The case
The money is the smallest part of it.
What a first check actually does, and why people do this for reasons that have nothing to do with a return.
Most people who write a first check into a company in their own state are not doing it for the return. They do it because they know the industry and can see the thing working, because they were backed once themselves, or because they can count the employers left in the county they grew up in. The return matters and it is at the bottom of this page with nothing left out. It is the last reason, and for most of the people who do this it is not the reason at all.
What a first check does
It decides whether the company exists.
This is the part that has been measured, and it is the strongest evidence on the page.
Researchers took two angel groups and compared the companies they funded against the companies they came close to funding and passed on. Same groups, same year, same room, businesses sitting either side of the line where a funding decision flips. The ones that got the check were 30 to 50% more likely to survive, to raise again, to hire, and to patent. That is as close to a controlled test as this field gets, and it says a first check changes what happens rather than merely picking the companies that were going to make it anyway.
better on survival, hiring and growth for the companies that got the check over the ones that nearly did
jobs added by firms in their first year. Net job growth comes through new firms, because existing firms are net job destroyers in most years
of rural bank branches closed between 2012 and 2017, against 9 percent in urban counties. More than 2,100 places in the country are more than ten miles from a branch, and over 1,500 of them are rural
income growth per head in counties with more small, locally owned employers. Where the employers are large and owned somewhere else, it grows slower
That last pair is the whole argument for doing this at home. A company owned in Georgia keeps its payroll, its suppliers and its decisions in Georgia. A county that loses its one employer does not get another one, and the grocery and the clinic go with it. None of that shows up in an internal rate of return.
Kerr, Lerner and Schoar, Review of Financial Studies 27(1), 2014. Kane, Kauffman Foundation, Census Business Dynamics Statistics 1977 to 2005. Federal Reserve, Perspectives from Main Street, November 2019. Fleming and Goetz, Economic Development Quarterly 25(3), 2011. United States figures.
What it asks of you
The check is the least of what it takes.
Four findings from the same research, all of them about time rather than money.
Read those four together and they describe something a person cannot do alone. Twenty hours of diligence per deal, inside your own industry, staying involved monthly, across fifty companies, is a full-time job for a group. The typical angel holds seven investments and is nowhere near diversified enough for the average return to be the one they should expect. That is the case for organising, and it is the case this organisation exists to answer.
It also does something for the companies. Ventures that angel groups funded did 30 to 50% better on survival, later financing, employment and growth than ventures the same groups came close to funding and passed on. That study compares companies either side of the line where a funding decision flips, which is as near to a controlled test as this field gets.
Wiltbank and Boeker, 2007, Kauffman Foundation and Angel Capital Education Foundation. Gregson, Bock and Harrison, Venture Capital 19(4), 2017. The American Angel, Angel Capital Association and Wharton Entrepreneurship, 2017. Kerr, Lerner and Schoar, Review of Financial Studies 27(1), 2014. United States figures. Gross, before fees, and survivorship biased: these studies observe exits and cannot see the investments that never reached one. Growth ventures only. These are associations across investors and cannot show that the behaviour caused the return. Gross, before fees, and survivorship biased: these studies observe exits and cannot see the investments that never reached one. Growth ventures only.
What a state gets back
Two programmes, both measured from outside.
Both are state-backed, so read them as what a place gets rather than as a forecast of any private return.
is the total state appropriation to MassVentures since 1978. It has recycled its own returns ever since. An outside study puts the annual average at $1.9B in state GDP and 4,346 jobs.
is about $661M of state investment in the Georgia Research Alliance since 1990, reported alongside about $7.8B leveraged. This is the Alliance’s own thirty-year reporting and has not been restated since 2020.
Georgia has already done a version of this and it worked. Invest Georgia’s portfolio employs 4,870 Georgia residents across 115 Georgia companies today.
UMass Donahue Institute, IMPLAN, May 2025. Georgia Research Alliance 30-year reporting. Invest Georgia 2025 Annual Report. United States figure.
Why here
Georgia is underweight its own capability.
The research is here, the companies are here, and the first money is not.
in venture capital drawn by Atlanta over five years. A Georgia Tech model of ecosystem strength implies about $66B. Atlanta ranked top tier on 15 of 16 metrics
growth in venture funding per dollar of state GDP in Georgia from 2012 to 2022. The national figure is 3.7x
of US innovation sector growth from 2005 to 2017 landed in five metro areas. The other 343 lost share
of American angels already live outside those coastal metros. This is the one layer of capital a state can build for itself
Georgia Tech Office of Commercialization, 2026. Metro Atlanta figure. PitchBook and BEA analysis, 2024. Georgia figure. Atkinson, Muro and Whiton, The Case for Growth Centers, Brookings and ITIF, December 2019. United States figure. The American Angel, Angel Capital Association and Wharton Entrepreneurship, 2017, from 1,659 accredited angels. United States figure.
Where it came from
One professor named it in 1983.
Before that it was a Broadway word for the person who paid for the show.
An angel put up the money for a production and was repaid only if it ran. If it closed in a week, nothing came back. The word carried that bargain with it.
William Wetzel at the University of New Hampshire interviewed 133 private investors across New England and published Angels and Informal Risk Capital. It is the first use of the word to mean somebody putting their own money into a company no institution would look at yet.
He founded the Center for Venture Research, which has counted the US angel market every year since. Until then nobody knew how big it was.
Hans Severiens invited twenty-five people to dinner in Menlo Park and twelve came. The Band of Angels was the first group of its kind, formed because venture firms had grown too large to write the first check.
