Operator Playbook: The Esusu Playbook — Turning Overlooked Data Into Financial Infrastructure

Executive Summary

Esusu is usually described as the company that made rent count toward credit. That describes the outcome and omits the method. The method is a six-step sequence that an operator can test on any overlooked stream of behavioural data: choose a recurring behaviour that institutions ignore, capture the data where it is already recorded, standardise it into a form an institution accepts, make that institution’s decision depend on it, attach further services to the same capture, and measure scale by how many decisions rest on the feed. Esusu is the evidence for each step.

The sequence also has failure conditions: fragmented data holders, institutions that do not change their decisions, and a feed that nobody agrees to pay for. The infrastructure business begins when somebody else’s decision depends on the data.

Why Is “Turned Rent Into Credit” the Wrong Lesson for Operators?

The conventional account holds that Esusu succeeded because rent is a good credit signal. That is directionally true and operationally empty. A signal is not a business. The case for rental data as credit data explains why the opportunity existed; it does not explain why one company captured it, because a good signal that nobody collects, formats and delivers remains a good signal and nothing more.

This article treats the company the way an operator would. It takes the company’s published claims as stated, marks which of them are the company’s own, and extracts rules that can be tested against another dataset. The founder’s path sits in Abbey Wemimo’s founder profile, and the technical chain from lease to credit file sits in the pipeline from lease to credit file; neither is repeated here. The question that remains is what had to be built between a monthly rent payment and a lender’s decision, and which of those pieces transfer.

What Did Esusu Recognise That Others Missed?

The recognition concerned location, not rent. A rent payment is recorded by the party that receives it, the property manager. The institutions that would price the payment, lenders and scoring models, sit elsewhere and could not see the record. Moving the record into a credit file requires the property manager to act as a furnisher, the term for a company that supplies account data to a credit bureau. FICO reports “a general aversion among potential furnishers” to the operational and compliance risks of doing so (FICO’s analysis of 5 March 2026).

Esusu’s own site presents the company as the rent reporting provider for property owners and managers, reporting to Experian, Equifax and TransUnion. The arrangement implies that the property manager does not have to build the bureau connection itself. That is the analysis in this article, not a statement the company makes in those words.

The location insight is the transferable part. Most overlooked data is not hidden. It is held by a party with no use for it and no route to the party that has one. The operator’s first task is therefore to map three positions: who records the behaviour, who would price it, and what stands between them. Where the answer to the third is effort alone, the opportunity is a service business. Where the answer is a compliance burden and a format mismatch, it can become infrastructure. The gap Esusu filled lay between a record and a decision, not between a renter and a score.

What Are the Six Steps of the Esusu Playbook?

The Flutterwave playbook starts from friction that repeats across markets. This playbook starts from a record that already exists and has no reader. Both end in a layer other parties build on, but the entry condition differs, and an operator should know which one they hold. The table sets each step beside a rule that can be tested, the Esusu evidence, and the signal that the step is failing.

Step 1: Choose a Behaviour That Is Recurring, Contractual and Unread

Three tests select the behaviour. It recurs, so the data accumulates without new effort. It is contractual, so the record is dated and disputable. It is unread, so the institution that would price it has no rival source. Rent passes all three. FICO’s analysis puts the number of consumers with a rental tradeline, a rent payment line in a credit file, at 2.7 million of roughly 77 million adults in rental housing, or 3.5%, against 270,000 consumers in 2017 (FICO’s analysis of 5 March 2026). By calculation from those figures, about 96.5% of renting adults in that sample had no rental tradeline.

A one-off purchase fails the recurring test; a behaviour that lenders already read fails the third.

The case that rent carries credit information is set out in the companion analysis of rental data as credit data, and the rule here is narrower: the selection tests come before any product decision. The first step selects a behaviour. It does not yet create a business.

Step 2: Capture the Data Where It Is Already Recorded

The cheapest capture route runs through the party that already holds the record. Esusu’s own site lists real estate clients that include Invitation Homes, Cushman & Wakefield, Nuveen and WinnCompanies, and quotes Patrick Richard of Stoneweg US: “We also immediately decided to offer this service to our residents instead of charging them for it.” The page, as read on 23 October 2026, states no general price for the reporting service, so the quote shows one client’s choice and not a pricing policy. It does show who decides: the property operator chooses whether residents are charged, which makes the operator the customer and the renter the beneficiary.

