Fintech Must Prove Its Impact

For years, fintech growth was relatively easy to describe.

More users. More transactions. More loans. More merchants. More financial activity moving onto digital platforms.

Regulators are increasingly asking a harder question:

What actually changes in the real economy after all that growth?

At FEKDI x IFSE 2026, Indonesia’s Financial Services Authority argued that financial innovation should ultimately improve the real economy by supporting areas such as SME finance, housing, downstream industries and infrastructure, lowering transaction friction and reaching underserved communities.[1] Bank Indonesia made a similar argument: digital transformation needs to move beyond transaction volumes toward productivity, value creation and inclusive, sustainable growth.[2]

The industry is moving from asking “How big is it?” toward “What good does it actually do?”

Scale still matters—but it is incomplete

User count, merchant count, disbursement, transaction value and loan growth remain useful indicators.

They prove adoption and scale.

They do not automatically prove better economic outcomes.

A lending platform can grow disbursement rapidly while borrowers remain financially fragile. A payments platform can process billions of transactions without materially improving merchant productivity.

Scale describes activity.

Impact describes change.

Digital lending is already substantial

OJK reported online-lending outstanding financing of Rp105.63 trillion in July 2026, up 24.76% year on year. Aggregate TWP90 stood at 4.32%, compared with 4.26% in June.[3]

Those figures reveal two simultaneous realities.

The sector continues to grow quickly.

Credit quality still needs to be watched closely.

Higher financing volumes are not inherently positive if borrower dependency or delinquency deteriorates.

Inclusion needs a deeper definition

Indonesia’s 2025 National Survey of Financial Literacy and Inclusion reported a financial inclusion index of 80.51%, while financial literacy stood at 66.46%.[4]

Fintech lending itself recorded only 4.40% inclusion, compared with 70.65% for banking.[4]

That does not mean fintech has failed.

It means inclusion should not be reduced to the number of apps installed or accounts opened.

The stronger question is whether people previously excluded from useful finance can now access an appropriate service.

New users are not always newly included users

Suppose a customer already has several bank facilities and then takes a fintech loan.

That is fintech adoption.

It may not represent additional financial inclusion.

By contrast, a small business previously rejected because it had limited collateral but strong transaction data may represent genuine additionality when a new scoring model gives it appropriate access.

The distinction is increasingly important.

A five-layer impact framework

One practical way to evaluate financial innovation is through five dimensions:

Reach → Cost → Usage → Outcome → Resilience.

This is not an official regulatory standard. It is a business framework for separating platform growth from economic impact.

Each layer asks a different question.

1. Reach: who is genuinely being reached?

Reach is not simply user count.

The more important question is whether the product reaches people facing genuine financial friction: small firms without traditional collateral, communities with limited banking infrastructure, informal businesses with usable transaction histories, or financially viable borrowers who do not fit conventional underwriting models.

SNLIK data still show geographic differences. Financial inclusion reached 83.61% in urban areas and 75.70% in rural areas, while financial literacy was 70.89% and 59.60%, respectively.[4]

Digital distribution does not automatically make access equal.

Access quality matters too

A customer may be able to open an account quickly but still receive an unsuitable product.

Are fees clear?

Are repayment obligations understandable?

Can users obtain support when something goes wrong?

Does the product solve the customer’s problem at a reasonable cost?

Technical availability without economic suitability is weak inclusion.

2. Cost: does the innovation reduce friction?

Fintech often claims efficiency through automation, APIs, digital distribution and data.

Those claims should be measurable.

In lending, companies can examine approval time, documentation burden, acquisition cost and borrower cost after adjusting for risk.

In payments, the calculation needs to include reconciliation, settlement, fraud, cash handling and accounting effort—not only transaction fees.

Subsidised prices are not structural efficiency

Digital platforms frequently accelerate adoption through discounts, cashback, free transfers and promotional pricing.

Those strategies are legitimate.

But they can hide the underlying economics.

A more demanding test asks whether the service remains cheaper or better after subsidies disappear.

If the answer is no, adoption may have been purchased rather than earned through structural efficiency.

3. Usage: active behaviour matters

A dormant account has limited impact.

A merchant receiving one digital payment per month has a different economic relationship with digital finance than a merchant using it daily for operations.

Repeat borrowing also needs careful interpretation.

It may indicate strong product value.

It can also indicate a borrower repeatedly refinancing a persistent liquidity shortage.

Usage needs context.

Engagement is not automatically good

Financial products are not social media.

More engagement is not always positive.

A savings product may be successful precisely because users automate healthy behaviour and rarely interact with the interface.

An investment platform can have high activity while encouraging excessive speculation.

Fintech therefore needs product-specific measures of healthy usage rather than generic screen-time metrics.

4. Outcome: what changes economically?

This is the hardest layer.

If an SME receives Rp50 million in financing, what happens six months later?

Does inventory turnover improve?

Does revenue grow?

Is margin sustainable?

Does the company hire staff?

Does the borrower eventually graduate into cheaper funding?

These outcomes are harder to measure than disbursement, but they are much closer to the purpose of inclusive finance.

Correlation is not causation

A borrower’s sales may grow after receiving financing.

That does not prove the loan caused all the improvement.

Sector demand may rise. A new customer may arrive. A location may improve. Competitors may leave.

Impact measurement needs discipline.

Cohort analysis, pre-and-post tracking, comparable customer groups and longitudinal data can provide better evidence than marketing anecdotes.

Payments need outcome metrics too

Digital payments are often measured by transaction count.

Merchants may need a different set of measures.

Did reconciliation become faster?

Did cash leakage fall?

Did checkout improve?

Did transaction records help inventory management?

Did better records improve access to financing?

Replacing cash with digital payments creates adoption. Improving business processes creates deeper impact.

5. Resilience: benefits must survive stress

OJK lists cybersecurity and consumer protection among its four priority areas for digital finance, alongside adaptive regulation, talent development and coordination.[1]

That matters because a product that grows rapidly while generating fraud, misuse of data, over-indebtedness or operational fragility is not a fully successful innovation.

Resilience means the model remains useful when conditions become difficult.

Trust is part of fintech infrastructure

Financial technology depends on users handing over identity, financial records, transaction data and money.

Trust can take years to build and one serious incident to damage.

Consumer protection therefore cannot be a compliance layer added after product development.

It belongs inside the design.

Sandboxes should test usefulness as well as feasibility

OJK’s regulatory sandbox has processed a growing set of technology models, including tokenisation, digital-asset custody, stablecoins and other financial innovations.[5]

Testing whether a technology works is essential.

A second question is equally important:

Does the business model create enough social and economic value to justify its risks?

Innovation should not receive value merely because the underlying technology is novel.

Start from the problem, not the technology

Technology companies are naturally excited by capability.

“We can tokenise this.”

“We can score this with AI.”

“We can connect this via API.”

A stronger product-development process begins elsewhere.

What problem remains unresolved?

Who experiences it?

Why are existing solutions insufficient?

How much friction can be removed?

Without a strong problem statement, new technology can become a feature in search of a purpose.

Alternative credit scoring is a useful example

OJK has brought Innovative Credit Scoring models into a more formal registration and supervisory framework following earlier sandbox development.[6]

The potential is substantial.

Alternative data may help lenders understand people and firms with limited traditional credit histories.

But prediction accuracy alone is not enough.

Do previously invisible borrowers receive fair access?

Does bias remain controlled?

Is the data legally and ethically sourced?

Can customers understand how consequential decisions are made?

Accuracy and fair access are different metrics.

Investors need different questions too

Technology investors traditionally focus on addressable market, users, take rate, customer acquisition cost and lifetime value.

Financial technology needs another layer:

risk-adjusted economic value.

Loan growth with worsening credit quality is not equivalent to healthy loan growth.

Customer acquisition driven by permanent subsidies is different from adoption driven by genuine utility.

Unit economics and social usefulness need to meet

A product cannot survive indefinitely if it delivers social value but loses money on every customer.

The opposite is also true: a highly profitable fintech serving only customers already well served by traditional finance may create limited additional inclusion.

The strongest models sit at the intersection:

commercially sustainable and economically useful.

That is harder than optimising one metric.

Government digitalisation provides another test case

Bank Indonesia reported that 518 of 546 regional governments, or around 94.9%, had reached the Digital category in regional-government transaction electronification by the first half of 2026.[7]

That is a major adoption milestone.

The next question is not simply how to reach 100%.

It is whether digitalisation improves collections, reduces leakage, accelerates public service, strengthens data quality and enables better fiscal decisions.

Adoption eventually needs to become outcomes.

AI raises the measurement bar

Fintech increasingly uses AI in fraud detection, credit scoring, service, recommendations and operations.

AI may create substantial efficiency.

It can also make decisions harder to explain.

If a company claims AI improves inclusion, the evidence should eventually show who gained new access, how credit performance changed and whether customer treatment remained fair.

“Powered by AI” is not an impact metric.

Financial innovation is not a feature race

Fast-moving industries naturally fear falling behind competitors.

But consumers do not need the maximum number of financial features.

They need problems solved.

Reliable transfers.

Affordable and understandable credit.

Safe savings.

Clear insurance.

Transparent investment.

The strongest fintech businesses may not be those with the most features, but those with the clearest utility.

An impact dashboard does not need to be complicated

Companies can start with a small set of measures.

Reach: share of genuinely underserved or new-to-formal-finance customers.

Cost: time, friction and financial cost before and after the solution.

Usage: active and healthy repeat behaviour.

Outcome: business, household or financial improvement.

Resilience: complaints, delinquency, fraud, downtime and security incidents.

Metrics should differ by product.

The important point is that management sees more than growth.

Impact metrics can become vanity metrics too

“SMEs served” sounds impressive.

But how is “served” defined?

Does one payment count?

Does a company with three existing bank loans qualify as financially excluded?

Definition discipline matters.

Otherwise impact reporting simply replaces one set of vanity metrics with another.

From financial inclusion to financial mobility

The first stage of transformation is access.

The second is usage.

The more meaningful stage is progress.

A small business becomes more bankable.

A merchant gains better records.

A household builds emergency savings.

A borrower eventually qualifies for lower-cost capital because its financial history improves.

That is the difference between financial access and financial mobility.

FEKDI signals a change in expectations

The regulatory message at FEKDI x IFSE 2026 was unusually consistent.

OJK connected innovation to the real economy.

BI connected digital transformation to productivity and value creation.[1][2]

That means the next phase of Indonesia’s fintech industry may not be defined simply by who processes the most transactions.

The harder question will be:

After the technology arrives, who becomes more productive, more resilient and more capable of moving forward economically?

If fintech can demonstrate that, it has created impact.

If not, perhaps only the technology has grown.