LinkedIn vs Xing vs ResearchGate vs GitHub: Where PhDs Should Actually Be Online

Most PhDs walk into the industry transition with one well-tended ResearchGate profile, a half-built LinkedIn from 2019 with a passport-style headshot and the headline “PhD Candidate at TU München,” possibly a Xing account they made because a supervisor told them to, and either a tidy GitHub or no GitHub at all. The instincts here are wrong for industry. The platform you have invested most into – ResearchGate – is the one industry recruiters never open. The platform you have neglected – LinkedIn – is the one that decides whether you get found.

This post is a comparison of the four platforms a transitioning STEM PhD in Germany has to think about: LinkedIn, Xing, ResearchGate, and GitHub. The goal is not to be diplomatic. It is to tell you which platform converts to interviews, which one is a vanity surface, which one to skip, and where your next four hours of effort should actually go.

What recruiters actually use

In Germany, for tech, data, ML, and engineering roles, the dominant sourcing tool is LinkedIn. Specifically, it is LinkedIn Recruiter – the paid version that lets recruiters search by keyword, filter by location and experience, and message candidates directly. It is the tool German tech recruiters live in. They run boolean searches at 9am, build candidate lists by lunch, and send InMails by 4pm. If your profile is not optimised for those searches, you are not in the candidate pool.

Xing still exists in this ecosystem, mostly for non-tech roles in older or more traditional German companies. ResearchGate barely exists in this ecosystem at all. GitHub matters intensely for some roles and is irrelevant for others. The rest of this post is the breakdown.

LinkedIn – the main platform

Verdict: ESSENTIAL.

If you only do one thing, do LinkedIn. Every German recruiter hiring for a STEM industry role uses it. Most pre-screening conversations start with “I found your profile on LinkedIn.” Most internal referrals start with someone forwarding your LinkedIn URL to a hiring manager. The platform is the front door.

The four sections that decide whether you get found, in priority order:

• Headline: not your job title. A keyword-rich line that says what you do in industry vocabulary. “PhD Candidate” is invisible. “Machine Learning Engineer | NLP & Computer Vision | PhD in Astrophysics” is searchable.
• About: three short paragraphs. What you do, what you have built, what you are looking for. No academic third person. No “passionate about.”
• Experience: each role rewritten in industry bullets – quantified, action-led, free of “novel” and “investigated.” Your LinkedIn experience section needs the same translation as your CV; the CV Bullet Translator works for both.
• Skills: ordered by what your target roles actually ask for, not by what you find most interesting. The algorithm matches recruiter searches against this list.

The way LinkedIn search works in 2026 is keyword fit plus activity signal. A profile with the right keywords beats a profile with the right experience but the wrong vocabulary. A profile that has posted or commented in the last 30 days surfaces above one that has been silent for a year. Neither effect is small.

Most PhDs already know LinkedIn matters and still under-invest in it because rebuilding the profile feels like marketing – uncomfortable, performative, off-brand. The fix is to treat it as a structured task, not a feeling. The 14-Day LinkedIn Optimisation Challenge is two weeks of 15-minute prompts, one section per day – built specifically for PhDs who hate writing about themselves. If you would rather do it yourself, the long-form walkthrough lives in LinkedIn Profile Tips for PhD Candidates.

Xing – the German-native alternative

Verdict: USEFUL FOR SOME ROLES – not a substitute for LinkedIn.

Xing is the German professional network that LinkedIn slowly ate over the last decade. It still has roughly 20 million users in DACH, but its centre of gravity has shifted: older crowd, traditional industries, more local SMEs and family-owned firms. If you are targeting a Mittelstand manufacturer in Baden-Württemberg, a regional bank, an insurance group, or a public-sector research institute, Xing is still where some of those recruiters look first. For tech, data, AI, and most multinational hiring, LinkedIn dominates by an order of magnitude.

What this means in practice: if your target list is Bosch, Siemens Energy, Allianz, a regional Sparkasse, a public research lab, or a traditional engineering firm, build a basic Xing profile and keep it consistent with LinkedIn. If your target list is SAP, Zalando, BioNTech R&D, Celonis, AWS, or any DAX tech function, Xing is a once-a-month check, not a daily presence.

Treat Xing as a mirror profile: same headline, same About, same experience bullets, same skills. Do not invest separately in posting, commenting, or community building. The activity ROI is too low. The point is presence so that a recruiter who searches Xing first finds you, not absence that flags you as a non-DACH applicant.

One technical note: Xing is stricter about completeness signals than LinkedIn. A half-filled profile actively works against you. Either fill it properly or delete the account – do not leave it as a placeholder.

ResearchGate – academic-first

Verdict: KEEP if you have it; do not invest more time in it for industry transition.

This is the one PhDs over-invest in. ResearchGate is excellent for what it was built for – academic networking, paper sharing, citation tracking, contacting authors of related work. For industry transition, it is close to useless. Most industry recruiters do not have ResearchGate accounts. Hiring managers at tech companies do not search there. The platform’s entire value system is calibrated to academic metrics – RG Score, citations, co-authorships – that industry hiring does not weight.

The mistake is not having a ResearchGate profile. The mistake is treating it as part of your industry job search. Hours spent grooming your publications list there are hours not spent rewriting your LinkedIn About section, and the marginal recruiter you reach via ResearchGate is roughly zero.