The rule that follows is to sell to the holder of the record and let the subject of the record pay nothing, or as little as the holder allows. Where the holder is a single-property landlord with no software and no reason to switch, the rule has no one to sell to. Distribution through the data holder is the capture strategy, and it fixes who the customer is.

Step 3: Standardise Into the Form the Institution Already Accepts

Standardisation here is mostly subtraction. Esusu’s own site states that Esusu reports to all three major credit bureaus and “only reports on-time rent payments and does not report missed or late rent payments.” A positive-only feed lowers the renter’s risk of participating and the property manager’s exposure as the supplier of the record. Fannie Mae’s pilot used the same design: Fannie Mae’s pilot announcement of 27 September 2022 describes a “positive-only initiative” in which renters who miss a payment are unenrolled automatically.

The receiving side was also ready. FICO states that rental data has been included in every FICO Score version since FICO Score 9, introduced in 2015, including FICO Score 10T (FICO’s analysis of 5 March 2026). The CFPB’s analysis shows the data can be tested against outside benchmarks: CFPB’s January 2025 introduction to its rental payment data compares contract rent and payment delinquency with other sources and benchmarks renters against American Community Survey data. The detailed chain is in the pipeline from lease to credit file.

For an operator the lesson is about sequence. The scoring side accepted the format before the supply of data existed, so the missing piece was the supplier. The feed that survives is the one with the fewest reasons for an institution to refuse it.

Step 4: Make an Institution’s Decision Depend on the Feed

Integration is the step that turns a data service into infrastructure. Fannie Mae’s pilot announcement of 27 September 2022 states that, beginning 27 September 2022, eligible Fannie Mae multifamily owners could send timely rent payment data through a vendor network to the three bureaus, and that Fannie Mae would cover the costs of collecting and disseminating the data for 12 months for borrowers using an approved vendor. It names Esusu Financial, Jetty Credit and Rent Dynamics as the approved vendors and lets renters opt out.

Two points matter to an operator. First, the institution lowered the cost of trying, so the data holder’s cost of collecting and sending the data was covered during the pilot period. Second, three vendors were approved, so the integration validated the category more than any single company, and differentiation had to come from the next step. The announcement states no terms after the 12 months, so who pays for the feed in steady state is not answered in that source.

A second route shows how integration can bypass a supplier. Fannie Mae’s release of 11 August 2021 describes Desktop Underwriter identifying recurring rent payments in an applicant’s bank statement data, with the applicant’s permission. In the Fannie Mae sample cited there, 17% of applicants who had not owned a home in three years and did not receive a favourable recommendation could have received an Approve/Eligible recommendation had rental payment history been considered. Integration is complete only when removing the feed would change a decision someone else makes.

Step 5: Attach New Utility to the Same Capture

Once the capture exists, the question becomes which other party has a question the same relationship can answer. Esusu’s own site lists identity services (income verification, identity verification, SSN verification and Esusu Passport), says that manual review “often misses fake pay stubs and stolen IDs, leading to bad debt,” and prices protecting net operating income at “as little as $0.99/unit.” It also lists an analytics product that tracks “payments, credit, and portfolio health in real time,” and resident services including Split Pay, Rent Relief and Financial Coaching.

The page does not say which of these products reuse rent payment data. That is the operator’s test, and it is the only one that matters: a second product earns its place if it is sold through the door the first one opened, to the same data holder, on the same capture. A product that needs a new buyer, a new data source and a new integration is a second company. The test of a second product is whether it reuses the first product’s door.

Step 6: Measure Scale by Dependence, Not by Customers

Infrastructure scale means that many decisions rest on the feed. FICO’s figures show how far that is from complete for rental data: 3.5% of renting adults with a rental tradeline, and only 0.07% of scorable consumers, fewer than 1 in 1,300, who became scorable because of rental data (FICO’s analysis of 5 March 2026). FICO writes that rental reporting is growing but has not reached the level where it would significantly move the needle.

