--- title: "Market Data API or Excel Exports for a Fintech MVP?" canonical: https://wavect.io/blog/market-data-api-or-excel-fintech-mvp/ language: en description: "Does your fintech pilot need a market data API? See how Wavect built a bond pricing report from quarterly Excel exports and when an API upgrade earns its cost." image: "https://wavect.io/img/blog/headers/header_market-data-api-or-excel-fintech-mvp.png" author: ["Kevin Riedl"] published: "2026-11-11" updated: "2026-11-11" reviewed: "2026-11-11" --- # Does a fintech MVP need a market data API, or will Excel exports do? [**Back**](/blog/overview/) [![Kevin Riedl](/img/team/kevin.webp)](/team/kevin-riedl/) [Kevin Riedl](/team/kevin-riedl/) https://linkedin.com/in/wsdt 4 min read · 11 Nov 2026 Last reviewed November 11, 2026 [**Next**](/blog/paid-pilot-vs-poc-vs-design-partner/) TL;DR A licensed Excel export can support a fintech MVP when the data arrives on the schedule the analysis needs. In our bond analytics pilot, asset managers received a PDF assessing issuance timing and pricing. We would reconsider the API when it buys useful freshness, correction history, less manual work or rights the product needs. Was the bond issued at a good time and at a fair price? Who got the better deal, the issuer or the buyer? An asset manager could open our first PDF report and explore those questions. The spreadsheet stayed in the backend. ## What did the first enterprise pilot need to prove? Our startup client wanted asset managers to assess a bond issue after the fact using a proprietary model with a scientific basis. The customer already had a licence for quarterly Excel exports. API access cost more than the startup could spend. Those were the working conditions the client brought to Wavect. We started with a report. The model processed the figures, and the PDF gave the reader an assessment of the timing and price of the issue. That gave asset managers something concrete to judge. Was the analysis sound enough to help them understand the terms the issuer and the buyers had received? The dashboards in the web application came later. Our [Bond Analytics case study](/case-studies/bond-analytics/) documents the platform behind that first report. ## How fresh did the underlying data need to be? The particular issuance dataset came out once a quarter. Between releases, the published figures stayed the same. A daily API call for those same figures would still have returned the latest available quarter. A desk making an intraday trading decision has a different question and would need to check the update schedule of its own dataset. For a similar pilot, I would write down the dataset, the intended readers and what they will receive. Then I would ask the provider to confirm the permitted use of that input and the resulting report. [LSEG’s description of market data redistribution](https://www.lseg.com/en/data-analytics/market-data/data-redistribution) explains why customer-facing analytics and derived outputs belong in that conversation. The agreement with your own provider determines what you can ship. ## What makes a quarterly file dependable enough for a report? The client supplied the file to the backend each quarter. As I remember it, the format stayed fairly consistent. Our import applied the same rules to each delivery before the analytical model ran. A stored quarterly snapshot let the team revisit the figures behind an earlier rating. A reader could receive a new analysis of the latest release without someone having to hand-edit a spreadsheet for that reader. One test I would ask a team to demonstrate is importing the same quarter twice. Does the second run leave a single set of observations? Then introduce a row without an issue identifier. Can the operator find the rejected row before anyone sends a report? These are two small demonstrations a buyer can witness. [PostgreSQL’s guide to constraints](https://www.postgresql.org/docs/current/ddl-constraints.html) explains how the database can enforce required fields and unique keys. The team still has to decide which fields identify an observation in its dataset. ## When is a market data API worth paying for? I would revisit the API quote when a change in delivery would improve a decision the customer makes or remove a recurring operational cost. Ask the vendor to show you which change the upgrade buys. For example, [S&P Global lists several delivery options for Credit Analytics](https://www.spglobal.com/market-intelligence/en/solutions/products/credit-analytics), including an Excel plug-in, APIs and bulk feeds. A delivery option and a dataset’s publication schedule are two separate items on the quote. | Buying signal | A file fits when… | Revisit the API when… | | --- | --- | --- | | Freshness | The report uses the last published quarter. | Earlier observations are available and would change a customer’s decision. | | Delivery effort | One operator can collect and validate each scheduled release. | Measured handling time or missed deliveries justify automation. | | Corrections and replay | You can retain permitted versions and reproduce an earlier report. | The provider offers correction notices or revision history the file workflow lacks. | | Usage rights | The agreement covers the data, report and pilot audience. | A different agreement covers an output or audience you need to serve. | ## What should a founder bring to the scoping call? Bring a sample you can share, the relevant licence terms and the question the first customer wants answered. Sketch page one. Pick a historical issue and ask what the reader ought to learn about its timing and price from that page alone. That conversation sets the scope for the model, the import and the first deliverable. It also gives you a concrete reason to ask for an API quote later. Our [guide to paid pilots, proofs of concept and design partners](/blog/paid-pilot-vs-poc-vs-design-partner/) covers choosing a commercial test. If the data workflow is the open question, our [software discovery guide](/software-development-guide/what-is-a-discovery-phase/) explains how to scope it. For an upcoming institutional pilot, [Wavect’s MVP development team](/services/mvp-development/) can help turn a sample and a buyer question into a first report. [Discuss the scope of your pilot](/contact/?intent=mvp). ## Questions about a file-based fintech pilot ### Did the asset managers have to open the spreadsheet? The customer supplied Excel exports to the backend. Asset managers first saw a PDF with the model’s findings. Later they could explore the analysis through dashboards in the web app. ### Would an API have made these quarterly figures fresher? For this dataset, new figures became available each quarter. An API could have shortened the time between publication and import. Earlier observations or interim revisions would have required checking what the provider released. ### Would a CSV export change the decision? A CSV changes how the import reads the fields. The report still depends on the publication schedule, stable identifiers, permitted uses and the ability to trace each result to the version received. ### What exactly did the first PDF assess? It helped an asset manager look back at a bond issue, assess its timing and pricing, and consider whether the terms favoured the issuer or buyers. The assessment came from the proprietary model. ### What should we ask the provider before pricing an upgrade? Ask for a sample, publication and correction schedules, available history, coverage and terms for your intended readers and outputs. Request an itemised comparison for your existing file delivery and the proposed API. Product strategy ## Read next Scope, product-market fit, credible releases and the decisions that shape a useful product. [Start here**A Measurable Road to Product-Market Fit**](/blog/road-to-product-market-fit/) - [Pivot the MVP or change the user flow? The Offlinery decision](/blog/mvp-pivot-or-change-user-flow/) - [Is AI Bringing Waterfall Back? How Much Upfront Planning Is Enough?](/blog/ai-waterfall-upfront-planning/) - [ChatGPT Search Console: what the reported Business Hub means for your business](/blog/chatgpt-search-console-business-hub/) - [Enterprise software: settle the business rules before AI writes the code](/blog/enterprise-business-rule-conflicts/) - [Scope vs Armature: Measure Agent API Adoption Beyond Mentions](/blog/scope-vs-armature-agent-api-adoption/) [Previous article**Pivot the MVP or change the user flow? The Offlinery decision**](/blog/mvp-pivot-or-change-user-flow/) [Next article**Is AI Bringing Waterfall Back? How Much Upfront Planning Is Enough?**](/blog/ai-waterfall-upfront-planning/) [**Back**](/blog/overview/) [![Kevin Riedl](/img/team/kevin.webp)](/team/kevin-riedl/) [Kevin Riedl](/team/kevin-riedl/) https://linkedin.com/in/wsdt 4 min read · 11 Nov 2026 Last reviewed November 11, 2026 [**Next**](/blog/paid-pilot-vs-poc-vs-design-partner/)