How performance data replaces relationship memory — the reputation system

The private mortgage market in Ontario runs on memory. A broker placing a deal remembers which lenders funded their last few similar files. A lender reviewing a submission remembers which brokers have sent them clean deals before. Both parties operate from a working list of counterparties built up over years — the lenders the broker has worked with successfully, the brokers the lender has come to trust. Relationships maintain the list. Phone calls, BDM visits, industry events, the occasional dinner. The market clears on the strength of the relationships that participants have built and maintained manually.

This system has worked for a long time. It has also produced specific costs that have grown with the market. A broker's effective lender network is bounded by what they can remember. A lender's effective broker network is bounded by who remembers them. As the market has grown — and Ontario's private mortgage market processed 65,233 transactions worth $32 billion in 2024, nearly triple the volume of a decade ago — the matching infrastructure built on relationship memory has not scaled with it. Brokers are placing more deals through smaller relative networks. Lenders are receiving more submissions from a relatively narrowing set of broker relationships. The familiarity that used to clear the market efficiently is now the bottleneck on it.

The Openfund reputation system is the infrastructure response to this constraint. Every broker and lender on the platform accumulates performance data with every deal — funding times, conditions-to-close rates, broker-lender ratings, deal quality on submission, response time on offers. That data becomes visible to counterparties at the moments when decisions get made. The system replaces what the market currently does from memory with what the platform can do from data. This article walks through what the reputation system tracks, how the data is presented, and why the structural shift matters for both sides of the market.

What the system tracks

The reputation system records three categories of data on every deal that flows through the platform.

The first category is operational performance. For lenders, this means the funding outcomes that follow from their offers — how often offers convert to funded deals, how fast funding occurs from offer acceptance, how consistently the lender's stated funding speed matches the funding speed observed in practice. For brokers, this means the quality and reliability of their submissions — how complete the application package is on first submission, how often the broker's deals close at the lender they originally selected, how the broker handles the period between offer and signing.

The second category is the counterparty ratings. After every deal closes, both the broker and the lender rate the other on a structured scale. The ratings cover specific dimensions — for the broker rating the lender, factors like communication clarity, condition reasonableness, and follow-through on commitments; for the lender rating the broker, factors like submission quality, responsiveness, and post-acceptance behavior. The ratings are not free-form testimonials. They are structured assessments that aggregate into a profile over time.

The third category is the deal-level history itself. Every deal a broker has submitted on the platform is recorded with its outcome. Every deal a lender has made an offer on is recorded with whether the offer was accepted and what happened next. The deal-level history is what allows the system to be evaluated against specifics — not “this lender funds fast” as a general claim, but “this lender funded 47 of their last 50 offers within their stated funding window.”

How the data is presented

The data does not appear as a single reputation score. The platform does not produce a “this lender is rated 4.6 stars” summary that flattens the underlying complexity. Reputation in private mortgage lending is genuinely multidimensional, and a single score would hide the dimensions that actually matter for any specific deal.

What the platform presents instead is dimensional. When a broker is evaluating competing offers on a deal, each lender's offer carries a profile of the lender's historical performance on the dimensions that affect this kind of decision. The lender's average time from offer acceptance to funding. The lender's consistency on funding speed across deals. The lender's history of conditions changes after offer submission. The lender's overall offer-to-funded conversion rate. The broker sees these metrics next to the offer terms, and the metrics inform which terms to trust.

When a lender is reviewing a submission from a broker they have not worked with before, the broker's history is similarly visible. The broker's submission completeness rate. The broker's response time during the underwriting period. The broker's history of accepting offers at the lender they originally selected versus pulling deals to alternative lenders late in the process. The lender sees these metrics before they invest review time, and the metrics inform whether to invest that review time.

The structural shift

The argument for the reputation system is not that it produces better data than the current relationship-based system. The current system, when it works well, already produces high-fidelity data — a broker who has worked with a lender for ten years knows exactly how that lender behaves on the dimensions that matter. The argument is that the reputation system extends what the current system can produce only at small scale to the full market.

In the current system, a broker's effective lender network is limited by relationship memory. The broker knows the lenders they have worked with. They do not know the lenders they have not. When a deal does not fit any lender they have worked with, the broker either places the deal at a less optimal lender they do know, or reaches out cold to a lender they have not worked with and asks them to take the file on faith. Neither outcome is good. The first costs the borrower competitive pricing. The second costs the broker time and may not produce a workable offer because the lender has no basis for evaluating the broker's submission quality.

