The LinkedIn Recommendation Playbook for Software Engineers
The LinkedIn Recommendation Playbook for Software Engineers
Most engineers treat LinkedIn recommendations as something that either happens to them or doesn't. A former manager writes one unprompted, or they don't. You get four in your first five years and then stop thinking about it.
That's not a strategy. It's a passive approach to one of the most credible social proof signals in software engineering hiring — one that hiring managers specifically look for when evaluating senior candidates, and one that you have far more control over than most engineers realize.
Recommendations are not the same as endorsements. Endorsements — the one-click skill confirmations — are better understood as keyword signals for LinkedIn's search algorithm: having 10+ endorsements for "Kubernetes" bumps you in recruiter searches for that skill. Useful, but low-trust. Recommendations are written testimonials that require real investment from the person writing them, which is exactly why recruiters weight them significantly more when deciding whether to move a profile to the "worth calling" pile.
Profiles with recommendations receive substantially more profile views than those without, and surveys of technical recruiters consistently show that a strong recommendation from a past manager or senior colleague is one of the faster ways to differentiate a profile when every other signal (title, company, tenure) looks similar between candidates.
Here's how to think about them — and how to be deliberate.
When Recommendations Actually Matter
Not every career moment calls for building your recommendation count. The times when they carry the most weight:
Active job search at senior levels. At 3–8 YOE, the difference between candidates on paper is usually marginal. Technical skills are assumed. What separates people is judgment, collaboration pattern, and what it's actually like to work with them. A well-written recommendation answers those questions before the phone screen.
Promotion packets. At companies that require calibration or 360-degree feedback, a strong LinkedIn recommendation from a cross-functional partner can be included in your self-nomination or cited in peer reviews. It signals external validation in a way that internal performance reviews can't.
Senior or staff-level role applications where you're competing with many technically equivalent candidates. The EM-to-director career path requires demonstrating organizational influence — a recommendation from someone you've influenced is evidence, not just a claim.
Post-layoff job searches. When you're explaining a gap or a company shutdown, a strong recommendation from a manager at the previous company is the fastest way to pre-empt the credibility question before it's asked. It's harder to doubt a candidate's performance when someone who managed them says it directly.
Who to Ask (and Who to Skip)
The instinct is to ask people who know you best and are likely to say yes. That's the right instinct — but with one important addition: ask people whose position relative to your work makes the recommendation meaningful.
High-value sources:
- Direct managers — the person who ran your performance reviews has the most credible vantage point on your impact and growth. One strong manager recommendation outweighs four peer ones.
- Senior engineers you've worked closely with — a recommendation from a staff or principal engineer who can speak to your technical judgment carries weight in hiring manager reads.
- Cross-functional partners (PMs, technical leads, designers) — these signal collaboration ability, which is frequently the unstated gate for senior roles.
- Managers from earlier in your career who've seen you grow — especially useful if you can get one from someone who's now at a more senior title than when they managed you.
Lower-value sources (not useless, but not your primary targets):
- Peers at the same level who haven't worked on something concrete with you
- Former colleagues you stayed in touch with but didn't collaborate closely with
- Mentors who know you conceptually but not through shared work
Three to five strong recommendations from the first category will do more than ten from the second category.
How to Ask Without Making It Awkward
The ask is the part engineers most often get wrong — either by procrastinating indefinitely or by sending a generic LinkedIn request with no context, which lands like a cold DM.
The approach that actually works:
1. Ask when the work is fresh. The best time to ask for a recommendation is within 60–90 days of completing meaningful work together — shipping a feature, wrapping a project, finishing an engagement. Memory is clearest, enthusiasm is highest, and it feels like a natural follow-on to something that went well. Asking someone to write about work from two years ago puts the entire burden of recollection on them.
2. Re-activate the relationship first. If it's been more than a year since you worked together, a cold recommendation request is jarring. Send a short update first — "I saw you're now at [company], congrats" or "I've been working on [project] recently and it reminded me of our work on [thing]." Give it a few days before circling back with the ask.
3. Make it easy and specific. The reason most people who are asked to write recommendations procrastinate is that they don't know what to write. Remove that friction entirely:
"Would you be open to writing a short LinkedIn recommendation? I know your time is limited — here are a couple of things you could draw on if helpful: the migration project we shipped in Q3 that reduced P95 latency by 40%, and how we handled the on-call handoff when the team was under-staffed. Feel free to ignore this entirely if it doesn't match what you remember. No pressure at all — if you'd rather skip, just say so."
That framing removes the blank-page problem, gives them concrete material to work from, and makes it low-stakes to decline. All three of those things increase the yes rate.
4. Give them an out. Ending the ask with a genuine "no pressure if it doesn't feel right" is not just politeness — it signals that you won't hold a no against them, which paradoxically makes a yes more likely. People say yes more readily when they know they can say no.
What to Write When Someone Asks You
Being asked to write a recommendation is a gift — it's a chance to do something genuinely useful for someone, and to do it well so they'll think of you the same way if you ever need one. Most engineers write weak recommendations because they use the wrong structure.
