Outbound first vs inbound first
four tests, written by people who sell one of them

I sell outbound for a living, so treat what follows with the appropriate suspicion and then check it against the tests. The reason this page exists at all is that outbound is genuinely the wrong instrument for a large share of the companies who ask us for it, and saying so after they have signed is a worse business than saying so before.

The framing that makes the argument resolvable is financial rather than philosophical. Outbound buys pipeline now at a roughly linear cost and stops producing the day you stop paying. Inbound builds an asset that produces later at a declining marginal cost and keeps producing after the spend ends. Those are different instruments with different risk profiles, and you choose between them on timing, on certainty, and on what you can afford to wait for.

The comparison that matters

Test 1: is your market big enough for outbound, and smaller than you think?

Outbound is viable when your addressable market is large enough to sustain the volume the arithmetic requires, and small enough that you can prioritise it without demand signals. Both halves of that sentence do work.

Run it backwards. If your model needs 10,000 meaningful sends a month and your realistic buying window means you can touch a company about once a quarter, you need on the order of 30,000 addressable companies to sustain the motion without burning the list. Below that, outbound works for a while and then runs out of market, which is a specific failure mode that looks exactly like declining performance and is actually list exhaustion. Above roughly 50,000 target accounts the problem inverts: you can send forever and you cannot tell which of them are in market, which is precisely where inbound and signal based targeting earn budget.

The counterintuitive half is that most teams get the size wrong in the small direction. Across the accounts we scope inside the group, clients underestimate their addressable market by 10 to 50x with striking regularity, because the market was built from the existing customer list and therefore describes who has already found them. Rebuild it once with your two least certain assumptions removed before you conclude outbound has no room.

Test 2: does your deal size support a human sales process?

Outbound arithmetic closes at roughly $10K in deal size and above and gets progressively harder below it. The mechanism is not subtle. Outbound has a real cost per meeting held, that cost is roughly fixed regardless of what you sell, and below some threshold it exceeds what the deal can carry over a payback period you would accept.

Work it in your own numbers. Take your fully loaded monthly channel cost, divide by meetings held rather than booked, and that is your true cost per meeting. Multiply by the meetings it takes to create an opportunity, then by the opportunities it takes to close one deal. Compare against your gross margin and your payback tolerance. If it does not clear, outbound has not underperformed and nobody executed it badly. It is the wrong instrument for your price point, and better copy does not move a unit economics problem. We publish no pricing figures anywhere on this site, so the cost inputs are yours to supply.

Test 3: how long can you actually wait?

The outbound ramp is not a soft estimate and it is the number most plans omit. Onboarding runs around 21 days for discovery, segment definition, infrastructure and list build. Domain warm up runs 4 to 6 weeks before you can send at real volume without wrecking sender reputation. Those overlap partially, which puts first meaningful volume near week six to eight and first pipeline shortly after. Anyone promising outbound pipeline in week two is either sending on infrastructure that will not survive the quarter or counting a meeting that has not happened.

That ramp is also where the industry loses people. Monthly client churn across this category runs in the 3 to 5% range, and a meaningful share of it is buyers cancelling during a ramp they were told about and did not price in. The first two months of a bad programme look identical to the first two months of a good one, which is why the review window has to be set past warm up before anything is signed.

Inbound is slower and it does something outbound structurally cannot, which is compound. An article that ranks or gets cited by an answer engine keeps producing without further spend, and the marginal cost of its tenth month is zero. That asymmetry is the strongest argument for starting inbound early even when it cannot be your first pipeline source, and it is why outbound first almost never means outbound only.

Test 4: can you hold the meetings you book?

The last test is skipped most often and is the cheapest to run. Outbound converts spend into meetings, and meetings convert into pipeline only if somebody competent runs them, on time, with a process behind them. Where calendar discipline is broken, booked meetings die at roughly a 50% show rate. That single number halves the return on the entire channel and it is invisible in every metric upstream of it.

Confirm four things before funding either motion, because inbound leads die on a bad calendar too. There is a named owner for every booked meeting. Reminders and reschedules are automated rather than remembered. The first call has a written structure. No shows are chased rather than written off. A team that cannot hold meetings will conclude the channel does not work, when what did not work was the calendar.

Where outbound first wins

Outbound first wins whenever the constraint is time and the market is nameable. You can be in front of a specific buyer at a specific company this quarter without waiting for them to develop an interest in searching. You control the volume, so the plan has a lever you can actually pull. And the measurement is fast and unambiguous when the fraction is stated properly: positive replies divided by emails sent, per sequence, after warm up, over a window agreed before launch.

It wins a second time on learning rate. Outbound is the fastest paid instrument for finding out which segment and which message land, because the feedback loop is days rather than quarters. Teams often buy outbound for pipeline and get more value from what it tells them about their positioning, which is a real return that no attribution model records.

Where inbound first wins

Inbound first wins when your buyers are already looking. If a category has established search behaviour, intercepting demand costs less per opportunity than manufacturing it, and the resulting conversations start further along. It wins again when deal size cannot carry outreach cost, which is not a marginal case. Below the point where cost per meeting held exceeds what the deal supports, outbound is not a channel to optimise, it is a channel to decline.

It wins a third time when the market is too large to prioritise. With a very large addressable set, the binding constraint stops being how many companies you can reach and becomes which of them are in market this month. Demand signals answer that and outreach volume does not.

And it wins on what it leaves behind. Outbound produces revenue that is visibly contingent on continued spend. Inbound produces an asset that keeps working through a budget cut and through a change of team. If you are building toward a transaction or toward reduced dependence on a single channel, that difference is worth more than a quarter of faster pipeline.

Who should pick which

Outbound first: deal size clears roughly $10K, you can name fewer than about 50,000 target accounts, you need pipeline inside a quarter, and your calendar discipline is real. Start the inbound asset build in parallel at whatever cadence you can sustain, because it cannot be backdated.

Inbound first: buyers actively research your category, or your deal size cannot carry per prospect cost, or your market is large enough that prioritisation matters more than reach. Accept the timeline honestly rather than expecting an early exception, and fund it to a cadence that will actually get indexed.

Run both from day one if you can fund both at honest scale. Half funding two motions produces two sets of inconclusive data and no decision, and it is the most common way a year gets spent without learning anything. Fund outbound to the volume the arithmetic requires or do not start it.

Fund neither if you have no closed deals yet. Every model behind this page needs a win rate and an average deal size from your own record, and estimated inputs produce false precision that gets quoted in a board meeting. The correct spend at that stage is founder led selling until fifteen to twenty opportunities have run their course and told you what your rates are.