Most early-stage funds start managing deal flow in a spreadsheet, an inbox and a shared folder. That works for the first few dozen companies. Then the same startup arrives through a partner, a scout and the website form, nobody is sure who took the last call, and a founder who applied two months ago is still waiting for a reply.
Good VC deal flow management does not need much software. It needs a few agreements the whole team keeps: what counts as a deal, what each stage means, who owns the next step, and how often you look at the whole pipeline together. The sections below set those out, with the numbers worth tracking and the mistakes that cause most of the pain.
What deal flow looks like in practice
Deal flow is the stream of investment opportunities a fund sees. Survey research shows how narrow the funnel is. In the study of 885 venture capitalists by Gompers, Gornall, Kaplan and Strebulaev, firms considered about 101 opportunities for every deal they closed. They met the management of about 28, reviewed about 10 at a partners meeting, ran due diligence on about 4.8 and offered about 1.7 term sheets.
The same study asked where closed deals came from. Over 30% came through professional networks, about 20% were referred by other investors, 8% were referred by portfolio companies, almost 30% were generated proactively by the firm, and about 10% came inbound from founders. Respondents reported spending about 15 hours a week on sourcing. These figures are averages from 2016 across many types of fund, but the shape is useful: deal flow arrives from many directions, and most of it never reaches a partner meeting.
Our guide on inbound vs outbound deal flow looks at the sources in more detail.
Set up VC deal flow management in five agreements
- 1One record per company. Whatever tool you use, each startup has exactly one row or record. When it arrives again through another channel, add the new contact to the existing record instead of creating a second one.
- 2Record the source. Note who or what brought the deal: a named person, a scout, an event, your application form, outbound research. You cannot learn which sources work if you do not write them down.
- 3Define stages with exit tests. Each stage has a plain sentence describing what must be true to move on, so two people would move the same deal the same way.
- 4Name an owner and a next step. Every open deal has one person responsible and a dated next action. A deal with no next step is a deal you have passed on without telling the founder.
- 5Commit to reply times. Decide how long a founder waits at each stage before hearing something, such as two weeks for an initial decision, and track it.
Pipeline stages and their exit tests
Stage names matter less than what each one means. A typical early-stage pipeline looks like this, with two side stages for deals that leave the main flow.
| Stage | What it means | Exit test | Typical owner |
|---|---|---|---|
| New | Arrived and recorded, not yet read | Fit checked against hard rules | Analyst |
| Screening | Being read for fit and quality | Advance, ask questions, or pass with a reason | Analyst |
| First meeting | Call or interview with the founders | Partner wants to spend more time on it | Partner or associate |
| Diligence | Key questions agreed and being worked | Findings written up, open issues known | Sponsoring partner |
| Committee | Decision being prepared and taken | Decision recorded with reasons | Sponsoring partner |
| Terms and closing | Term sheet offered, documents in progress | Signed and funded, or withdrawn | Partner with counsel |
| Passed (side stage) | Declined at any stage | Founder told, reason recorded | Owner at that stage |
| Revisit (side stage) | Declined for now with a reason to look again | Revisit date or milestone reached | Named team member |
Keep side stages separate from the main flow. A passed deal with a recorded reason, and a revisit trigger such as "when they reach a stated revenue level", is far more useful a year later than a row marked "no". Our guide on how to screen startup applications covers the screening stage in detail.
Run a weekly pipeline review
A short weekly meeting, thirty minutes or less, keeps deal flow moving. The aim is not to discuss every company. It is to catch deals that are stuck, unowned or overdue.
- New this week: how many arrived, from which sources, and how many are still unread.
- Overdue replies: any founder past your committed reply time, and who will write today.
- Stuck deals: anything in the same stage for longer than your usual pace, with a decision to advance or pass.
- Decisions needed: deals a partner must look at before next week.
- Passes: a quick read of reasons recorded, to spot drift from the thesis.
Once a quarter, look at the numbers behind the pipeline: conversion from each stage to the next, conversion by source, median time in each stage, and median time to first reply. Do not set targets for these until you have a few quarters of your own data. Use them to ask better questions, such as why one source produces many meetings and no diligence.
Common deal flow mistakes
- Duplicate records. The same company in two rows leads to two analysts doing the same work and, worse, two different answers to the founder.
- Stages without meaning. "Interesting" and "warm" are moods, not stages. Use exit tests.
- Silent passes. Paul Graham's essay How to Raise Money warns founders that some investors never say no and simply stop replying. Every pass should end with a message.
- Losing the evidence. Notes in personal inboxes and files in private folders vanish when someone leaves. Keep the deck, notes and decisions with the deal record.
- No source tracking. Without it, you cannot tell which partners, scouts or events bring deals that progress.
- Treating the tool as the process. A new system without agreed stages and owners recreates the same mess with better formatting.
Where Delsivo fits
Delsivo is designed to give each deal one record that holds the application, the deck, the thesis screening, the claims and the decisions. In the pilot release, founders apply through your branded portal, through private referral links, or an analyst creates the record. Possible duplicates are suggested by company domain and founder identity, and an analyst confirms any merge. Applications go into an analyst queue, and every change is kept in an audit history.
Suggested stages run from New through Screening, First meeting, Diligence, Committee, Terms and Closed, with Passed and Revisit as side stages. Interviews arrive in the first release. CRM import and export, an intake mailbox, scout links, tracking passed deals for a revisit and weekly partner briefs are planned for a later release, so in the pilot most funds would run Delsivo alongside their existing CRM or spreadsheet. People move every deal and approve every message to a founder. The VC deal flow software page has more on the product side.
Sources
- Gompers, Gornall, Kaplan and Strebulaev: How Do Venture Capitalists Make Decisions? (NBER Working Paper 22587, September 2016) (checked 6 November 2026)
- Paul Graham: How to Raise Money (September 2013) (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.