The State of Data Delivery 2025

1. Overview

Welcome to the first Monda Data Provider Market Report on The State of Data Delivery 2025.

Data delivery is one of the most critical processes for any data provider. When it’s done well, it ensures you’re satisfying customers, maintaining data security, and keeping your bottom line in check. How? For these reasons:

• Delivering data products to customers on their preferred platforms, in a timely manner, increases their loyalty to you
• Sharing data using security-accredited technology means you avoid costly data breaches
• Cost-effective data delivery means you avoid spending masses on engineering staff and data sharing software.

However, many data providers struggle to optimize their data delivery process, focussing instead on the preceding stages in the data commerce journey. With lots of industry attention going to data productization and data marketplaces, there’s a tendency for data providers to devote lots of resources to developing and marketing their data products and fewer on delivering data successfully.

To better understand these trends and challenges, it’s important to first define what data delivery actually involves—and why it plays such a central role in the success of data providers.

1.1. What is Data Delivery?

In data commerce, data delivery is the process by which data providers package and transport data from their systems to their customers. Think of it like a wholesaler shipping goods—except instead of pallets and boxes, it’s datasets and files. Just as a wholesaler must prepare orders in the right format, quantities, and schedule for each customer, data providers must deliver data in the right structure, frequency, and destination.

Whether it's streaming data to a live dashboard or sending a weekly CSV file to a cloud bucket, the goal is the same: ensure the right data reaches the right place, in the right way, and at the right time.

‍

1.2. Why Data Delivery Matters

Data delivery is where all the upstream work of collecting, preparing, and analyzing data becomes valuable to the end user. It’s the final leg of the data commerce journey—the moment the product reaches the customer. If delivery is delayed, inconsistent, or too costly, it erodes trust and satisfaction, just like unreliable shipping in retail.

With customers using a wide range of tools and platforms, each with their own requirements, and expecting data in specific formats and timelines, efficient and reliable delivery is critical.

It’s not just a technical task—it’s a key business function. The smoother and smarter the delivery process, the more scalable, profitable, and competitive a data provider can be. ‘The State of Delivery 2025’ reveals important trends when it comes to how providers deliver data today.

‍

2. Summary & Findings

Customers expect more flexibility of data destinations, ad-hoc data access, and customized delivery—and demand for unstructured data is growing thanks to AI.

Highlighted charts:

2.1. Key Findings

Our research investigates how 27 leading data providers deliver data. Using unique first-party insights, the report uncovering trends in data delivery & sharing, and can help data providers benchmark their own delivery capabilities. These five insights stood out most:

• There's a clear trend that adoption of data APIs and of cloud data warehouses is growing
• Still, a third of data providers are storing data full on-premise and are not cloud-native
• Demand for unstructured data has boomed—probably catalyzed by the need for AI training data
• Customizing data for each customer is one of the biggest challenges which data providers encounter when delivering data
• Only 20% of data providers are working with external tech vendors to support them with data delivery

‍

2.2. Executive Summary

Our survey of commercial data providers confirms that data delivery is one of the most fundamental processes for data-as-a-service companies, and that increased demand for data will only emphasise the importance of robust data delivery in the future.

Though our sample size of 27 companies is small and a larger survey would substantiate our findings, we nonetheless see a clear trend: data providers are currently handling delivery in similar ways. This is despite their different customer profiles and data products. As their customers expect data be delivered across different cloud destinations, sometimes even in real-time and in formats from CSV to Parquet, around half of providers are leveraging new data management, warehousing, and delivery software to keep up.

However, data delivery remains time- and cost-intensive for 40-50% data providers. Tech vendors can provide much-needed solutions which solve data providers' pains associated to data delivery, which in turn facilitate an open, cross-cloud data sharing ecosystem.

Over 20% of data providers are making it a 2025 priority to support more destinations and delivery methods. We expect that data APIs will be even more commonplace by the end of the year, and that most cloud data warehouses will be supported by 25-50% data providers.

A crucial insights is that, in spite of various ways to approach data delivery and its challenges, data providers rarely let delivery compromise a customer's order. When there is a deal on the table, commercial data providers ensure that they can deliver.