The same discipline applies to company metrics. Esusu’s own site reports an average credit score increase of 53 points for active renters at active properties from enrolment through Q4 2025. The footnote defines the method: the score at enrolment is compared with the latest score, and renters with no score at enrolment are treated as having a score of zero. Hypothetical, for illustration: under that method a renter whose first score is 640 contributes an increase of 640.

The figure therefore blends two different events, a score moving and a score appearing, and the page adds that results are not guaranteed. The page read for this piece shows no readable count of renters or units, so this article does not size the company.

Scale in infrastructure is dependence, and dependence is measured from the other side of the feed.

Where Does the Esusu Model Not Work?

The model fails under five conditions, and each is visible before capital is committed.

Fragmented data holders. FICO notes that much of the rental market consists of single-property landlords, which makes national-scale reporting difficult (FICO’s analysis of 5 March 2026). Step 2 needs holders concentrated enough that one sales relationship reaches many records.

Institutions that do not move. A feed that made only 0.07% of scorable consumers scorable has limited power to change lending, whatever its growth rate. Step 4 fails quietly when the institution accepts the data and leaves its decisions unchanged.

A feed nobody pays for. The Fannie Mae pilot covered costs for 12 months, and the source states nothing about the period after. A business whose supply cost is carried by an institutional subsidy has not yet shown that the holder will pay once the subsidy ends.

Bypass by another route. Where an institution can read the same behaviour from permissioned bank statement data, as Desktop Underwriter does, the supplier’s position depends on what it offers beyond the signal itself.

Markets without a data holder. Hypothetical, for illustration only: in a market where rent is paid in cash between individuals with no intermediary keeping records, step 2 has no one to sell to, and no amount of institutional demand creates a record. Whether London or Lagos contains a concentrated data holder and a credit bureau that will accept such a feed is a question this article does not answer, because no UK credit reference agency or Nigerian bureau document was read for it.

A playbook is only as portable as its dependencies, and the dependencies here are concentrated holders, an institution willing to change a decision, and a party willing to pay.

What Should an Operator Test on Monday?

The six steps convert into five questions an operator can put to any dataset, in this order:

• Who records it? Name the party that holds the record and confirm it is concentrated enough to sell to.
• Who would price it? Name the institution whose decision the record could change, and the decision.
• What stands between them? List the format, compliance and trust gaps, and decide which the product removes.
• Who pays once the pilot ends? Identify the party whose budget carries the feed in steady state.
• What else can the same door sell? List the second products that reuse the capture and drop the rest.

Why This Matters for Product Leaders and Capital Allocators

For product leaders, the sequence matters more than the idea. The Esusu evidence shows the receiving side ready before the supply existed, which means the work was capture and standardisation, not persuasion of the scoring model. An operator who begins with a product for the data subject, and not with the data holder, builds step 6 before step 2.

For capital allocators, the diligence questions are different from those for a consumer fintech. Which decisions already rest on the feed? Is the integration specific to the company or shared across approved vendors? Who pays when the incentive ends? These are the questions that separate infrastructure from a service with a good signal. They also fit the thesis in The Rise of Diaspora Operators, where diaspora operators build the layers that others rely on, and Esusu is one case of it. The housing-market consequences of rent reporting are the subject of a forthcoming infrastructure thesis in this cluster.

Key Takeaways

• Esusu’s lesson is a sequence, not a signal: behaviour, capture, standardisation, integration, utility and scale, each with a test.
• The opportunity lay between a record and a decision, which makes the data holder, not the data subject, the first customer.
• Standardisation is subtraction: positive-only reporting to all three bureaus left institutions with little reason to refuse the feed.
• Integration is category-level when several vendors are approved, so differentiation has to come from services that reuse the same capture.
• The model fails where data holders are fragmented, institutions leave decisions unchanged, or nobody pays for the feed once an incentive ends.

Conclusion

Esusu’s story is told as a company that made rent count. The playbook underneath is a way of finding a record with no reader, placing a service between the record and a decision, and building until someone else’s decision depends on the feed. The evidence shows the sequence working in one market and, in FICO’s figures, still early in that market. An operator who applies it should expect the first step to be easy to name and the fourth to be where the business is decided.