From about ten angel groups in 1996 to more than three hundred by 2013. What a group buys a person acting alone is shared deal flow and shared diligence, and the research below says those two things are most of the difference between a good result and a bad one.
Van Osnabrugge and Robinson, Angel Investing, Jossey-Bass, 2000. Wetzel, Angels and Informal Risk Capital, Sloan Management Review 24(4), 1983, and Sohl and Harrison, Venture Capital 20(3), 2018. University of New Hampshire, Center for Venture Research. Band of Angels. Angel Capital Association. United States figures. The standard account across the angel investing literature. No primary record of a first use was verified for this site, so it carries no date. Popular accounts date this to 1978, the year the fieldwork began. The memorial written by his successor at the Center for Venture Research dates it to the 1983 article, which is the published record. The claim to being first is the group’s own. The earlier counts circulate through association materials and contemporary trade press. No single primary series was located.
How big it is now
Angels fund more companies than venture capital does.
In much smaller amounts, and in many more places.
of angel money in 2023, into 54,735 companies at about $339,000 each
of US venture capital in 2024, across about 15,260 deals. Far more money, reaching far fewer companies
active angel investors in the country in 2023, a number that rose while the dollars fell
of them live outside Silicon Valley, New York and Boston, and write larger checks than the ones inside
That last figure matters more here than anywhere. Angel capital is the one layer of the stack that was never concentrated on two coasts. Georgia does not need permission from anyone to build this.
University of New Hampshire Center for Venture Research, The Angel Market in 2023. PitchBook and NVCA Venture Monitor, Q4 2024. The American Angel, Angel Capital Association and Wharton Entrepreneurship, 2017, from 1,659 accredited angels. United States figures. The angel and venture totals come from different years and different counting methods, so they sit side by side and do not divide into each other.
Who is allowed to do it
Ten times as many households as in 1983.
Nobody voted for that. The thresholds simply never moved.
An individual qualifies as an accredited investor on income above $200,000, or $300,000 with a spouse, or a net worth above $1 million outside the home. Those numbers were set in 1983 and have never been adjusted for inflation. In 1983, 1.8 percent of US households cleared them. By 2022 it was 18.5%, some 24.3 million households, and the SEC’s own staff put the change down mostly to the thresholds standing still.
The JOBS Act of 2012 was supposed to open this further, and in law it did: Regulation Crowdfunding let the general public buy equity in a private company for the first time. In practice it stayed small. $1.3B was raised under it in total between 2016 and 2024. One year of ordinary angel investment is more than ten times that. The people writing first checks in this country are still mostly individuals with money, meeting in rooms.
SEC staff, Review of the Accredited Investor Definition under the Dodd-Frank Act, December 2023. SEC Division of Economic and Risk Analysis, crowdfunding market analysis, May 2025. Angel Capital Association. United States figures. The SEC describes its proceeds total as a low estimate, because issuers vary in how they file the closing form. Self-reported by the association.
What the research does not cover
All of it describes one instrument.
Equity in a company that gets sold. Most Georgia companies will never be sold.
Every figure above measures the same thing: an equity stake in a company that reached an exit. That is the right instrument for some businesses here and the wrong one for most of them. A machine shop in Dalton can be a good business for forty years and never be sold to anybody. The published angel research has nothing to say about it, which is a limit of the research rather than of the business.
It is worth knowing what the equity route has done at the institutional level too. Kauffman reviewed its own twenty years and roughly a hundred venture funds and found that after 1997 more cash went into venture capital than came back out of it. is the year after which more cash went into venture capital than came back out of it. Kauffman reviewed its own twenty years and roughly 100 fund investments and found returns had not meaningfully beaten public markets since the late 1990s. So we write the first check either way and take the return in whatever form the business can pay.
Mulcahy, Weeks and Bradley, We Have Met the Enemy and He Is Us, Kauffman Foundation, 2012. United States figure.
And the money, last
Most angel investments lose money.
Here it is with nothing left out, at the bottom of the page, where it belongs.
average return over 3.5 years, about a 27 percent IRR, across 1,137 exits. Wiltbank and Boeker collected it from 539 angels in 86 groups. This is the number the practice is usually sold on.
of those same exits returned less than the money that went in. Both numbers describe the same set of companies. A few returned many times the capital and carried the rest.
Only 7% of exits returned more than ten times capital, and the top five percent of investments produced 57 percent of all the cash that came back. In the 2016 follow-up the multiple barely moved, to 2.5x, while the share of outcomes returning less than capital rose to 70%. That is the survivorship problem stated as a fact rather than as a disclaimer: losers take longer to resolve than winners, so a dataset looked at later has more of them in it.
The best evidence in the field is not a survey at all. Norwegian tax records cover every individual investor in every early-stage firm in the country from 2004 to 2018, including everyone who lost. There, 1 in 3 angel investments is a total loss and only about a quarter return more than went in. Those numbers are worse than the American survey numbers in exactly the way the survivorship critique predicts, which is the strongest reason to trust both.
Wiltbank and Boeker, 2007, Kauffman Foundation and Angel Capital Education Foundation. Wiltbank and Brooks, Tracking Angel Returns, Angel Resource Institute, 2016. United States figures. Gross, before fees, and survivorship biased: these studies observe exits and cannot see the investments that never reached one. Growth ventures only.
Karlsen, Kisseleva, Mjøs and Robinson, NBER Working Paper 33231, December 2024. International figure. The only angel return study built on population tax records instead of a survey of angels who chose to answer, which is why its numbers are worse than the American ones. It was a working paper at the time of writing and had not completed peer review.
None of that was about the return.
It is why the support and the founder cover exist, and why this is organised instead of a list of people writing checks alone.