‍

‍

3. Data Warehousing & Storage

Data warehouses and data storage are essential for data providers because they’re the backbone for organizing, storing, and managing vast amounts of structured and unstructured data. Sticking with our wholesale anaology, the first step for a wholeseller ensuring that the goods are delivered successfully to a customer is locating the ordered goods quickly and finding them in a sellable condition. In other words, they need a well-structured, secure warehouse and storage facility for their goods.

Data warehouses and data storage systems work in the same way. These systems allow providers to efficiently consolidate data from multiple sources, ensuring consistency, accuracy, and accessibility when they’re preparing data for delivery.

A well-structured data warehouse supports complex querying and analytics, enabling faster insights and better decision-making for clients. In the data delivery process, these storage solutions play a critical role by ensuring that data is clean, reliable, and ready for distribution. Without robust data warehousing and storage, maintaining data quality and meeting client expectations would be difficult, if not impossible.

We investigated which data warehousing and storage solutions data providers are using for which data types, to understand this first crucial step in the data delivery process better.

‍

3.1. Cloud Data Warehouses

33% of data providers do not use cloud data warehouses, instead opting for on-premise solutions/their own servers. Of those leveraging cloud data warehouses, 22% were using Snowflake, and 15% Google Cloud BigQuery.

Databricks was less popular amongst respondents as a data warehouse solution, with 4% answering that they use it as their primary cloud data warehouse (although as we’ll see later, many more providers support it as a delivery destination).

It’s notable how many data providers are not yet cloud-native. Additional research would need to be carried out to ascertain the reasons for this: security concerns, the hassle of migrating from legacy setups, or reluctance to pay cloud service providers—all are possible explanations.

Also unclear is the impact on-premise warehousing has on a data provider’s ability to delivery data to different cloud destinations. Though most cloud destinations are well-covered by data providers, it’s possible that data providers could improve time-to-deliver and reduce costs by using clouds for both warehousing and for delivering.

‍

3.2. Data Management

MySQL was by far the most-used data management technology, with 63% of data providers answering that they used it. MySQL was followed by Postgres, which is used by 41% of providers. Both MySQL and Potgres are open source relational databases, unlike the third most-used technology, Apache Hudi, a data lakehouse platform.

‍

3.3. Cloud Data Storage

Amazon S3 is still the uncontested leader in cloud storage, with 59% of data providers using its file storage solution. 19% of data providers use Google Cloud Storage— yet the same percentage of providers host their data fully on-premise.

Microsoft Azure Blob storage comes in 4th most-used, with 15% of data providers answering that they used it, in line with the previous two questions, where Microsoft’s cloud data warehouse and data management solutions were 5th most popular behind alternatives.

‍

3.4. Data Formats

Overwhelmingly, data providers are using file formats compatible with smaller data transfers: 82% use CSV and 74% use JSON. For big data use cases, Apache Parquet leads the pack, with 37% of respondants saying they use this file format.

Also notable is that 4% of data providers are using WAV as a format to deliver audio files. It’s possible that this is connected to increased demand for unstructured data (audio, image, and textual data) as companies need it to train generative AI models.

‍

3.5. Data Types

As we saw in the previous question, demand for unstructured data is on the rise, with 48% of data providers delivering textual data, 11% delivering image data, and 4% delivering audio data. All the same, tabular data remains the dominant data type, with 93% data providers using it.

In summary, the data warehousing and storage landscape in 2025 reveals a sector in transition, with providers using both legacy and next-gen methods for delivering data. While many data providers still rely on on-premise solutions, cloud adoption is steadily increasing, with Snowflake and Google BigQuery leading the way among cloud warehouses. MySQL and Postgres dominate data management, and Amazon S3 remains the top choice for cloud storage. Despite the rise of cloud-native tools and unstructured data types, tabular data and lightweight formats like CSV and JSON continue to be the norm.

These trends suggest that while the industry is beginning to embrace more scalable, cloud-first solutions, legacy systems and traditional formats still hold significant ground.

