blend many signals into account scores. Useful for enterprise ABM orchestration. Different use case from person-level outbound.
G2’s intent is strong if buyers review their shortlist on G2 (many do). Much narrower pixel footprint than a cross-site network.
Orbit is per-person, per-topic, daily. That means you can set a floor of 80 on “CRM Software” and get a list of people (not companies) researching that category at that strength right now.
For a deeper side-by-side, see Orbit vs Bombora and can Orbit replace Bombora.
Tuning the score for your use case
Pick your threshold based on the downstream action.
| Use case | Recommended minScore | Why |
|---|---|---|
| SDR outbound to target accounts | 70 | Standard meaningful-research floor |
| Executive/founder outreach | 85 | High signal, low volume, manual action |
| LinkedIn Matched Audiences retargeting | 60 | More volume, ads forgive noise |
| ABM watchlist of 200 domains | 70 with minTopicOverlap: 2 | Depth via overlap, not just score height |
| Competitive intercept | 80 on competitor topics | Cleaner signal on branded competitor terms |
| Content strategy / trend spotting | 50 and look at trend direction | Use movers and topic trends, not audiences |
If your audience is too big, raise the floor by 5. If it is too small, drop by 5 or loosen an ICP filter (not the score).
What a score of 80 tells you, concretely
A person showing 80 on a topic today is a person who:
- Has been active on this topic in the last 7 to 14 days.
- Has viewed 6 to 15 on-topic pages across at least 2 sites.
- Has at least one high-weight page (pricing, comparison, alternatives) in their history.
- Is in an active research phase but may not yet have a shortlist.
That is the profile you want an SDR reaching out to. Early enough to influence the shortlist. Late enough that the person is genuinely a buyer, not a researcher.
Where to go next
If you want the underlying topic anatomy, read what’s inside an Orbit intent topic. If you want the early-detection timeline the score is tuned against, read how early Orbit detects in-market buyers. For the category fundamentals, the buyer intent glossary and in-market buyer intent cover the ground.
Orbit resolves intent at the person level against a deterministic identity graph — the difference between “an account is in-market” and “this director at this account is researching today.” Try Orbit →