The reputation system removes this bottleneck. A broker placing a deal on the platform sees offers from lenders they have never worked with, accompanied by the operational data those lenders have built up across their other deals on the platform. The unfamiliar lender is no longer an unknown quantity. Their history is visible. The broker can evaluate the offer not just on its terms but on the lender's reliability of delivering at those terms. The trust that previously had to be established through a phone call, an industry event, or a BDM visit is now available immediately from the deal record.

The same structural shift happens on the lender side. A lender reviewing a submission from a broker they have not worked with previously is operating with no basis for evaluating the broker's submission quality. They review the file with appropriate skepticism, which costs operator time and frequently produces a decline because the lender cannot extend the trust the broker has not earned with them. On the platform, the broker's submission history is visible. A broker who has consistently submitted complete, accurate applications on previous deals arrives with a track record that lets the lender invest review time confidently. The lender's effective broker network expands from the brokers they have personally worked with to the brokers whose data demonstrates they are worth working with.

Why this matters for the market

The market-level consequence of this shift is that matching becomes less dependent on which counterparties happen to know each other and more dependent on which counterparties actually fit the deal. In a system where a broker has access to fifteen lenders' worth of relationship memory, the broker's clients are served by whichever of those fifteen lenders has the best fit for the specific deal — and if none of the fifteen has good fit, the deal is placed sub-optimally. In a system where the broker has access to every eligible lender's operational data, the broker's clients are served by whichever lender on the entire platform has the best fit. The denominator changes from “lenders the broker remembers” to “lenders whose criteria fit.”

The same logic operates on the lender side. A lender's deal flow is no longer limited by which brokers happen to remember them. It is determined by which brokers are placing deals that match their criteria, regardless of whether those brokers have any prior relationship with the lender. The denominator changes from “brokers who remember to send deals” to “brokers placing deals that fit.”

This is not a small efficiency improvement. It is a structural change in how the market matches supply and demand. The relationships do not disappear. The brokers and lenders who have built strong working relationships will continue to benefit from those relationships — the platform does not penalize relationship-based deal flow, it just stops requiring it as the price of access. What changes is that the broker or lender who has not had the time, resources, or network to build extensive relationships is no longer locked out of the rest of the market. The data does the work the relationship used to do.

What this produces for early participants

The reputation system is built around accumulated data. Brokers and lenders who join the platform early begin building their reputation profile before competitors join. The data they accumulate becomes the basis for how counterparties evaluate them as the platform scales — and the historical data that early participants accumulate cannot be replicated by later entrants. A lender who has funded fifty deals on the platform by the time competitors arrive has a fifty-deal track record that those competitors cannot retrofit. A broker whose submission quality data shows ninety-five percent first-submission completeness across two hundred deals carries that record into every new lender relationship they engage with through the platform.

This produces an asymmetric incentive for early participation. The platform's matching engine, the Openfund Report, and the deal room are valuable to every participant equally. The reputation system specifically rewards those who build their profile before the market catches up. The brokers and lenders who define the platform's data layer in its first year are the ones whose reputations will carry the most weight as the platform scales — not because the system favors early users, but because they are the ones with the longest data history.

The reputation system, like the rest of the platform, is infrastructure. The data accumulates whether participants think about it or not. What changes is that the work of building professional credibility — currently spread across years of relationship maintenance, conference appearances, and BDM visits — is now produced as a byproduct of doing the work itself. The deal you fund builds your reputation. The submission you make accurately builds your reputation. The Openfund Report you generate builds your reputation. The platform records what you do, presents it to counterparties at the moments it matters, and lets your work speak for itself in markets you have not yet entered.

Ontario's private mortgage market has tripled in size over the last decade. The infrastructure that supported it at one-third the volume cannot support it at scale. The reputation system is what infrastructure looks like when it replaces what relationship memory used to do — at scale, in real time, with the dimensional specificity that real lending decisions require.

Ringo So is Co-Founder and Head of Strategy and Growth at Openfund. He has built and operated mortgage businesses in Ontario for over a decade across brokerage, lender administration, and platform infrastructure.