{{cta-1}}

‍

4. Data Destinations, Volume & Frequency

In the process of data delivery, data destinations, volume, and frequency play a critical role, much like a wholesaler managing shipments of goods to customers across various countries. Each customer may require different amounts of product, delivered at different intervals, and each destination may have its own import regulations and logistical constraints. Similarly, data providers must consider how much data needs to be delivered (volume), how often (frequency), and where it needs to go (destination).

Some delivery methods, such as APIs or streaming services, are optimized for real-time or high-frequency, low-volume data, while others like file transfers or batch uploads are better suited for large volumes delivered periodically. Additionally, customers are spread across a range of platforms and tools—such as cloud storage, databases, or analytics dashboards—each with specific integration requirements.

Understanding and adapting to this diversity is essential for ensuring smooth, timely, and reliable data delivery.

In this chapter, we look at the most in-demand destinations, the most-used delivery methods, and the volume & frequency of data delivery.

‍

4.1. Data Destinations & Delivery Methods

Shockingly, 70% of data providers are still delivering data to customers using email, in spite of how manual, restrictive (in terms of file size and type), and not secure this method is. By way of explanation, 11% of providers are prioritizing supporting delivery via Amazon S3 in 2025/2026, 8% via SFTP, and a huge 22% are prioritzing API delivery. This shows that data providers are moving away from legacy methods like email, and towards more robust methods.

And many data providers already are using such robust methods. Currently, 67% use Amazon S3, 63% use APIs, 59% use SFTP.

Snowflake, Google Cloud Storage, and Databricks are also popular delivery methods. Interestingly, although Databricks is the least-used of the three at the moment, it was the method which the most data providers say they want to use in the future. If their priorities stay the same, we can expect to see 77% of data providers delivering via Databricks in future.

‍

4.2. Data Volume

There’s a noticeable split between data providers sending ‘small data’ and those sending ‘big data’. 19% of respondents send under 1GB of data per month, another 11% send between 1 and 10 GB, and another 7% send between 11 and 100GB, meaning 37% of providers in total send less that 100 GB of data each month.

In contrast, 45% of providers send at least 1 TB a month, with 19% of respondents saying that they share over 10 TB of data with customers each month.

No providers fell in the middle of this ‘small vs. big’ data split. A larger study into data volume would establish whether or not this pattern is replicable, and therefore whether we can broadly group providers in two according to the volume of data they’re delivering each month.

‍

4.3. Number of Customers

Interestingly, there’s another—albeit less pronounced—split between data providers which deliver data to maximum 100 customers a month, and those which send to over 1,000.

It’s not uncommon for data providers to have a small cohort of monthly customers, with 33% of providers sharing data with just 10 different customers a month—in fact, most respondants answered that they have fewer than 10 monthly customers. This shows that both the high velocity, low ACV model of data-as-a-service (DaaS) works just as much as low velocity, high ACV does.

All the same, there’s a clear need for automation when it comes to data delivery: 38% of providers share data with at least 101 customers a month. Handling all of these different orders manually comes with significant DataOps strains.

‍

4.4. Data Delivery Frequency

It’s most common that providers delivery data on a daily basis, with 26% delivering at this frequency. The second most-common frequency was one-off, suggesting that single file sends are still popular—which makes sense, considering how often data providers must share one-time files with prospective customers to sample. It also shows that though the DaaS subscription model is widespread, one-time purchases are still commonplace.

Notably, only 4% of providers support real-time data delivery. It’s worth monitoring whether this number increases over the years as more real-time data streaming vendors pop up to make instantaneous order-to-fulfilment—plus constant updates—a viable reality for data providers and their customers.

‍

4.5. Data Delivery Teams

Reinforcing that data delivery is a fundamental process for data providers, 59% of respondents told us that they have a dedicated team for handling it.

‍

4.6. Tech Vendor Adoption

For the most part, however, data providers handle data delivery in-house. Only 20% of respondents said that they work with a data sharing tech vendor to assist their data delivery. The data providers which answered ‘Yes’ are using Amplify and Snowflake, showing that both new players and industry veterans are playing a role in shaping the state of data delivery.

‍

{{cta-2}}

‍

5. Key Challenges of Data Delivery

Although it's a fundamental part of the value chain, data delivery is a process that puts significant strain on data providers. Much like a wholesaler trying to deliver custom orders to multiple customers on time while staying profitable, providers face challenges around customization, timelines, and cost. Each client may require a different “package” of data—specific formats, frequencies, or integrations—which adds complexity and effort to the process.

