I have a whole new appreciation of Business Intelligence now

As many of you know, I grew up as a BI and Data warehousing guy – I have implemented BI for a lot of users across the world . I have collected requirements from shipping clerks as well as CXOs , and in at least a dozen different languages. I have spent countless hours thinking and rethinking data models, how best to transform data and how best to present data to my users.

Along the way, I became a manager and then an executive, and thus became an active consumer of BI myself. But in the big companies like IBM and SAP that I worked at – I learned to live with someone else’s BI design. As far as I can remember, I never had to create significant new requirements . These were rather stable businesses that could be run with minor tweaks to existing BI capabilities. 

And then few months ago, I joined MongoDB and my whole perspective on BI changed. We are for the most part still a startup. We don’t have a huge IT arm that can cater to endless requirements from me and other leaders of the business. Our IT landscape is almost completely SaaS based. If we can hire one more person – we would rather hire to fill a front line technical role to make the product and customer experience better,  or a sales or channel type role to make the business better. For foreseeable future, I don’t expect that to change either.  

We are a global business – and we are growing incredibly fast. And to keep that pace – we need good data, especially when it comes to customer facing business whether it is direct sales or channels (which I run) . Having grown up as a programmer and then a BI guy after that, I have a great affinity for making decisions by numbers. With the speed at which we grow and our lean policies, I don’t really have a lot of time to wait for information to see how things are going – which essentially means I need good quality operational BI at all times. 

We use Salesforce.com for our CRM. I am a first time user of this solution – which might surprise a lot of people. My past experience with CRM has all been in Seibel and SAP CRM. The best part of salesforce.com was that I did not need any training to use it – none at all. My past experience was almost immediately transferable to use the system as a non-expert. My primary use is not as a transactional user who creates or updates opportunities etc. My main goal is that of gaining quick visibility into the aggregate opportunity to order process for channel business, with the ability to drill down into details as needed.

Once I got settled in my new role, and got to know my team better – my immediate priority was to get a full view of the global business. I mocked up initial requirements into spreadsheets and discussed it on phone with my partner manager Guillaume in Dublin, who is an experienced salesforce user. From my past consultant life – I estimated the effort required as a few months in the technologies I grew up with (assuming I got the most skilled people I could find). Next day morning, I saw Guillaume already had 3 dashboards ready for me which showed most of the information I needed. And then in 2 more daily scrums – I had the 6 dashboards I needed to view the business from every dimension I care about. That is much less than the time it would have taken me to write a proposal for a customer for this work in my past life. 

What did I learn from this experience ?

A lot of good things for sure

1. Business users like Guillaume (and Luca, his boss who runs channels in EMEA for MongoDB) are better BI consultants than anyone I could have ever hired from outside. He not only knew the technology well, he knew my business well and could challenge my assumptions and give me new ideas. It has convinced me that rest of my team including me should step up our skills in salesforce.

2. The technology to build operational reports should be extremely simple so that business teams can iterate quickly. Till I saw it with my own eyes, I did not believe that it could be this easy.

3. From prior life as a programmer and a BI guy, I am well aware of the limitations in reporting – so I can minimize the churn in requirements gathering and make good compromises on what needs to be measured.

4. The simplicity of reports and the report writing technology – and my big time aversion to any transformations (having seen how data loses meaning way too many times) – helps us stay nimble and make changes on the fly. 

There are also some areas of improvement of technology , which I am sure Alex Dayon and team will fix at some point, hopefully soon .

1. Charting and visualization is very limited – so when multiple graphs are put next to each other it is quite a strain to discern information quickly. Granted, the ease of changing things on the fly is more important to me than flashy reports.

2. Only 20 controls possible in a dashboard. I can compromise on it for now, but as business grows – this is a pretty serious limitation for me to get a global view across everything I need to monitor and act on. 

3. Reporting across objects looks limited – but this could just be my lack of experience.

4. Operational reporting does not replace the need for a data warehouse . I still need some other place to combine the lead to order process with information from from Finance, HR etc. For my current purposes, I have work arounds – but if all the SaaS vendors for CRM, Finance, HR co-operated and built a BI solution to seamlessly provide me with integrated data – I will swipe my credit card happily to buy it. 


