Proof assets for job seekers: show the work behind your claims

You're updating LinkedIn after a quarterly check-in. Looking through your past roles, you remember the six-month stretch where you ran a customer-onboarding overhaul that cut time-to-value from 14 days to 3. The meetings. The resistance from the support team. The launch. You remember the bullet on your resume that summarizes it as "redesigned customer onboarding, reducing time-to-first-value by 80%." What you don't have is anywhere to point a hiring manager who wants to know whether you actually ran that project or whether you stood next to someone who did. The bullet is a claim. The receipts live in a Notion you no longer have access to.

A resume claims. A proof asset demonstrates. Proof assets are one of the four artifacts the pillar argues compound across a search, with a caveat: they compound across applications when you build them once and reuse them; across jobs, less reliably, because the receipts that prove they're yours sometimes live in workspaces you no longer have access to. That asymmetry shapes everything that follows.

Why proof beats claims

Ashby's first-party ATS data shows applications per hire roughly tripled from 2021 to 2024, reaching 291 per hire by early 2026. Gem's 2025 benchmarks put the average recruiter on 56% more open requisitions and 2.7 times the applications they handled three years earlier. Remote roles concentrate it further: LinkedIn's Economic Graph found fully remote listings were about 8% of US postings on LinkedIn in 2024 while drawing about 40% of the applications there.

That does not mean every posting is a stampede. NBER's analysis of 125 million applications to tech job postings found a median of four applicants per job, and one in five drew nobody at all. The pile is wildly uneven. What is consistent is the reading side: the attention budget per application keeps shrinking. Ladders' eye-tracking work put the first pass over a resume at about 7.4 seconds, and that was back in 2018.

In seven seconds, "increased revenue by 30%" reads the same as twenty other bullets that say "increased revenue by 30%." A document a hiring manager can open in 60 seconds and decide "this person actually thinks like that" does not.

Hiring managers are not reading your application looking for reasons to say yes. They're looking for reasons to say no. Every interview question, from "Tell me about a time..." to "Why should we hire you?" is really asking: how do I know you won't be a costly mistake? Proof assets answer that before anyone has to ask.

Four proof asset formats that work

You don't need all of these. Pick the one that fits your role and build one strong example. If someone has told you to build a career portfolio or a professional portfolio, this is what goes inside it. The container matters far less than having one artifact in it that a hiring manager can open.

1. Case study (the SAOL structure)

Works for almost any role. Structure it as Situation, Approach, Outcome, Learning:

• Situation. What was the context? What problem existed?
• Approach. What did you do? What decisions did you make and why?
• Outcome. What happened? Use numbers where possible.
• Learning. What would you do differently? What did this teach you?

A case study can be a Google Doc, a Notion page, or a clean PDF. Keep it to 1-2 pages. The goal is to show your thinking process, not just the result. These same case studies are what your source resume bullets are quoting from when you tailor the argument for a specific role.

Tip: the "Learning" section is what most people skip. Don't. It signals you can assess your own work honestly, and hiring managers almost never see that in applications.

2. Personal site

A personal URL signals effort. It tells a hiring manager you care enough about how you present yourself to actually invest time in it.

You don't need to be a developer. Tools like Lovable, Framer, Carrd, or Webflow make this accessible to anyone. Include:

• A clear headline that says what you do and who you help
• 2-3 highlighted projects or accomplishments with brief context
• Links to relevant profiles (LinkedIn, GitHub, portfolio)
• Contact information

Keep it simple. One page is fine. You're not trying to win a design award. You're giving someone a single link that tells your professional story better than a resume ever could.

3. Portfolio (creative and visual roles)

If you're in design, content, marketing, or any role where the work is visual, a portfolio is expected. But more is not better.

Curate ruthlessly. Show 4-6 of your strongest pieces, each with brief context:

• What was the problem or brief?
• What was your specific role?
• What was the result?

Six excellent pieces with clear context will always beat twenty screenshots and no explanation. A hiring manager deciding whether to keep reading gives you moments, not minutes. Act accordingly.

4. Proof link

Sometimes one strong artifact does the job. This is a single link you can drop into an application, email, or LinkedIn message:

• A Loom video walkthrough of a project or process you built
• A slide deck that demonstrates your strategic thinking
• A detailed write-up of how you solved a specific problem
• A before/after analysis showing measurable improvement

The proof link works especially well for follow-up emails and LinkedIn conversations. Low friction for the recipient, high signal for you.

Handling confidentiality

The most common objection: "I can't share my work because it's confidential."

