Today’s sourcing should not exist.

Whenever you hear “relationship-driven process” or “proprietary sourcing” described as a source of returns, think of another word: inefficient.

A banker should not pitch a company to a firm because of a good relationship with its investors. A banker should pitch a company to a firm because it fits that firm’s mandate, and the firm has the operating expertise to help it grow. A sale should reach every buyer for whom that is true.

It doesn’t, because reaching buyers is expensive. A company worth hundreds of millions is priced by a handful of firms. The cost shows up in the price: private companies sell at a 15 to 30 percent discount to comparable public companies, because sellers in a thin market have no credible alternative to the buyer in front of them.¹

Buyers bear the cost too: for every hundred companies a private equity firm considers, it closes about six.² The rest is search. Search costs keep the market thin, and thin markets hand companies to the wrong owners.

And once a private company has a buyer, it rarely has a second. About 1 percent of independent private targets drew another recorded bidder after the initial bid, against 8 percent of public ones.¹

For a decade, cheap debt hid the problem. When money is nearly free, buying a company at the wrong price still produces strong returns, and buyout pricing tracked credit conditions more than the businesses being bought.³ That era is over, and the inefficiency it concealed is now exposed.

AI removes most of the cost of search and evaluation, and far less of the time needed to close a deal. Firms will screen and value many more companies, while management meetings and final negotiation remain human. When every firm can see every company, no firm wins by finding one first, and sourcing stops being an advantage.

What remains is the only advantage that was ever legitimate: making a company more valuable than it was before. Founders capture more of the price as serious bidders compete for their companies.

The paper ends by asking what would replace sourcing rather than automate it. Vendu is that answer, in three parts.

• Garde

  French for “guard”

  Holds your company's data, handles the buyers who contact you, and represents you when you decide to sell. It answers only to you.

• Ryo

  Japanese: the gold coin of old Japan

  Prices every company the way your fund would, from your own mandate, model and limits, and reads data rooms for your team. It answers only to your fund.

• Agora

  Greek: the open square where people traded

  Where Garde and Ryo meet. It runs the rules of a competitive sale and never negotiates for either side.

1. 01

   Tiered disclosure

   Standardized figures go broadly, because disclosing more verified information raises prices. Sensitive material goes only to invited bidders, through clean teams when they own a competitor.⁴

2. 02

   Agents that answer to their own side

   Each fund defines its own agent: its criteria, its assumptions, its model. A venue that wrote buyers' agents itself would hold every firm's valuation logic and could steer outcomes.

3. 03

   A neutral venue

   Fees that don't depend on the price or the winner, sealed bids, and no negotiating for either side. A venue that held bargaining power itself would recreate the intermediary that keeps prices away from competitive levels.⁵

4. 04

   Live alternatives

   Exclusivity is where sellers lose leverage today. Keeping runner-up bids live through diligence preserves the owner's outside option and limits re-trading.