The Data Warehouse Automation Tools Survey Results
As you might know, I started a survey on the usage and opinion about Data Warehouse Automation (DWA) Tools.
Today I am presenting the results.
I sincerely want to thank all the participants for sharing your thoughts.
First some insight into the persons that filled in the survey. You were not with so many (62 completed surveys).
In statistics there is a formula to compute the minimum sample (number of participants) given a certain margin of error, reliability level and total population. If I fill in 10% for error margin, 90% for reliability level and 20000 for total population (this number does not have much impact on the end result) the minimum sample (number of participants) should be 68.
Please note that with 10% error margin and 90% reliability level that there are considerable chances that the answers are not representative for the entire population of data warehouse professionals.
Participants by country
Participants are spread over 23 countries on several continents, with a concentration in the Netherlands and the United States.
|Country||No of participants|
Jobroles of participants
Almost half of the participants selected Data Warehouse Architect as best describing their current job role.
Furthermore there were 12 ETL developers and a number of other job roles, as shown in the graph.
|Jobrole||No of participants|
|BI & Analytics/Data Science manager||2|
|Data warehouse architect||30|
|Front end BI Developer||1|
|Multi-disciplined BI Professional||4|
Familiarity with the investigated tools
First question to answer is: how familiar are the respondents with those tools?
I used 5 levels of familiarity with a short name and explanation. The option values have the following meaning:
- Don’t know: I do not know this product.
- Heard of: I have heard of this product but I have not worked in a project where it was tried or used.
- Tried: I have tried this product but not used it in production environments.
- Used: I have worked in a project where this product was used.
- Using: I am currently working in a project where this product is being used.
The results in a graph:
And the exact figures, first absolute, then percentages:
|Issuer – Productname||Don’t know||Heard of||Tried||Used||Using|
|AnalytiX DS – CatfX / Data Vault Code Gen Bundle||44||14||3||0||0|
|Attunity – Attunity Compose (formerly BiReady)||36||21||4||0||1|
|Gamma Systems – Data Warehouse Studio||55||5||1||0||0|
|Javlin – CloverETL||51||7||2||1||0|
|Magnitude – Magnitude Kalido||39||20||1||1||0|
|Qosqo – Quipu||43||12||3||2||1|
|TimeXtender – TimeXtender||42||14||3||0||2|
|Trivadis – biGenius||52||5||1||1||2|
|Varigence – BimlExpress / Bidshelper||43||3||4||3||8|
|Varigence – Mist / BimlStudio||46||4||9||1||1|
|WhereScape – RED||35||10||6||5||5|
|Issuer – Productname||Don’t know||Heard of||Tried||Used||Using|
|AnalytiX DS – CatfX / Data Vault Code Gen Bundle||72,1%||23%||4,9%||0%||0%|
|Attunity – Attunity Compose (formerly BiReady)||58,1%||33,9%||6,5%||0%||1,6%|
|Gamma Systems – Data Warehouse Studio||90,2%||8,2%||1,6%||0%||0%|
|Javlin – CloverETL||83,6%||11,5%||3,3%||1,6%||0%|
|Magnitude – Magnitude Kalido||63,9%||32,8%||1,6%||1,6%||0%|
|Qosqo – Quipu||70,5%||19,7%||4,9%||3,3%||1,6%|
|TimeXtender – TimeXtender||68,9%||23%||4,9%||0%||3,3%|
|Trivadis – biGenius||85,2%||8,2%||1,6%||1,6%||3,3%|
|Varigence – BimlExpress / Bidshelper||70,5%||4,9%||6,6%||4,9%||13,1%|
|Varigence – Mist / BimlStudio||75,4%||6,6%||14,8%||1,6%||1,6%|
|WhereScape – RED||57,4%||16,4%||9,8%||8,2%||8,2%|
I think most remarkable is that most of the tools are not very well known: the answer “Don’t know” is on the average given by 72,3% of the participants.
The product satisfaction question gave the insight that there are no ‘bad’ tools. All tools fulfill a need, and the digits go from 3 to 5 (on a scale of 1 to 5). So the products with the lowest mark for satisfaction, still score a ‘C’ (or as we say in the Netherlands, a ‘vijfje’).