Meeting delivery timelines is critical, especially when clients depend on fresh data for real-time decisions, but doing so consistently can strain resources. Supporting multiple destinations, each with unique technical requirements and standards, also increases the operational burden.

Balancing all of this—customization, speed, and accuracy—while managing costs and maintaining efficiency is a constant challenge.

That’s why it’s essential to understand these delivery hurdles, so vendors can build better technologies and processes that enable more time- and cost-efficient data delivery. Read on to learn which challenges are most pressing for data providers when it comes to data delivery at present.

‍

5.1. Data Delivery Challenges

Customization emerged as the most challenging aspect of data delivery. Over half (56%) of respondents agree that they need to customize data for every delivery, pointing to the complexity and manual effort still involved in meeting varied client needs.

Time and cost are also major concerns. Around 41% say they spend too much time on data delivery, and 33% feel the need to cut costs—indicating that efficiency and scalability remain difficult to achieve.

Supporting multiple destinations is another strain, with 37% agreeing it requires significant effort. This reinforces the need for cross-cloud and cloud-agnostic data delivery solutions, which solve this pain.

Despite all of these pains, however, it seems that, when a deal is on the table, data providers will deliver to get it over the line and close successfully. Only 15% report losing business due to unsupported destinations, suggesting providers are managing to meet demands, but possibly at a high operational cost.

‍

5.2. Cost of Data Delivery

This report is the first of its kind to provide a benchmark into the cost of data delivery, and it shows that 26% of providers spend over $132K per year just on this process. Though this value is relative to other business costs and business revenue, it’s a sizeable chunk of annual expenditure.

Interestingly, 30% of respondents answered ‘Don’t know’, suggesting that there’s a lack of straightforward pricing for data delivery. It’s difficult to calculate the cost of engineers’ time, compute, and infrastructure, which could explain why over a third of data providers seemingly have no dedicated budget for data delivery. But when it’s such a fundamental—and potentially expensive—process, it seems there’s need for more transparency around exactly how much data delivery costs data providers.

‍

5.3. Time Spent on Data Delivery

The majority of data providers are spending a notable, yet manageable, portion of their time on data delivery tasks. 77% of respondents report that their teams spend up to 25% of their time on data delivery, with 44% falling into the 11–25% range.

However, a significant minority—19%—say they spend between 26% and 75% of their time on delivery, suggesting that for some teams, data delivery is a much more time-intensive part of their operations. This reinforces the idea that delivery workflows can become a bottleneck, especially when scaling up or supporting more complex customer needs.

The absence of any respondents spending more than 75% of their time on delivery suggests most teams have found ways to prevent it from completely dominating their workload. Still, the time spent—even at lower percentages—represents a non-trivial operational cost. These insights highlight the opportunity for tools and automation that reduce manual delivery effort and help teams reclaim valuable time.

‍

6. Outlook

As data delivery continues to evolve, we can expect to see a new wave of solutions emerge—designed to help providers manage this critical process with greater precision, speed, and scale.

The growing demand for seamless, secure, and efficient data sharing will push the industry toward more standardized practices and innovative delivery models.

As the landscape shifts, Monda will continue to publish benchmark reports, offering ongoing insights into how the data provider market is adapting and setting new standards for the future.

We'd like to thank the data providers who told us how they're handling data delivery so that we could create this report.

Your feedback helps others understand the data delivery market so that we might drive further progress in creating an open data sharing ecosystem.

Deliver data easily using Monda: monda.ai/features/data-sharing

‍

7. Methodology

We surveyed 27 employees of data providers with 71% in tech roles and 93% with management responsibility

All respondents to our survey work for data providers (i.e. companies whose primary revenue is generated by selling data products). These companies are referred to as 'data providers' throughout this report. The data providers are based in different countries and offer different categories of data.

Of the individuals we surveyed, over 70% worked in the Product & Tech team at their company. 41% held VP, Head, or Lead positions, and 41% were C-level.

In cases where data providers answered 'Not sure', we included these responses in our analysis.