IT Central Station is now PeerSpot: Here's why

How do you plan for an enterprise architecture review for business intelligence tools?

Nurit Sherman - PeerSpot reviewer
Content Operations Manager at PeerSpot (formerly IT Central Station)

Is it required in your company to conduct an enterprise architecture review before purchasing a business intelligence tool? What are the common materials you use in the review? Do you have any tips or advice for the community? Any pitfalls to watch out for?

PeerSpot user
88 Answers

Robert Neundlinger - PeerSpot reviewer

In our group it is necessary to conduct a vendor selection and look at multiple options. The categories suggested in the first post make sense for such a selection, same as the ability to link back to the business process systems mentioned above. Although the latter may depend more on the ability of your business process system to work with deep links and provide API endpoints. Either way, a catalogue of functional and non-functional requirements makes sense to be generated to evaluate any tool under consideration. This catalogue needs to cover specialized BI topics such as slice-and-dice, handling multiple currencies and languages (if required), hierarchies, versioning, and many others. I agree on getting outside support if you are new to the topic.

You may want to differentiate between standard reporting (i.e. the same set of reports and dashboards get used over and over again so can afford to standardize) and ad-hoc reporting (i.e. need to work with new data sources fast, may not need to re-use it so much). In some cases one tool is good at one thing, but not the other. This is also related to the centralized versus decentralized model mentioned above.

Obviously the decision between Cloud and on-premise is an important one to take as each way has their pros & cons.

Joe Fernandes - PeerSpot reviewer
Real User

In my organization, we broadly have three major vendors. Viz. SAP, Microsoft, and IBM. At the time we purchased SAP, the business function which was going in for SAP ERP adopted SAP's tool for Business Intelligence. Similarly, the users who were using Microsoft software preferred to go with Microsoft BI. IBM too demonstrated their BI offerings but since we already had SAP tools and Microsoft BI we did not procure Cognos.
So, we did not do any Enterprise Architecture review for the BI Tool. But, for an organization embarking on implementing BI, it would be advisable to devise your BI Strategy in consultation with business users then see which product fits in with that strategy.  Ideally, this should be driven from the top to avoid various groups of users procuring too many tools which may then be difficult to integrate.
While selecting a BI Tool,  factors which can be used to assess a good fit would be -1. Cost 2. Features the product provides 3. Ease of use 4. Resource availability with required skills for development and maintenance 5. Support vendor provides6. Standing of the vendor in the BI market7. Scalability8. Cloud deployments 9. The performance with workloads contemplated10. Security features 11. Integration possible with existing software 
If one is new to BI, it is advisable selecting an implementation partner who brings in the requisite knowledge, skills, experience, and expertise.

James-Anderson - PeerSpot reviewer
Real User

Obviously you want to make sure that the tool can handle different types of sources like SQL server, Oracle, REST, Web Files, GeoAnalytics, Google, etc. Make sure you understand who is going to be creating the reports. Do you want a traditional IT approach or do you want the tech savvy business users creating the reports? Centralized vs Decentralized. Also consider any security and privacy needs like HIPAA. One consideration I always look at is if you can integrate back into the systems you use. Can I look at a report and then link back into my core system using hyperlinks or code? Good luck!

Hit Mistry - PeerSpot reviewer
LeaderboardReal User

There are a plethora of BI tools and solutions out there and architecture is absolutely the second thing that should be on the list. Understanding the business needs and long term goals is essential to building the foundation (architecture). It's important to determine whether the reports need to be interactive with drill down capabilities or can a static report suffice. Cost will be heavily impacted as a result. Distribution model needs to be considered as well. Will reports be sent via email, produced in batch, generated on demand, etc. Last but not least, a good security model not only for the reports but ensuring the data presented on the reports is properly secured.

Artour Aslanian - PeerSpot reviewer
Real User

Hi guys,
I will be really short..

1 Means a company budget and a number of a user wise licenses required to support BI enterprise
2 Means a platform change in case it is matching above and needs to be a "multi" or bringing all to a same one.. :)
3 DW/Front End systems possible migration/integration and a Reports migration/rework
4 A DW History migration process - can be costly from a legacy point.. :)
5 Timelines required to migrate estimation -> based on a resources available/required
6 A user wise company level of a culture.. maybe it is just better to use Excel for everything?
Maybe, if everybody knows what this is and can use Excel as a guru.. :) Pivoting is a most difficult thing still to understand for almost all of young guys, though all had a courses.. :)

Finally, I would recommend a globalizing all existing systems in a company bringing to a single platform provider like MS. PowerBi is a piggy and buggy but it is really invested and shows a result. Please, avoid products that are not invested and thus getting worse.. though advertisement stays great.. like a dynamic cubes from Cognos.. No words, sorry..

Having an understanding of a product development investments trend by a provider and a user wise/licensing expenses related -> may help to find a real balance between a products of an interest on a market that you will finally bring into the existing at your particular company infrastructure..

P.S. Any product support fees and a quality of a support can be crucial in terms of a planned fast migration/upgrade.. :)

Robert Nendza - PeerSpot reviewer
Real User

I agree with the recommendations from the first two answers: make sure you know your stakeholders' pains and requirements, make sure all the requirements are explicitly addressed by the potential vendors, and do an architecture review to uncover possible points of incompatibility or performance challenges. And I'll add one more that has worked well for both our company and our customers. Consider requesting a proof-of-concept (POC). Most BI tools should be able to connect to your real data sources and produce real output without a lot of custom work. Your POC should include smoothly connecting to your data, formatting the output to satisfy your stakeholders, performance under real-world conditions and usability by real users. Vendors may offer this at no charge, or for a small fee. Either way, it will be time/money well spent.

Jessica Schwab - PeerSpot reviewer
Real User

Hi - That is best practice is to have a architecture review however, everything is online to purchase if you would like to buy without this step. Common Material we use to review would be Diagrams based off the customers environment, best practices of customers we have successfully deployed with. Tips--> I would make sure to have a sound understanding of your business, people and process prior to engaging in BI tool. Really understand your organization goals for now and future. Unfortunately many do not encompass future and only right now and then end up spending a lot of $$ later to re-implement/architect because their short term goals were price motivated. Ends up hurting you later if you do not plan for future.

Amy Armstrong - PeerSpot reviewer

I am beginning a BI review right now. I have the same questions so will be anxious to see how this is answered.

Buyer's Guide
BI (Business Intelligence) Tools
August 2022
Find out what your peers are saying about Microsoft, Tableau, SAP and others in BI (Business Intelligence) Tools. Updated: August 2022.
621,703 professionals have used our research since 2012.