The Sourcing Playbook for Mobility Teams: How Direct Channels Fill What Job Posts Can't

A hiring manager I know ran a job post on LinkedIn for six weeks. Two hundred twelve applications. Zero passed the phone screen. The req sat open for another four months before we talked.

The problem was not the job post. The job post was fine. The problem was that the candidates who would have passed the screen were not applying. They were employed, not looking, and not spending Sunday evenings scrolling LinkedIn Easy Apply. The posting was reaching exactly the wrong population.

This is not a niche failure mode. It is the default outcome for ADAS and autonomy reqs when the sourcing strategy is "post and wait." The candidate pool for safety-critical embedded software, sensor fusion, perception, and motion planning is small. The overlap between that pool and the set of people actively applying to jobs at any given moment is smaller still. Job posts work when passive candidate density is high. For technical mobility roles, it rarely is.

What fills these reqs is a different mix: direct outreach to passive candidates, a structured referral engine, and community sourcing. Most teams use one of the three. The ones filling reqs in 42 days use all three in parallel, activated before urgency is live.

Channel One: Direct Outreach

Direct outreach is not cold messaging. Done right, it is a warm, specific, credible message to a person you have already researched and a role you understand well. The failure mode is the spray-and-pray InMail: generic, volume-first, easy to ignore.

The sequence that works starts before the req is open. When you know a function is scaling, begin building a target list three to four months ahead. Use LinkedIn Recruiter filters for title, company, tenure, and geography. Cross-reference with GitHub for engineers, conference speaker lists for technical specialists, and published papers for autonomy and AI researchers. Build a list of specific names, not job titles.

The outreach message has three components: what caught your attention about their specific background (one sentence, verifiable), what you are building and why it is worth their attention (two sentences, specific), and a low-commitment ask (a 15-minute call, not an interview). Response rates on messages that follow this structure are consistently higher than on template blasts. The candidates who respond to a specific, credible message are also more likely to move through the process and close.

Response timing matters. Messages sent Tuesday through Thursday between 7 and 9 a.m. in the candidate's time zone outperform messages sent on Friday afternoons and Monday mornings. This is not speculative: it is a pattern that holds across the AV and eVTOL searches I have run.

Channel Two: Employee Referrals

Referral programs fail at most companies for one reason: the ask is too passive. You post the req on the intranet, send a Slack message asking people to share it, and wait. The team has 40 open browser tabs and forgets by lunch.

The version that works is active and specific. When a req opens, the hiring manager and recruiter sit down together and name five to eight people who would be excellent in the role. Not job titles, not "someone with sensor fusion experience," actual names. Those names become the active ask: "Do you know this person? Have you worked with anyone like them? Who in your network has done this kind of work?"

Referral timing is also underrated. The best time to activate your referral network for a role is in the first week, not after direct sourcing has stalled. Referrals from the first week of a search have a higher close rate and shorter time-to-offer than referrals surfaced at week six when the hiring manager is frustrated.

Referral incentives matter less than most companies think. A $2,000 bonus announced in a Slack message gets less engagement than a direct ask from a peer who believes in the role. The incentive is a tailwind, not the engine. The engine is specificity and relationship.

Channel Three: Community Sourcing

Technical mobility professionals organize themselves in ways that job boards do not index. Autonomous vehicle engineers participate in forums around ROS, CARLA, and specific sensor platforms. eVTOL specialists attend AIAA conferences and follow FAA certification working groups. Battery and BMS engineers cluster in IEEE communities and a handful of LinkedIn groups that have actual signal-to-noise discipline.

Community sourcing means being present in these spaces before you have an open role. Contributing something: a write-up, a data point, a question worth answering. When you reach out to someone you have interacted with in a technical community, the context transforms the message from cold to warm. You are not a recruiter who found them on LinkedIn. You are someone they have seen show up in their space.

The setup cost is real: it takes months to build credibility in a technical community. But the payoff compounds. A sourcer who is known in the Joby, Archer, or Aurora engineering communities has a substantially different response rate than one who sends cold InMails from outside that world.

Channel by Function

Not every channel works equally for every function. Here is how I allocate effort across the sourcing mix for common mobility roles:

How This Worked at AFEELA

When I ran recruiting for the U.S. launch of AFEELA at Sony Honda Mobility, the req list included 48 hires across 9 functions in 12 months. There was no agency on the account. Every hire came through direct sourcing, structured referrals, or a combination of both.

The 42-day average compares to a 60 to 75 day industry average. The difference was not speed for its own sake. It was sourcing architecture: for each function, we had a primary channel active before the req opened, a referral ask in the first week, and a named target list on day one. The process was not reactive to the applicant pool. It created its own pipeline.

At Fisker, where the org scaled from 300 to over 1,700 employees at 17 hires per month on average (51 at peak), the same model applied. The months where sourcing channels were running in parallel before urgency hit were the months that met target. The months where we were reactive to job post volume were the months that missed.

The Timing Problem

The single most common sourcing failure in mobility hiring is not a channel problem. It is a timing problem. Teams activate sourcing when the req is open and the hiring manager is impatient. At that point, direct outreach has a two to six week runway before it generates interviews. Referral activation takes a week to start producing names. Community credibility takes months you do not have.

The fix is pre-req sourcing. When you know a function will scale in the next quarter, start the sourcing motion now. Build the target list. Make the referral ask. Engage in the community. By the time the req is formally open, you have pipeline. The 42-day fill does not start at req open. It starts three months earlier.

What to Build First

If you are starting from scratch, the order matters. Direct outreach infrastructure comes first: a working Recruiter license, a target list methodology, and a message template you can personalize in under five minutes per candidate. This is the highest-yield channel for hard technical roles and the one with the most leverage on time-to-fill.

Referral activation comes second, not because it yields more hires but because the setup is low-cost and the pipeline appears quickly once the ask is specific. A structured referral ask in week one of a search takes 30 minutes and regularly surfaces one to three strong names.

Community sourcing is a longer investment. Start by identifying the two or three communities where your target candidates actually congregate. Engage consistently. The payoff is a pipeline that does not require you to restart from zero on every search.

Job posts belong in the mix, especially for operations and business functions. They do not belong as the primary strategy for the roles where passive candidate density is lowest, which are the hardest roles to fill and the ones most likely to slip timelines.

The companies filling ADAS and autonomy reqs in under six weeks are not doing something exotic. They are running three channels where most teams run one, and they start before the urgency is live rather than after it arrives.