A hundred studied, one funded. Nobody can find the other 99

A hundred companies studied, one funded. The other 99 do not disappear: they raise, they hire, they come out of beta, and some will be worth a second look in eighteen months. You met them. The question is whether you can find them again, with the reason you passed, the date, and the name of the person who held the relationship. At most funds, the answer is no.

This guide describes a fund's operating system. Two rules sum it up. Every relationship carries a last-exchange date and an owner, otherwise nobody is steering it. Every pass carries an observable reopening condition, otherwise you have filed an opinion. None of it rests on data-entry discipline, because no data-entry discipline survives three months.

01What a fund loses between two investment committees

A fund's losses appear on no dashboard, because they produce no line anywhere. A company seen two years ago, passed on for a reason nobody can articulate any more, that you now find in a competitor's portfolio. An intro promised to a portfolio company, never made. A referrer who brought you three deals, and no contact in eight months. A founder followed up on the same day by two partners unaware of each other. None of those four scenes is negligence. All four are the same failure of the system.

The information does exist, and it is complete. It sits in everyone's inbox, calendar, LinkedIn thread and WhatsApp conversation, cut into as many silos as there are people on the team. It is not lost. It is private. Private information serves the fund only from the moment it is captured, dated, attributed, and readable by someone other than the person who produced it.

The scenario is not exceptional. CNBC reported in April 2025 an acceleration of senior partner departures at established firms, after a 2024 already heavy with movement. In a business where the top of the queue is produced by people, every departure takes a share of next year's sourcing with it. Unless the system recorded it first.

Keeping the trace does not mean taking a relationship away from the person who holds it, or turning your partners into sales reps filing call reports. It means the fund keeps the history, the freshness and the owner of every relationship, without asking anything of anyone. That is exactly what a CRM should do. It is exactly what a sales CRM does not do.

02Why a sales CRM cannot model a fund

A sales CRM is built to push a prospect from cold to signature, through a repeatable process and a predictable funnel that empties out. A fund holds the opposite: a multi-year queue that keeps growing, fed by relationships rather than by leads. Four differences are enough to make the tool a poor fit.

• Non-linear : a passed deal is not lost. It goes back into the queue and may become your best investment three years later.
• Multi-year : a relationship often builds over several years, and your system has to outlive it, including the departure of the partner who owned it.
• Network-driven : the top of the queue is produced by people, not received through a form. A sales CRM knows how to score an inbound lead. It cannot tell you which of your partners holds the strongest tie to a founder.
• Multiple tempos : sourcing, due diligence, portfolio monitoring and LP relations run on distinct rhythms that a single pipeline cannot model.

03What feeds the queue: your network, not your inbound

Those hundred companies studied for every one backed come from somewhere, and Gompers, Gornall, Kaplan and Strebulaev, published in the Journal of Financial Economics in 2020 from a survey of 885 venture capital investors across 681 firms, say where they come from: over 30% of deals come from partners' professional networks, close to 30% from proactive self-generated sourcing, 20% from other investors, 8% from portfolio companies, and only around 10% from inbound founder outreach. Nine deals out of ten are produced by an action or a relationship. A fund that treats every deal as an inbound lead is ignoring its own mechanics.

Affinity's 2025 benchmark adds a nuance that looks contradictory and is not. A note on method first: Affinity is a CRM vendor measuring its own customer base, nearly 3,000 firms across more than 65 countries, with top firms defined by Dealroom's investor ranking. The absolute levels hold for that population, not for the whole market; the gaps between the two groups are measured on a constant method. Those firms work on more than twice as many deals as the average. And they closed 2024 with 55% fewer deals done year over year. The same report therefore measures both ends of the queue: the network widens what you study, selectivity happens at the bottom. A strong fund looks at more and signs less.

04Qualifying a relationship: four criteria and a ritual

If the relationship produces the deals, then the relationship is the unit of production, and a unit of production gets measured. Relationship intelligence maps who knows whom across the entire fund and scores each connection from observable signals, not from partners' memory. It answers the only question that matters when you reopen a file or go after a target: who on the team has the strongest and most recent connection. A strong relationship, whether a referrer, an operator or an expert, is rare and depreciates without upkeep. Four criteria are enough to qualify one.

05Capture the relationship where it lives

The four criteria above are only worth as much as the signals reaching them. And an investor relationship does not live in a form. It lives in email, on LinkedIn and, increasingly, on WhatsApp. A CRM that only captures the inbox scores your ties on a fraction of your exchanges. The word that matters in the table below is "natively". A connector each partner has to install themselves is not capture, it is an intention. The test before you choose: take your last three months of exchanges with your five best referrers, and count which channel they actually happened on.

Email

• 
• Affinity
• Attio
• folk
• HubSpot
• Salesforce

Calendar / meetings

• 
• Affinity
• Attio
• folk
• HubSpot
• Salesforce

LinkedIn messages

• 
• Affinity
• AttioIntegr.
• folkManual
• HubSpotIntegr.
• SalesforceIntegr.

WhatsApp

• 
• Affinity
• Attio
• folk
• HubSpotAdd-on
• SalesforceAdd-on

Phone calls

• 
• AffinityIntegr.
• AttioIntegr.
• folkIntegr.
• HubSpot
• SalesforceAdd-on

Meeting transcription

• 
• Affinity
• Attio
• folkIntegr.
• HubSpotAdd-on
• SalesforceAdd-on

Channel		Affinity	Attio	folk	HubSpot	Salesforce
Email
Calendar / meetings
LinkedIn messages			Integr.	Manual	Integr.	Integr.
WhatsApp					Add-on	Add-on
Phone calls		Integr.	Integr.	Integr.		Add-on
Meeting transcription				Integr.	Add-on	Add-on

Affinity opened the relationship intelligence category for private capital: automatic email and calendar capture, a proprietary relationship score, built-in enrichment. Two limits, observable as of this guide's publication date: Affinity does not publish its pricing and shares it on request, and its relationship score is computed on the vendor side, so it does not follow your data the day you leave. Attio is modern and flexible with transparent pricing, but it is not natively designed for venture (you design the model yourself) and a company can only sit at one stage of one list at a time. folk, HubSpot and Salesforce cover email and calendar well without being built for venture.

06Putting your network to work twice: the portfolio

The network you just mapped works twice. Once to source, once to help your portfolio companies, and it is that second use that separates you the day a founder chooses between two term sheets. Affinity's 2025 benchmark shows it on both counts: 48% of firms put portfolio support among their priorities for the year, and top firms make at least 28% more introductions per quarter than the average, between their network and their portfolio companies. Value beyond the cheque is not a pitch, it is a countable volume of intros.

07What is not counted gets no credit

Unmeasured value is invisible exactly where it counts: at LP re-up, and in a competitive round where the founder is comparing two funds. Hence the counting discipline in the previous section, and hence the reporting. LP reporting is a fundraising instrument, not a chore. Re-up decisions turn on reputation and consistency, and reporting is what builds them between two funds. Transparency, write-downs included, triggers re-ups; opacity derails a raise even on good numbers. Anchor on the standards, deadlines included. The ILPA Principles 3.0 expect quarterly reporting within 60 days of quarter end and audited annual financial statements within 120 days of year end. Reporting Template version 2.0, published in January 2025, replaces the 2016 one for funds formed on or after January 1, 2026. France Invest publishes the equivalent for the French market. Those are dates, not intentions: put them in the team calendar before you promise them in a raise.

08The network checklist: healthy deal flow

You do not need Kasar to start. Open your last eighteen months of history, pull out the companies you met and passed on, and write down for each the condition that would bring you back. The number of lines you get is the real size of your queue, and nobody on your team knows it today. The checklist below audits the rest.