Put real numbers on your funnel and two things become obvious: how fast a first look has to be, and how much sourcing you actually need to support the portfolio you are trying to build.
12 min read · Level 2
The number that tells you most about an investor is not their hit rate, it is how many hours they spent on the opportunities they declined - and almost nobody measures it, because declined work leaves no artefact to count. A high hit rate can mean excellent sourcing or it can mean a pipeline so thin that everything in it gets taken; the two are indistinguishable from outside and, worryingly, often from inside too. Hours spent on declines is harder to fool yourself about: it is the direct cost of a filter that is not doing its job, and it is paid out of the same budget that should be funding deep work on real candidates and genuine support for companies already owned. I would rather see an investor's calendar than their portfolio if I wanted to know whether their process was real. The portfolio shows what they chose; the calendar shows what choosing cost them, and whether the expensive hours went to the decisions that were genuinely close.
A funnel is four numbers: what arrives, what survives a first look, what gets real work, and what you invest in. Most investors have never written them down, and writing them down is uncomfortable because it makes the time budget arithmetic inescapable.
Take an investor aiming to make 6 investments a year, with the ratios that are common in early-stage private investing:
Stage Count Time each Total hours Arrive 400 5 min 33 First look 80 45 min 60 Real diligence 18 20 hours 360 Invest 6 - -
Total 453 hours
That is roughly 11 working weeks, before any portfolio work, any board meetings, any time spent with companies already owned. And the ratios are not pessimistic: 400 to 6 is a hit rate of 1.5%, which is unremarkable for early-stage.
The arithmetic immediately constrains three decisions.
How long a first look can take. At 80 first looks, an hour each is 80 hours. At 20 minutes each it is 27. The difference is a month of working time, and it comes entirely from whether you have a disqualifying question ready or start from scratch each time.
How much diligence is affordable. Twenty hours each on 18 opportunities is 360 hours - by far the largest block. Halving the number reaching diligence, by screening better rather than by lowering standards, buys 180 hours: enough to do genuinely deeper work on the ones that remain.
Whether the top of the funnel is big enough. Run it backwards. If you need 6 investments and your real conversion from serious diligence is one in three, you need 18 diligence-worthy opportunities, which at a 20% first-look survival rate means 90 first looks, which at your channels' actual rates means whatever number that implies. Most investors discover their top of funnel is either far too small for their ambitions or far too large for their calendar.
One warning about the ratios: they are yours, not universal. A specialist investor in a narrow niche might see 60 opportunities a year and invest in 6, because their sourcing has already filtered. A generalist seeing everything might be at one in two hundred. Copying somebody else's ratios tells you nothing; measuring your own tells you what to change.
Because without the arithmetic, the constraint shows up as a feeling rather than a number. Investors who have not done it describe being busy, or behind, or unable to get to things - and respond by working longer rather than by changing the funnel, because they cannot see which stage is consuming the time.
It also exposes a trap specific to private investing: the diligence stage is where the hours are, and it is entered by default. Every opportunity that is not clearly declined drifts into real work, so the effective decision about how to spend 360 hours gets made by not deciding, forty minutes at a time. The funnel makes that visible.
The third reason is that it tells you whether your portfolio plan is achievable at all. An investor planning 6 investments a year with a top of funnel of 40 is planning to invest in 15% of what they see, which means either their sourcing is genuinely exceptional or their standards will have to fall to meet the number. That is worth knowing in January rather than in November.
And it reframes what a fast no is worth. Declining something in 20 minutes rather than five hours does not just save 4 hours and 40 minutes - across 60 declined opportunities it is most of a working month, spent instead on the handful where the answer was not obvious.
Educational, not personalised financial advice. Nothing in the Investor Academy is a recommendation to buy, sell or hold anything — what is right for you depends on your own circumstances and the rules where you live.
An investor measures their own funnel for the first time. They had assumed they were reasonably efficient.
Stage Actual What they assumed Arrived 310 about 300 First look 140 about 60 Real diligence 31 about 15 Invested 5 5
The discovery. Nearly half of everything that arrived got a first look, and a fifth of what arrived reached real diligence. Their assumption had been that the funnel narrowed sharply at the top. It did not - it narrowed at the very bottom, after the expensive work had already been done.
The hours. At an honest 12 hours average for the diligence stage, 31 opportunities consumed about 372 hours. Twenty-six of those investigations ended in a decline. Roughly 310 hours went into work that produced no investment, and the investor had spent the year feeling short of time to support the companies they already owned.
What they changed. Not the standards, and not the volume. They wrote a disqualifying question derived from their thesis and applied it at first look, with a pre-decided wrong answer. The following year:
Stage Year 1 Year 2 Arrived 310 295 First look 140 102 Real diligence 31 14 Invested 5 6
Diligence hours 372 168
The result worth noticing. Investments went from 5 to 6 while diligence hours more than halved. The 17 opportunities that no longer reached diligence were not opportunities they had been about to take - they were the ones where the answer had always been no and the work had been spent establishing it slowly. The 204 hours released went into deeper work on 14 real candidates and into the companies already owned.
The honest caveat. One year is a small sample and one of the six could yet fail. The claim here is not that tightening the funnel improved returns - that takes a decade to know. It is that the same number of investments came from half the hours, which is a fact available immediately.
Improving the funnel by lowering the standard at the bottom rather than sharpening the filter at the top. Both reduce diligence hours and they are opposite decisions. If the number reaching real work falls because you started asking a good disqualifying question early, that is efficiency. If it falls because you started investing in the first few that looked reasonable in order to clear the backlog, that is a deployment problem wearing a process costume.
Arrived, first-looked, diligenced, invested. Use your calendar and your inbox rather than memory, which systematically undercounts the middle stages because declined opportunities are forgotten. Most people are surprised by the second number.
Use honest averages including the meetings, the follow-up reading and the calls you did not log. The output is a total hours figure, and it should be checked against how many hours you actually have, because the two are frequently incompatible in a way nobody has noticed.
It is almost always real diligence, because it is measured in days rather than minutes. That tells you the highest-leverage intervention is the filter immediately before it, not the one at the very top.
Start from the number of investments you intend to make, apply your measured conversion rates, and see what top of funnel that requires. If the required number is far above what you see, the plan needs either more sourcing or fewer investments - and deciding that deliberately is much better than discovering it in the fourth quarter.
The first-look-to-diligence ratio is the one that drifts, because standards slip gradually under deployment pressure and no single decision looks wrong. It is the early warning that the funnel has stopped filtering.
Do this once before setting any annual plan, and again whenever you feel persistently short of time without being able to say where it went. It is most valuable for anyone making a target number of investments in a period - a fund with a deployment window, an angel with an annual allocation - because that is where the arithmetic binds hardest.
Do not run funnel arithmetic on your first year in a category, where you have no stable conversion rates and every opportunity is teaching you something regardless of whether you invest. Optimising a funnel before you know what you are filtering for encodes your earliest and worst assumptions as policy. Measure first, optimise once the ratios have stopped moving.
An investor reaches a clear negative view on an opportunity within the first few hours of real work - the numbers do not support the story, or a load-bearing assumption has failed. The investigation nonetheless continues for weeks. The reasons given are usually about thoroughness or fairness to the founder, and occasionally about not wanting to appear to have decided quickly. Measured across a year, this pattern often accounts for more hours than any other single inefficiency in the funnel, and it is invisible in the ratios because the opportunity is correctly declined in the end. The count of diligences looks fine; the hours per diligence is where it shows.
Track hours per declined diligence, not just the count. Work that continues after the decision has been made is the most expensive and least visible waste in the funnel.
Create a free account to read the rest and apply it to your own investing.