Yeah, probably. But that doesn't get you off the hook. You have options:

• Anonymize details. Change company names, product names, and specific numbers. Keep the structure and thinking intact.
• Focus on process. Describe your approach and methodology without revealing proprietary outcomes.
• Use a different project. Side projects, volunteer work, and freelance engagements all count.
• Create something new. Build a proof asset specifically for your job search. A spec project for a target company or a case study of a public challenge in your industry works well.
• Build on public or sample data. A template filled with invented numbers, or an analysis of a public company's filings, shows the skill with nothing to leak.

Tip: nobody's asking you to leak trade secrets. They want to see how you think.

When the door locks behind you

The hardest confidentiality case is the one most people don't plan for. You leave a company. IT revokes your account the next morning. The proof asset you could have built (the Notion page, the case study draft, the shared deck) is somewhere you no longer have access to. The bullet on the resume is what survives.

That asymmetry is why this entire piece argues for building proof while you still have access. The case study you write in week six travels with you. The one you try to reconstruct in month eighteen rarely does. If you're currently employed, treat the week you give notice as the deadline, not the start. If you're already out, work from what you can verify externally: public artifacts, references, signed credentials, work you can rebuild from memory and external evidence.

Role-specific proof asset examples

Designers have portfolios and engineers have code. If your work is neither, the hard part is knowing what to make.

A rule of thumb that works for most roles: show the kind of thing you'd be asked to produce in the job. Work samples, in the plainest sense of the phrase. That can be something you already made, cleaned up and anonymized, or a small sample you make now. It lets a hiring manager picture your work on their team instead of guessing at it.

A reusable asset proves you've done the work. A small sample made for the company you're applying to shows how you'd do it for them, and it's the fastest route when you have no past work you can share. "Here's how I'd rewrite your refund article" hands them their own problem with your thinking already applied. Each list below ends with one way to do that.

Tip: keep company-specific samples small: one page, a few hours at most. You're showing your approach. Don't do the job for free. Label anything you invented, like sample tickets or dummy numbers, as a sample.

You're also working without their data, so write it as what you'd test or ask about first. A verdict on what they got wrong won't land.

Sales

• A deal narrative walking through how you identified, qualified, and closed a specific opportunity
• A cold outreach sequence with anonymized response rates and conversion data
• A territory plan showing how you prioritized accounts and allocated time
• Made for them: a one-page account plan for a customer they'd want to win: why them, why now, who the buyer is, and your first three touches

Marketing

• A campaign deck with strategy, execution, and results (even with anonymized metrics)
• Content samples paired with performance data: impressions, engagement, conversion
• A competitive analysis or positioning document
• Made for them: a walkthrough of their signup or landing page, with the three tests you'd want to run first and why

Engineering

• GitHub projects with clean READMEs that explain the problem, approach, and how to run the code
• Technical blog posts that walk through your problem-solving process
• System design documents showing how you think about architecture and tradeoffs
• Made for them: if they maintain open-source code, pick up an issue they've marked as open to contributors and follow their contributing guide (let the work speak, and keep the job pitch out of the pull request)

Data and Analytics

• An analysis of a public dataset, with the question, the method, and the "so what" stated at the top
• A dashboard built on sample data, with a note on who it's for and which decision it supports
• A metric definition: how you'd define "active user" for a product, and where the obvious definition misleads
• Made for them: an analysis of something public about their product or market, such as a hundred app store reviews you read and categorized by hand, ending in one recommendation

Customer Support

• A help article you rewrote, shown before and after, with a note on what you changed and why
• Three sample replies to tickets you invent: an angry refund request, a vague bug report, and a feature request you have to decline
• A bug report written for engineers: steps to reproduce, expected versus actual behavior, and who it affects
• Made for them: go through their help center as a new user and note the five places you got stuck and what you'd try

Customer Success

• Documentation you've created for customers or internal teams
• A process improvement case study showing before/after metrics
• An onboarding playbook or customer health framework you developed
• Made for them: a first-quarter success plan for one of their customer types, with the early signs that an account is at risk

Product Management

• A product requirements document (PRD) or product spec for a feature you shipped (anonymized if needed)
• A roadmap framework showing how you prioritize and make tradeoffs
• A launch retrospective with what went well, what didn't, and what you'd change
• Made for them: sign up for their product and write a one-page brief on the first thing you'd look into: the problem, who has it, and how you'd know a fix worked