The common mistake: describe the person.
"Aisha is a talented engineer who always goes above and beyond. She's a great communicator and a real team player."
This is useless. It describes traits — not behavior. It says nothing a hiring manager can evaluate. It applies to everyone and no one.
The right structure: show, don't tell.
A recommendation should do what a reference call does: answer the question "what was it actually like to work with this person, on this kind of problem?" The components that matter:
- Context: Who were you to each other, and what were you working on?
- Specific behavior: What did they do, concretely, in a situation that tested the trait you're claiming they have?
- Observable result: What changed because of that behavior?
- The thing you'd still call them for: What kind of problem would you immediately bring to them if you worked together again?
Example (weak):
"David is an excellent backend engineer with strong system design skills. He's collaborative and easy to work with. I'd recommend him for any engineering role."
Example (stronger):
"I worked with David as his PM on a high-stakes backend migration at [company] — we were moving a core payments service to a new provider under a hard regulatory deadline. David caught an edge case in our data reconciliation logic that would have caused silent failures on refund processing. He flagged it three weeks before launch, documented the failure mode, proposed two solutions with tradeoff analysis, and got consensus from three teams in two days. That kind of clarity under deadline pressure is why I'd go to him first for anything involving distributed system reliability."
The second version gives a hiring manager something to verify, a concrete situation to ask follow-up questions about, and evidence of a specific trait — not just an assertion of it.
Length: three to five sentences is usually enough. Longer is fine if it stays concrete. Vague and long is the worst combination.
The Anatomy of a Weak vs. Strong Recommendation
Here's what separates recommendations that actually influence hiring decisions from ones that get skimmed and forgotten:
| Signal | Weak | Strong | |---|---|---| | Specificity | Lists traits ("collaborative," "hardworking") | Describes a specific situation and what they did in it | | Context | Generic ("a great engineer") | Clear relationship and working context | | Technical signal | Vague ("strong technical skills") | Specific domain ("distributed systems," "ML infra," "API design at scale") | | Uniqueness | Could apply to anyone | Describes something that distinguishes this person | | Author's position | Unclear relationship | The author's role and proximity to the work is evident |
The goal is a recommendation that couldn't plausibly have been written about someone else.
How Many Do You Need?
There's no magic number, but there's a practical floor: three strong recommendations is the minimum for a senior-level profile to feel credible. Below three, a hiring manager will wonder if there are no strong advocates, or if you just haven't gotten around to it (which implies you're not thinking carefully about how you present professionally).
Above five, marginal returns drop unless the new recommendations are qualitatively more credible (a director you worked under, a staff engineer at a well-known company) than the ones already there.
For engineers with a promotion push in the near-term future: one strong recommendation from a cross-functional partner who isn't your skip-level is often more valuable in a promotion conversation than two peer recommendations, because it demonstrates influence outside your immediate team — which is frequently a staff-level criterion.
The Reciprocal Move
If you ask for a recommendation, offer to write one in return — and do it without waiting to be asked. This isn't transactional, it's just useful: most people don't ask because they feel like they're imposing. If you write a strong one first, the dynamic shifts.
Good reciprocal recommendations also require good judgment about timing. Writing one before asking makes sense when you'd write it regardless (it's true and useful). Writing one primarily as a currency exchange gets visible in the phrasing.
The deeper version of this: your LinkedIn network is worth thinking about as a long-horizon asset. A recommendation you write today for a former colleague who's now a staff engineer somewhere interesting creates a real relationship, not just a profile line. The engineer career graph is dense — people you worked with at 24 will be CTOs and VPs at 36. The engineers who navigate careers well treat those relationships as worth tending.
TL;DR
- Recommendations outperform endorsements. Endorsements help with search ranking. Recommendations influence whether a recruiter or hiring manager picks up the phone.
- When they matter most: active senior-level job searches, promotion packets, and post-layoff reputation recovery.
- Ask managers and senior engineers who can speak to concrete work. Three strong recommendations beat ten generic ones.
- Ask when the work is fresh. Within 90 days of a completed project. Re-warm the relationship before asking if it's been more than a year.
- Make the ask easy and specific. Give them two or three concrete things to draw from. Include a genuine no-pressure out.
- When writing one, show don't tell. Describe a specific situation, what they did, and what happened — not a list of their traits.
- Three is the floor for senior profiles. Aim for three to five from high-context sources.
- Write them proactively for people you'd want in your network. The long game is the relationship, not the profile line.
Related: The Engineer's LinkedIn Playbook — the full profile optimization system that recommendations slot into.
Related: The Engineer's Internal Promotion Playbook — how to build the full evidence package for a promotion, of which recommendations are one component.
Related: The Software Engineer's Referral Playbook — recommendations and referrals operate on similar social proof mechanics; here's how referrals work in the hiring funnel.
Your LinkedIn profile is one signal. Your resume, GitHub story, and overall career narrative are the rest. Wrok builds the full career profile — AI-powered resume, GitHub integration, and portfolio — so every recruiter conversation starts from a stronger position. Try it free →