Published by Vijay Vijayasankar

Son/Husband/Dad/Dog Lover/Engineer. Follow me on twitter @vijayasankarv. These blogs are all my personal views - and not in way related to my employer or past employers

11 thoughts on “I have a whole new appreciation of Business Intelligence now

  1. Hi Vijay,

    Interestingly, I found that as an SAP employee the biggest challenge in getting data out of any internal system was navigating all the internal approvals and business case requests from IT before we could get anything concrete done.

    Having said that, it has been my experience that most reasonable and recent software solutions provide some form of data inquiry and dashboarding tool that the average business user can jump on to pretty quickly and get a half way decent result regardless of the tool set…but of course once you get a good tool you can really shine.

    Certainly that has been my experience here at Acumatica, with our Generic Inquiry function and the ability to turn these inquiries in to meaningful pivots and dashboards.

    Once you have a basic understanding of the data model things get pretty easy from there…IMHO

    Richard Duffy


  2. Vijay – during your tenure at SAP did you had some experience with the Cloud for Sales solution? I guess that would make a more applicable comparison than Seibel/SAP CRM. Not sure about the reporting capability and object flexibility there though


    1. Was that the one renamed from Sales on Demand ? I never had to use it as a consumer. For me – Seibel and SAP CRM were the only ones I used personally .

      Also – for operational BI, why should the deployment model of CRM matter for UX ? Not sure I understood your point correctly


      1. Yes – I appreciate that as a techie . As a CRM user , this is an implementation detail that is not a big factor for me .

        As a techie – I can poke a lot of holes in Salesforce too ( to be fair – also in sales on demand ) . But as a business user – I realistically will favor ease of use over tech elegance . Sales on Demand from a UX perspective was far superior to comparable functionality in SAP CRM . But if we look at the market – very few even know anything other than SFDC , sap , dynamics and seibel .


  3. Great post Vijay! Like yourself, I’ve been a long time in the BI business, I do however believe that ERP & CRM vendors should provide just enough operational reporting capabilities to serve their users while staying focused on their core business. ‘just enough’ being the tricky part for their PMs..

    Beyond this, agile analytic tools are necessary, specially to address the variety of data sources that you’ll want to use. Off course I have to +1 @dahowlett on his Alteryx + Tableau recommendation.



    1. Unless of course the customer already uses 3 or 4 BI tools–common, in which case why should customer pay for commoditization over again and again? Why not bolt on the value added differentiation? What’s strange to me is that so few BI vendors will partner. I think the reason is simply perceived low risk of migration, supported by stats. Trouble is making sense of it all can cost for more than entire IT budget many times over. .02


  4. As one of the named future of BI a couple of years ago by leading analyst firm, suppose I felt compelled to keep updated and to some extent observe a bit of the change we helped instigate. After many years of little innovation, quite a bit taking place now. Financials are commoditized now–dozens offer decent reports, many with good drill down capabilities. Of course I’m doing my best to remain ahead of the pack–it’s morphing into distributed network operating systems, which Kyield began moving towards years ago, with several pre-packaged apps involving machine learning and deep learning, novel security, new algorithms–a great deal of new capabilities frankly. I’m kind of surprised myself at times what we can do now. Only a very few are keeping current.

    One of the better places to review #BBBT http://www.bbbt.us – can join for free and listen to podcast and/or view video presentations. Sort of a network of independent consultants, a fair portion of related vendors have presented. I’ve discussed with Claudia – we’ll likely present Kyield once a couple of the current large unification efforts in pipeline are mature & fully functional system is demonstrable via full cloud version, perhaps before. I’m more shy than most on IC/IP/TS, but then we have more to protect than most.


  5. Salesforce.com reporting is very limited and has been that way for a number of years. I thought they were going to sort it out about 4 years ago but the project disappeared when they dropped working with in-mem as an option. Methinks you’ll be doing a lot of finagling in the short term to get much that will add value for your decision making beyond what you have today.

    If you’ve not already done so then I’d deffo consider Alteryx + Tableau (or Qlik) for your inevitable ETL when spanning systems + data viz.

    Oh – and yes – I’ll have a lot more to say once I return from DATA, Tableau’s customer event. I’m lined up for some training with a live data set I want to parse. Stay tuned 😉

    My 2cents


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: