Job ads or direct approach: not a duel, a sequence.

// 01Inbound first, always

The job ad is not the enemy — it's one of the attraction channels you should lean on, and often the main channel companies use. The real strategy, the one that pays long term, is to maximise every inbound channel — ads, career page, brand, network — so that good direct applications come in by themselves. That is work in its own right: in Module 1, we go through these channels one by one, with what can be maximised on each.

If your roles fill well through ads today, keep going — and invest so they fill even better. You don't need to turn your team into a sourcing unit for roles the market comes to fetch on its own.

// 02Where inbound stops

A job ad captures demand that already exists: people searching, scrolling job boards, applying. Direct approach goes after people who weren't looking for anything. On rare profiles — senior, R&D, deeptech, structurally scarce roles — the demand that exists is close to zero: these profiles are employed, solicited constantly, and don't read job ads. The queue in front of your storefront simply doesn't exist, whatever the advertising budget.

That's the pedestrian street image, PAT's founding metaphor: you can polish your storefront and wait for passers-by to walk in — or go walk where the profiles actually circulate. On rare roles, the flow of passers-by in front of your storefront is near zero, no matter how good the storefront is.

Add a perverse effect anyone who has opened a sharp role knows: on a rare role, the ad mostly attracts off-target applications. You don't gain candidates, you gain screening work. When inbound, even maximised, doesn't bring the volume or the quality you need — then you go outbound: sourcing.

// 03Sourcing: technical, time-consuming, not scalable

When you have to go, take the time to do it properly. Sourcing is technical, time-consuming and not scalable — that's its honest cost, and it's also what makes it defensible on a critical role: nobody else does it seriously.

• A real search competence. Finding is a craft. Recruiter badly used produces shallow pools — the same 30 profiles everyone contacts. Used well, what we observe with trained teams: talent pools 3 to 10× deeper on existing searches.
• Clean framing upstream. A sloppy intake means weeks of searching beside the target.
• An outreach message worthy of the name. Done badly, direct approach becomes spam — and on a small market where everyone knows everyone, it's your brand that burns.

It's an investment in team competence, not media budget. It survives turnover and improves with practice — that's its real economic difference with the ad, which you pay again for every role.

// 04The false middle: "post and pray, improved"

The classic trap: post the ad, then send three template InMails to the first profiles that show up, "to activate sourcing". It's the worst of both worlds — neither maximised inbound, nor serious outbound. The sequence works when each side is done properly for what it does best: inbound captures the active flow, sourcing covers the invisible market.

// 05The verdict, situation by situation

Your situation	The sequence that works
Large talent pool, active candidates	Well-maximised inbound — sourcing would be a luxury
Common technical profiles, tight market	Maximised inbound + targeted sourcing on the short-list
Rare profiles, senior, R&D, deeptech	Sourcing is indispensable — ads don't reach them
Confidential search or replacement	Sourcing only
Very strong brand on your segment	Inbound first, sourcing as an exhaustivity complement

// 06And AI changes (almost) nothing

AI speeds up both sides — writing, screening, synthesis. But despite the promises of automated systems "guaranteeing it's handled", a bad signal amplified over an incomplete database simply cannot work. It will work at the margin — and you will never get a guarantee of exhaustivity.

For exhaustivity, you will have to do it yourself, and understand the system. That is exactly what The People Attraction Theory is: understanding how the system works, reaching exhaustivity, actually going to find people — and it's a full chapter of Module 1.