Applications arrive unevenly. A fund with an open form might get a trickle most weeks and a flood after a partner's podcast appearance. An accelerator gets most of its volume in the last few days before a deadline. In both cases the same few people are reading, and the hundredth application of the week gets less attention than the first.
Knowing how to screen startup applications is mostly about removing that unevenness. A screening process that works has four parts: a form that collects what you need, written criteria, a triage step that sends each application to the right next action, and a reply to the founder. Each part is simple. The discipline is in running all four every time.
Why screening needs a process
Venture firms see far more companies than they can back. In the survey of 885 venture capitalists by Gompers, Gornall, Kaplan and Strebulaev, the average firm screened about 200 companies a year and made about four investments. For every closed deal, firms considered roughly 100 opportunities and met the management of only about one in four. The first screen is where most companies leave the funnel, so it is where consistency matters most.
Structured methods help with that. Google's re:Work guide to structured interviewing describes using the same questions and standardised rubrics for every candidate, and reports that research has found structured assessment more predictive than unstructured interviews, with smaller differences between demographic groups. Screening startups is not hiring, but the lesson carries over: decide what you are looking for before you read, and apply it the same way to everyone.
Design an application form you will actually use
Every field on your form costs founders time and costs you reading time. Ask for what helps you make the first decision, and save the rest for later stages.
- Company basics: name, website, where it is incorporated and operates, stage, sector.
- The round: amount raising, instrument, how much is committed, who is leading if anyone.
- The business in the founder's words: what it does, for whom, and why now, in a few short answers with word limits.
- Traction with definitions: if they report revenue or users, ask how each is defined and for which period.
- Team: founders, roles, and relevant experience.
- The deck: as a PDF upload, so you can read it as the founder wrote it.
- Consent and privacy: what you will do with the information, who sees it and how long you keep it.
Skip questions you will not read at the first screen, such as detailed financial models or cap tables. Asking for them early signals that you are not respecting founders' time. Our VC application form template has a field list you can copy, and our guide on founder data privacy for VC funds covers the consent part.
How to screen startup applications in three passes
Split screening into passes that take different amounts of time. Most applications only need the first.
- 1Fit pass, a few minutes: check the hard rules from your thesis, such as stage, sector, geography and round size. Mark each as met, mismatched or unknown.
- 2Quality pass, longer: for applications that fit, read the deck and answers for the problem, the customer, traction with definitions, the team and the ask. Note each material claim and where it came from.
- 3Decision pass: choose a next action, write one or two sentences on why, and record it where the whole team can see it.
| Outcome | When to use it | Next step | What the founder hears |
|---|---|---|---|
| Advance | Fits the thesis and the quality pass raises real interest | First call or interview, with a named owner | An invitation with a date and what to prepare |
| Ask | Fits, but key facts are unknown | Send two or three specific questions | The questions, and when you will reply after they answer |
| Second read | Borderline, or the first reader is unsure | Another team member reads it independently | Nothing yet, unless your reply deadline is near |
| Pass: outside thesis | Clear mismatch on a hard rule | Record the rule and close | A short, prompt decline that says why |
| Pass: not now | Fits, but too early or evidence too thin | Record what would change your view | A decline with what you would want to see later |
The "ask" outcome matters most. Many early applications leave out the round size or the definition of a metric. Treating that as a failure rewards founders who know how venture forms work, not founders with good companies. Our guide on how to evaluate a pitch deck covers the quality pass in more detail.
Keep screening consistent across readers
- Write the criteria down before the cycle starts, and do not change them mid-cycle without telling everyone.
- Calibrate at the start: have every reader screen the same five applications independently, then compare and discuss differences.
- Read blind where you can. Hiding names and photos in the first pass keeps attention on the business.
- Use a second reader for borderline cases, and do not show them the first reader's decision until they have made their own.
- Score substance, not style. A clear answer in plain language is as strong as a polished one. Accent, fluency and presentation polish are not evidence about a business.
- Review your passes monthly. Look at which rules decline the most applications, and whether the reasons recorded still match your thesis.
Our guide on bias in startup screening goes further into these safeguards, and the deal screening scorecard gives a shared format for the quality pass.
Reply to every founder, on a schedule
Founders remember how a fund treated them when it said no. Commit to a reply time, such as two weeks from submission, publish it on your application page, and keep it. Y Combinator, for example, says on its application page that it typically makes decisions the same day as an interview and gives everyone who interviews feedback on its decision. You do not need to match that, but you do need to say what founders can expect.
A decline should be short and honest. If the company is outside your thesis, say which part. If it is not the right time, say what would make you look again. Avoid vague promises to stay in touch that you will not keep. Our guide on how to say no to founders has wording for common cases.
Common mistakes at this stage: letting applications sit with no owner, replying only to the ones you like, sending a generic decline months later, and asking founders to fill in a long form before telling them the fund does not invest in their country.
Where Delsivo fits
Delsivo is designed around this process. In the pilot release, founders apply through your branded portal on your own subdomain, with save and resume, a receipt and a secure link to their application. Each PDF deck is read and its facts extracted with slide citations. Your thesis rubric is applied criterion by criterion, with results of Met, Mismatch or Unknown, and Unknown becomes a question for the founder rather than a failure.
Applications land in an analyst queue with thesis fit, evidence quality and diligence readiness shown separately. A mismatch is surfaced for review, never an automatic rejection. Analysts decide every stage change and pass, and AI-drafted messages to founders need a person's approval before they are sent. Live AI founder interviews, which a founder starts whenever suits them, with a text alternative, are planned for the first release. The startup application portal page describes the intake side.
Sources
- Gompers, Gornall, Kaplan and Strebulaev: How Do Venture Capitalists Make Decisions? (NBER Working Paper 22587, September 2016) (checked 6 November 2026)
- Google re:Work: A guide to structured interviewing for better hiring practices (checked 6 November 2026)
- Y Combinator: Apply to Y Combinator (checked 6 November 2026)
Delsivo publishes this guide and makes software for venture funds, so we have a stake in the topic. Facts about other companies, research and regulations come from the sources listed, on the dates shown, and can change. Corrections are welcome through our contact page.