ETL pipelines help companies extract, transform, and load data so it is ready to provide insights and value to the company. 

But running a smooth data operation depends on building reliable and scalable data ingestion pipelines.

SaaS vendors like Keboola and Azure Data Factory take away the heavy lifting. 

They build software that collects data, transforms and cleans it, and loads it to a destination of your choice, without having to worry about coding the solution yourself, maintenance, or scaling problems.

In this article, we explore the 8 critical differences between Keboola and Azure Data Factory (ADF), to help you better decide which one is best for your DataOps:

1. Overview - What are Keboola and Azure Data Factory used for?

Both Keboola and Azure Data Factory are ETL tools (or data ingestion tools) that help you speed up the build and later automate your data pipelines. 

Relying on a software vendor for your ETL (extract, transform, load) has multiple advantages:

1. Automated operations. The tool automates all the steps necessary for collecting and cleaning data, so valuable engineering hours are freed up for more critical and revenue-generating work. 
2. Transparency and traceability. By relying on a single tool to carry out all the data pipeline, it is easier to follow the data as it moves throughout your data warehouse. This brings greater transparency to your operations and allows traceability, which is critical for regulatory compliance (GDPR, HIPAA, SOC2, ....) and for debugging (for example, with Root Cause Analysis)
3. Speed. Outsourcing the heavy lifting to 3rd Party Apps means you get data faster (less debugging, no maintenance of tooling), so you can move faster, get the insights before your competitors and iterate on your data products to grab a larger share of the market.
   

Both solutions go beyond simple ETL. As end-to-end platforms, they both offer monitoring, data governance capabilities, and managed DataOps deployment and scaling via the use of cloud technologies. 

2. Integrations - What sources and destinations can you connect in your data pipelines?

The main use of an ETL tool is to facilitate the movement of data from one location to another without much supervision.

ETL tools achieve this with integrations - often called “connectors” or “components”. Each connector is specialized for one data source or data destination. For example, the “Facebook Ads” connector would be specialized for extracting data from the Facebook Ads platform.

How do the two tools compare on integrations?

Quantity. Keboola offers more many integrations (280+) than Azure (<100).

Coverage. Azure Data Factory is almost exclusively focused on covering the Microsoft datascape (integrates with Azure Blob Storage, Azure Data Lake, …) with the rare exception of other vendors. For instance, Amazon Redshift for warehousing, SalesForce for CRM, etc. Keboola, on the other hand, covers the most common data use cases in a company:

1. Datascapes: file systems (Google Sheets, CSVs, Excel, …), data DAMs (DropBox, Google Drive, …), databases (Postgres, MySQL, MongoDB ...), data warehouses (Amazon Redshift, Google BigQuery, Snowflake), data lakes (S3, 
2. CRM and sales software: Hubspot, Salesforce, Pipedrive, etc.
3. Advertising platforms: Facebook Ads, Google Ads, Linkedin Ads, Twitter Ads, CJ Affiliate, BingAds, and others.
4. Marketing communication: CustomerIO, Drip, Mailchimp, ActiveCampaign, Klaviyo, ...
5. Project management: Asana, Jira, Emails, Freelo, Google Drive, OneDrive, etc.
6. Support and customer success software: Delighted, Zendesk, Intercom, Freshdesk, etc.
7. Accounting solutions: QuickBooks, Fakturoid, FlexiBee, …
   

Sometimes you will have to extract data from sources and 3rd Party Apps that are not covered by the vendor. How are the two solutions prepared for custom sources? 

2.1 Custom integrations

Both Keboola and Azure Data Lake offer users the capacity to build custom data extractors. 

Keboola offers its Generic component to import data from almost any REST API and countless other APIs. The component acts as a customizable HTTP REST client. So, you can build your custom data source extractor. The universal extractor does not require programming prowess, you can write your extractor as a JSON.

Azure Data Factory, on the other hand, makes custom integrations available for do-it-yourself type of queries via HTTP endpoints and REST endpoints.

3. Transformations - How do Keboola and Azure Data Factory help you clean and refine your data?

Data transformations include all the data wrangling and cleaning necessary to prepare data for analysis and insights extraction. For instance, aggregating data into metrics, removing outliers and corrupted metrics, standardizing formats, etc.

Both vendors offer a wide variety of transformations in script-form, using a scripting language.