Operations

• Process documentation that shows your ability to systematize
• A policy document with clear rationale for the decisions made
• A project plan demonstrating how you scope, sequence, and manage work
• Made for them: pick a process you can see from outside, like how they handle applications or returns, and write the one-page procedure you'd want a new hire to follow

Finance and Accounting

• A month-end close checklist with an owner, a due day, and a review step for each task (it shows process thinking without a single confidential number)
• A year-over-year analysis of a public company's results that explains the three biggest movements, or a three-statement model built from their annual report
• A one-page memo explaining a technical treatment, like revenue recognition on a multi-year contract, to a founder with no finance background
• Made for them: if they're public, a one-page read of their latest filing with three things you'd want to understand in your first week; if they're private, a close calendar sized for a company at their stage

People and Recruiting

• A 30-day onboarding plan for a remote hire: who they meet, what they deliver, and how you'd know it's working
• An interview scorecard for one role, with the rubric that keeps a panel consistent
• A policy you rewrote in plain language, shown before and after
• Made for them: take one of their open job descriptions and draft the version you'd post, with the reasoning behind each change

If your role isn't listed, the rule of thumb still holds. Ask what you'd be expected to produce in your first month, and make a small version of it.

When and how to use proof assets

Building a proof asset is half the job. Knowing when to deploy it is the other half.

In applications. "I've attached a brief case study on a similar challenge I tackled at my previous company." This immediately differentiates you from candidates who submitted a resume and nothing else.

On LinkedIn. Add proof assets to your Featured section. When someone visits your profile (and hiring managers will), they see evidence, not just claims.

In follow-up emails. "I put together a quick overview of how I'd approach the challenge you described. Would love to hear your thoughts." This turns a generic follow-up into a value-add.

In interviews. "I actually wrote up my thinking on this. Happy to share it after our conversation." Signals preparation and seriousness. The interview framework walks through how a proof asset becomes the STAR+R answer when the question lands.

Tip: proof assets shift the dynamic from "please consider me" to "here's what I bring." That's a different conversation.

Non-traditional proof asset options

Not everything has to be a PDF or a portfolio site. Some of the best proof assets are unconventional:

• Claude artifacts. Use AI to help you create polished analyses, frameworks, or documents. The thinking is yours; the tool just makes the output cleaner.
• Interactive demos. Build a simple prototype or interactive page using tools like Lovable to demonstrate a concept.
• Blog posts. A well-written post about a problem in your field shows expertise and communication skills simultaneously.
• Video walkthroughs. A 3-5 minute Loom video explaining how you solved a problem is more engaging than any document.
• Before/after analyses. Take a public example in your industry and show how you'd improve it. Original thinking applied to real problems is compelling.

What this looks like inside the toolkit

The proof-asset-creator skill in the open-source job-hunt-skills repo turns a half-hour conversation about a project you actually ran into a clean SAOL case study, anonymization decisions and all. Use whatever LLM you already pay for.

Build the receipt

Most candidates submit a resume and hope. That's the floor, not a strategy.

You don't need a portfolio site or six artifacts. You need one piece of evidence that shows how you think when nobody's watching, and you need to build it while you still have access to the work that proves it.

Build the receipt. Keep a copy.

The full thinking behind why proof assets compound across applications but not always across jobs: Nothing earned is lost: the case for a job search that compounds. Remotivated Pro bundles saved searches, alerts, and the tracker that help you ship proof faster. The open-source toolkit that runs this loop with whatever LLM you already pay for: github.com/Remotivated/job-hunt-skills.

Sources

• Ashby Recruiter Productivity Report, 2026 Talent Trends edition: first-party ATS data; applications per hire roughly tripled from 2021 to 2024, averaging 291 per hire by Q1 2026. ashbyhq.com
• Gem, 2025 Recruiting Benchmarks Report: first-party data across 140M+ applications; the average recruiter manages 56% more open requisitions and 2.7x more applications than three years prior. gem.com
• LinkedIn Economic Graph, "The Remote Work Gap" (2024): fully remote roles are about 8% of US postings on LinkedIn and attract about 40% of the applications there. Figures are LinkedIn-platform rates, not whole-labor-market rates. economicgraph.linkedin.com
• Davis, Steven J. and Brenda Samaniego de la Parra, "Application Flows": NBER Working Paper 32320, April 2024 (revised December 2025); 125 million applications to 7.5 million tech job postings; median of 4 applicants per posting, one-fifth attract zero. nber.org
• Ladders eye-tracking study (2018, via HR Dive): recruiters spend an average of 7.4 seconds on the initial resume scan. hrdive.com