Keep your ResearchGate profile if it already exists – it is reasonable to keep your academic record visible, and you will occasionally get DMs from former co-authors or people referencing your work. Answer those DMs. Do not delete the profile. Do not invest more time in it. If a recruiter ever finds you on ResearchGate (rare), they will then look you up on LinkedIn before doing anything else, so the LinkedIn profile is still the one that decides the outcome.

The same logic applies to Google Scholar, Academia.edu, and ORCID. Maintain them. Do not optimise them for industry. The audience is not there.

GitHub – portfolio platform for technical PhDs

Verdict: ESSENTIAL FOR TECHNICAL ROLES, IRRELEVANT OTHERWISE.

For ML, data science, software, computational biology, computational physics, robotics, or any role where you will write code in production, GitHub is your code CV. Hiring managers genuinely click through to it. Recruiters do not always understand what they are reading, but technical interviewers do, and they form an opinion fast.

What a clean GitHub looks like for a transitioning PhD: a pinned set of two or three projects, each with a proper README, a one-paragraph problem statement, a results section with numbers, and a runnable example. Not your entire PhD codebase. Not 47 forks of tutorials. Not a graveyard of half-finished side projects with no README. Two or three projects you can talk about for ten minutes each in an interview.

The bar is lower than people think. You do not need a viral repo. You need legible work that signals: this person can scope a problem, ship code, and document it for someone else to read. For the longer playbook on what those two to three projects should actually look like – what to build, how to write the README, how to make a small project look like a real one – read Building a Data Science Portfolio for Europe.

If you are a humanities PhD, a pure-theory PhD, or transitioning into a non-technical role (product, consulting, scientific affairs, MSL, communications, policy, regulatory), GitHub is irrelevant. Do not force a portfolio you do not need. Spend that time on LinkedIn instead.

The 80/20 split

For most STEM PhDs targeting Germany, the time allocation is straightforward:

• 80% LinkedIn. Profile build, weekly activity, targeted recruiter outreach, post or comment once a week. This is where the conversions happen.
• 20% GitHub if you are technical. Two to three pinned, documented projects. Do this once, properly, and stop.
• Xing: a mirror profile, checked once a month for messages. No active investment.
• ResearchGate: answer DMs only. No new investment.

The split shifts only if your target market shifts. If you are deliberately aiming at the German Mittelstand or public sector, raise Xing to 20% and lower GitHub to 10%. If you are pure-theory or humanities, drop GitHub to zero and raise LinkedIn to 95% and Xing to 5%.

The thing to notice: there is no allocation in this list for ResearchGate or Academia.edu. That is intentional. Those platforms have an audience; the audience is not the people who hire you into industry.

What to actually do this week

A short, ordered checklist. Do these in sequence; do not skip ahead.

1. LinkedIn headline. Rewrite it today. Target role + two to three core skills + your PhD as credential, not as job title. Twenty minutes.
2. LinkedIn About. Three paragraphs: what you do, what you have built (with one specific result), what you are looking for. Forty minutes.
3. LinkedIn Experience. Rewrite your last two roles in industry bullets – quantified, action-led. The CV Bullet Translator handles this. One hour.
4. LinkedIn Skills. Reorder so the top three match what your target job descriptions actually ask for. Ten minutes.
5. CV alignment. Your CV and LinkedIn experience section should say the same things in the same words. If they disagree, the recruiter will notice. The ATS-Ready CV Template for PhDs gives you the structural document; the LinkedIn content should match it. See the full ATS CV breakdown for what recruiters look for.
6. GitHub (if technical). Pin two or three projects. Add a README to each with a problem statement, a result, and a run command. Two hours.
7. Xing mirror. Copy LinkedIn headline, About, experience, skills. Do not over-think it. Thirty minutes.
8. Skip: any further work on ResearchGate, Academia.edu, ORCID grooming, Google Scholar updates, or building a personal blog. None of these are blocking your next interview.

If you would rather have all of this done in two weeks of guided 15-minute prompts, with templates, examples, and accountability checkpoints, the 14-Day LinkedIn Optimisation Challenge is the structured version. PhDs who want a guided rebuild of all surfaces – LinkedIn, CV, GitHub, narrative, target list – usually go through Career Bridge instead.

For the broader question of what hiring managers actually weight when they look at your profile, read What Recruiters Look For in PhD Profiles. For the toolkit context this fits inside, see The Academia to Industry Toolkit. The free diagnostic tells you which surface is actually blocking you, and the participant testimonials show what fixing the right surface looks like in outcomes.

The honest version

Networking platforms are surfaces. They do not get you hired. They make you visible to people who could hire you, and they decide whether the first impression supports your candidacy or undermines it.

Most transitioning PhDs lose months optimising the wrong surface. ResearchGate is well-tended; LinkedIn is empty; Xing is stale; GitHub has 47 forks of tutorials and no real projects. The fix is not more effort. It is reallocated effort. Ninety minutes on LinkedIn this week beats nine hours on ResearchGate this month, every time.

Pick the platform that maps to your target market, do the work properly once, and stop. Then go do the harder thing – tailor applications, rehearse interview stories, negotiate the offer. The platforms are the entry point, not the work.