Cold outbound starts from a list and works it. Intent-based outbound waits for a signal and acts on it. The real difference is not reply rate, it is volume: a signal-led motion can only contact people who did something, which caps your reachable market and changes how you staff it.
Disclosure: Signl sells signal-led prospecting, so read the recommendation accordingly. The case against it is made properly below, including the ceiling that makes it a poor fit for some teams, because a comparison with no honest concession is an advert.
Key facts
- Both models are list-building. They differ in what puts someone on the list.
- The published performance claims for intent-led outbound contradict each other and carry no methodology.
- Signal-led outbound has a hard volume ceiling that list-led outbound does not.
- The two need different operating rhythms and do not combine well inside one sequencer.
- Neither model fixes a product nobody wants, which is the usual actual problem.
The two models, structurally
Cold outbound defines who should buy, builds a list of everyone matching, and contacts them on a cadence. The organising assumption is that in any sufficiently large group of people who fit, some proportion happen to need you this month, and volume finds them. It is a search problem solved by brute force.
Intent-based outbound watches for an observable event, and contacts the people attached to it. The organising assumption is the reverse: that timing is the scarce variable, so the job is to detect it and move quickly. It is a search problem solved by waiting in the right place.
Both end up with a list and a message. The difference is upstream.
| Cold outbound | Intent-based outbound | |
|---|---|---|
| Starting point | Filter criteria | An observed event |
| Who you contact | Everyone who fits | The few who acted |
| Scales by | Adding volume and sending capacity | Adding sources and watching longer |
| Scarce resource | Deliverability and domain reputation | Qualified events |
| Fails when | The list is undifferentiated and the market is saturated | Your buyers do nothing observable |
| Message material | Role, company, a researched line | What the person said or did |
| Operating rhythm | Build, sequence, measure in cohorts | Read daily, write individually |
| Honest ceiling | Your domains and your patience | The number of signals your market produces |
What the published numbers actually say
This is the part most comparisons skip, and it matters because the numbers are the main reason people adopt the model.
Reading the vendor pages ranking for this topic on 25 October 2026, claimed reply rates for intent-led outreach ran from 30% to 45%, against cold baselines quoted as under 1% in one place and 1% to 3% in another. One page put cost per qualified call at $83 against $1,300 for traditional methods. None of them published a sample size, a time window, a definition of "reply", or who ran the test.
Those claims are not comparable with each other, let alone with your business. A 45% reply rate on 20 contacts and a 2% reply rate on 5,000 are different businesses, and most of the pages quoting the first number are selling the tool that produces it.
We publish no reply rate for this motion. We have not measured one across customers in a way we could reproduce, and a number we cannot reproduce is worse than no number. What we can say is structural, and that is what the rest of this article is.
The volume ceiling
Here is the limit nobody selling signal-led outbound puts on the page.
You can only contact people who did something you can observe. In a given month, that is a small and fixed fraction of your market. If you watch 25 sources and they produce 60 people who pass an ICP check, your pipeline from this motion is 60 people, no matter how good the messages are. Adding effort does not add people; adding sources does, slowly, and only while good sources exist.
Work it backwards before you commit. If you need 200 conversations a quarter at a 10% reply rate, you need roughly 2,000 contacts. Signal-led sourcing will not produce 2,000 qualified events a quarter in most niches. That gap is the honest answer to whether this model can be your only one.
The flip side is the thing list-led outbound cannot buy: every person on a signal list did something, which means every message can answer something real. The maths of a smaller list that converts better is good right up until you need the absolute number to be large.
Where each one fails
Cold outbound fails when the market has been worked. Everyone in a defined segment has heard from eleven companies this quarter, reply rates fall, and the response is usually more volume, which accelerates the decline. It also fails on deliverability: the model's scaling mechanism is the thing that gets your domain flagged.
Intent-based outbound fails when your buyers are silent. Plenty of markets do not discuss their problems in public: plant managers, hospital procurement, most of the public sector. No amount of listening finds a signal that was never emitted. It also fails when the team treats signals as a list and runs them through a sequencer, which discards the only advantage the model has.
A third failure is shared and more common than either: the product has no urgency attached to it. Signals tell you someone has a problem. If your product is a nice improvement rather than a fix for something that hurts, a perfectly timed message still gets ignored.
Which to run
| Your situation | Start with |
|---|---|
| You need large absolute volume this quarter | Cold outbound, with signals layered on later |
| Your buyers are active on LinkedIn or in public communities | Intent-based |
| Your market is small and you have already worked it | Intent-based, because volume is not available anyway |
| Your buyers are not online in any observable way | Cold outbound, and reconsider the channel |
| Deliverability is already a problem | Intent-based, which is low volume by nature |
| You have no clear ICP yet | Neither. Fix that first, or both models fail identically |
| Deal sizes are large and accounts are named | Both, with account-based listening on the named list |
The last row is the one most mid-market teams land on, and it is covered in account-based social listening.
Running both without ruining both
The common mistake is operational rather than strategic: teams buy a signal tool, pipe its output into the sequencer that runs the cold motion, and wonder why it performs like the cold motion.
Keep them separate in three places.
Different queues. Signal people get read individually. List people get sequenced. A person who appears in both should be removed from the list motion, not messaged twice.
Different volumes. Forty considered sends a week from signals is a healthy number. Judging that queue by the volume standards of the list motion kills it.
Different measurement. Cold outbound is measured in cohorts and conversion percentages. Signal-led work is measured per source: which sources produced conversations, which produced nothing. The model for that is in how to measure social listening ROI.
Where the signal is public engagement specifically, the ordering that keeps it cheap is in qualify before you buy an email, and the strength of each kind of signal is in LinkedIn buying signals. The taxonomy of signal types, including the ones you buy rather than observe, is first-party vs third-party intent data.
Limits
- No model fixes positioning. If the reply is "we do not have that problem", neither timing nor volume is the issue.
- Signals are public, so your competitors see them too. On the strongest signals you are in a race, not a private channel.
- Attribution is harder for signal-led work, because the touch that mattered was often a public reply nobody logged.
- The compliance questions are identical for both. A signal is not consent and public visibility is not a lawful basis; see GDPR and B2B prospecting. This is not legal advice.