Keboola focuses on the lingua francas of data scientists and data engineers and exposes transformations in the programming language your engineers and scientists love best (SQL, Python, R, Julia, Spark …). 

Azure Data Factory is more imposing regarding transformations. The solution puts PowerQuery as the main transformation tool (better learn it!). You can also pick other Big Data technologies like Spark transformations or Hadoop MapReduce, but you need to configure the transformation environment yourself before accessing the additional languages.

Both vendors understand that there are multiple paradigms and choices of data architectures, so you can run transformations under either the ETL or ELT paradigm. 

4. Advanced Analytics - Which BI and AI software can you use with Keboola and Azure Data Factory? 

Azure Data Factory offers advanced analytics, BI, and artificial intelligence by integrating into Microsoft's app ecosystem. Users can send their data to Power BI for business intelligence or machine learning apps like Azure Cognitive Services (check the full list here).

Keboola, on the other hand, lets the user decide what applications they would love to use for their competitive advantage. 

It offers a wide variety of integrations to BI tools (Tableau, Looker, Power BI, GoodData, Sisense, ...), machine learning applications (like Google NLP, Azure ML, ...), or pre-built scaffolds (automated recipes) that automatically run common machine learning (and data engineering) tasks.

5. Automation - How do Keboola and Azure Data Factory orchestrate and automate jobs? 

Both vendors offer the power of automation to their users via orchestrated jobs. 

The jobs can cover any data engineering task, such as extraction, cleaning and transformations, and loading data to the destination of your choice. 

The orchestration aspect shows maturity in both solutions, since jobs can be scheduled, chained, parallelized, automated, and monitored with little work.

6. Support Quality - What help is offered?

Both vendors give users an extensive set of documentation for self-service support (Keboola’s docs & Azure Data Factor’s docs). 

The main difference is in the technical support - when you request help or submit a ticket. 

Keboola always offers extensive support at no extra charge to you for any custom inquiries. And it is not just free, it is also of superb quality:  users have consistently rated Keboola’s customer service with 5/5. 

Unlike Azure Data Factory that only offers paid support starting at $29/month (“developer” support package for non-production environments or during trial and evaluation) to $1000/month (“professional direct” support package for medium-size to large companies using Azure). The support price tag is hidden under Azure’s service charges.

7. User Satisfaction - What do users say about Keboola and Azure Data Factory?

Both vendors are on top of their fields, with extreme user satisfaction in comparison to their competitors. Looking at the Gartner reviews, Keboola scores a whooping 4.9/5 overall satisfaction, while Azure Data Factory scores 4.6/5. 

8. Pricing - How much money will you spend on Azure Data Factory vs Keboola?

Pricing is a critical factor for making a tooling decision.

Azure Data Factory has a complicated pricing model, where plan costs are computed as a combination of hardware, compute hours, number of orchestrated jobs, and read/write (I/O) operations. The pricing model is hard to understand unless you are familiar with cloud computing cost estimations (you can use a calculator here to make your life easier).

The bottom line is the following:

1. Azure Data Factory is rather expensive in comparison to the competition.
2. There is no free plan. It is pay-as-you-go (PAYG).
3. You are charged for each read/write operation, so do not experiment too much when building pipelines.
   

Contrast this to Keboola.

Keboola offers three plans: an always-free plan ($0/month), a pay-as-you-go-plan (PAYG), and a subscription plan. 

You are charged for the Time of processing. Every account gets 300 free minutes a month (free plan), then pays 14 cents per additional minute. In case your needs are greater, custom subscription plans are available.

Summary - Which tool to pick?

Both Azure Data Factory and Keboola are state-of-the-art tools in the data integration vertical.

We compared both ETL and DataOps tools across 8 dimensions:

Ultimately, the choice of tools will depend heavily on what your enterprise’s needs will be.

But to make the choice easier, here are three important questions to consider:

1. Will you need non-Microsoft integrations? Keboola offers a wider range of data sources and destinations outside of the Microsoft stack.
2. Is support important to you? When things go wrong, Keboola offers better-quality and cheaper support. ($0 is much less than Azure's $29/month during trial)
3. Are you price-sensitive? Whether you would like to get the most money for your dollar or would like to run a test drive without breaking the piggy bank, Keboola is the more affordable option. With its always free trial and transparent pricing, it is much more affordable than Azure Data Factory.

Feel free to give Keboola’s always free trial a go or reach out to us if you have any questions.