Value for money
Value for money is obviously a very subjective matter. What one person thinks is expensive, the other might think of as cheap. Most vendors are not transparent about their prices.
Marks vary from 1,80 to 3,46. What we can conclude from this is that some tools have the perception of a relatively low value for money, while others provide a better value for money, according to the participants of the survey.
The products that have an average score of 3 or higher (again on a scale of 1 to 5) are both Varigence products (BimlExpress/Bidshelper and Mist/BimlStudio), AnalytiX DS (CatfX / Data Vault Code Gen Bundle) and Trivadis (biGenius).
By the way, the question in the survey had the following scale: 1=cheap and 5=expensive. I translated this to a value for money by reversing the answers (so 1 became 5 and vice versa and 2 became 4 and vice versa). In this way, a higher score means a ‘cheaper’ pricing perception, thus a better value for money.
Other tools that were mentioned
A number of other tools were also mentioned, many of them only once or twice.
Remarkable is that also tools that do not meet the “DWA Tool” definition were mentioned, for instance Microsoft BI Products.
What I learned from this is that I should have given a definition of what I meant with a “Data Warehouse Automation Tool”.
“Data warehouse automation or DWA refers to the process of accelerating and automating the data warehouse development cycles, while assuring quality and consistency. DWA is believed to provide automation of the entire lifecycle of a data warehouse, from source system analysis to testing to documentation. It helps improve productivity, reduce cost, and improve overall quality.”
Worth mentioning is the open source tool ODE. I didn’t know it but tried it recently. It’s a sql based, metadata driven tool without graphical user interface, and I could not get an “AdventureWorks” example working due to errors that occurred during working through the example. So I think this tool needs to become more stable or at least should have a better manual on how to use it. But it’s Open Source and that puts a smile on my face.
From a company/human resources perspective, working with such a tool can however lead to a hidden cost because using it can be more laborious than working with a commercial, ‘more stable’ product. And once you start changing the source code (for yourself or your company, not as member of the Open Source Development team) you can’t upgrade to newer versions anymore and are stuck with the maintenance of your copy of the tool. And if you join the Open Source Team you are now doing two projects instead of one!
In the Netherlands we have an old-fashioned saying “Bezint eer gij begint” which means “think well before you start”, and I think this is applicable here as well. A license for a commercial tool might be worth the money.
What also appears to be the case, that a number of those tools are not for sale, but are tools that consultancy firms use for doing projects.
List of all the other tools that were mentioned:
|Issuer – Productname||Mentions|
|Adante – Datawarehouse on Demand||1|
|Automic – One Automation||2|
|Birst – Birst||1|
|Centennium – TBD engine (template based development)||1|
|Cubido – Cubikit||1|
|DIKW – DHW Accelerator||1|
|Grip Op Data B.V. – GDIP||1|
|Infent – Infenture Insight||1|
|Informatica – Powercenter||2|
|Informatica – B2b data exchange||1|
|Informatica – Informatica||1|
|Informatica – MFT||1|
|I-refact – I-refactory||1|
|Kadenza – Smart Datawarehouse generator||1|
|Microsoft – MSBI(SSIS,SSAS,SSRS)||9|
|Microsoft – Power BI||1|
|Optimalbi Limited – ODE (Open source)||1|
|Oracle – ODI||1|
|Oracle – OWB||1|
|Own build – Excelbased tool||1|
|Pentaho – Kettle||2|
|Pentaho – Big Data||1|
|Rabobank – Foundation Toolkit||1|
|SAP – Data Services||2|
|SAP – Information Steward||1|
|Tableau – Server||1|
|Talend – Data Fabric||1|
|Talend – Open Studio||1|
|TopBi – Data Vault Generator||1|
|TopBI – HSG||1|
|Varigence – BimlFlex||1|
|Varigence – BimlOnline||1|
|Viz Analytics – Data Warehouse Studio||1|
Comments given about all the tools
Participants could give their opinion about the tools in free text format – and they did!
Here are just the bare comments.
Attunity – Attunity Compose (formerly BiReady)
- “Like features, makes work flow easily “
Qosqo – Quipu
- “This one could be split up between the free and paid variant for this research.”
- “Most advanced options on data vault modeling and generation in one tool. Does not support data marts, requires Groovy knowledge to change templates.”
TimeXtender – TimeXtender
- “vendor lock-in, no best practices applied when generating SSAS cubes”
- “lock in, very good features, easy to use (even for business users)”
Trivadis – biGenius
- “if they extend SSIS generation no lock in, very flexibel, fast, cool”
- “No Vendor-Lock (generated Code can be scheduled and executed native), Microsoft and Oracle-Support, Data Vault Support”
- “Integrated product, short learning curve, easy to adapt”
Varigence – BimlExpress / Bidshelper
- “Great to building SSIS packages.”
- “Free, but you need to write the generating code first. Ideal for consultants though.”
- “Nice combination, but needs to mature.”
- “no lock in, very cool that you generate SSIS/SSAS, free, a lot of work coding, not for everyone”
- “Very flexible, but long learning curve”
- “Second best tool to generate SSIS packages. Requires a lot of exceptions if you want to alter parts of code for specific tables. Debugging is hell.”
Varigence – Mist / BimlStudio
- “Has potential but is progressing slowly. Pricing is fair.”
- “No vendor lock-in, as the result of BIML are just SSIS packages”
- “More mature version.”
- “easier, expensive, nice you can easily re-engineer from existing DWH”
- “There is no reverse engineering module.We can not convert old package to use BimlStudio”
- “Very flexible”
- “Best tool to generate SSIS packages. Still requires a lot of manual labor for low level changes.”
WhereScape – RED
- “We were able to increase development speed with a factor 3. “
- “Vendor-Lock (needs App-Server for Scheduling, Processing), Microsoft not fully supported (only Beta), Data Vault Support”
- “Creates stored procedures which clutters db’s, powerful tool, but has some annoyances”
- “Several products which are not completely integrated, some features not “ready”, several limitations (especially when working with datasets, expensive for what you get, only based overseas (except smalll base in UK)”
- “You have to click and type almost everything, then it “generates” a stored procedure. Also stores its repository in the same database and schema as your data warehouse.”
- “Changed projects into initiatives enabling business analytics. “
- “Like lineage, like support for multiple generation types (such as data vault), dislike performance, dislike excessive storage, dislike cost of consulting, dislike unreliability, dislike UI, dislike lack of adapters, dislike reliance on database procedures, dislike scheduler, dislike target specific repository”
Comments for other tools (not all available for sale)
Adante – Datawarehouse on Demand
- “Very easy to use automation tool for Dimensional modeling”
Centennium – TBD engine (template based development)
- “SQL server based metadata based staging, vault en datamart generator. Developed by Centennium, for the larger part made adult by me.”
I-refact – I-refactory
- “A very solid, well thought solution for the data management. It’s not just for dwh, but for the whole data platform of your company.”
Kadenza – Smart Datawarehouse generator
- “Model driven generator c-sharp”
Talend – Data Fabric
- “Talend is a very friendly for completing complex operations like Data Integration, Data Quality and Big Data (towards Hadoop) with minimum programming skills”
Viz Analytics – Data Warehouse Studio
- “Custom built from the ground up, with own ETL engine, very quick to build and implement DWs.”
One of the questions was about the technology that is being used (from Gartner Magic Quadrant for BI and Analytics). As more than one answer was allowed, the total number of responses is more than the number of participants, because some participants use two or more technologies.
In the Top 5 we see: Microsoft, Oracle, Tableau, SAP and Qlik. To be honest, nothing that surprises me.
Conclusion / Wrap up
With this survey I tried to give the business intelligence community more insight into the available Data Warehouse Automation Tools.
Due to the low number of participants, conclusions come with large uncertainty margins. I think this is a pity.
Nevertheless this investigation could be good starting point for further analysis on DWA Tools or a DWA Tool selection.
(c) 2017 hansmichiels.com – Do not steal the contents – spread the link instead – thank you.