The Focal Point of Data Integration: Did Salesforce Get It Right?

We've all been brought up believing that the focus of data integration is a back office function and that this process should be something that is completed by large ERP and CRM applications deep in the bowels of the IT department. All ERP vendors promote their own add-ons that can facilitate the data integration process.

The problem with data integration is not the transfer of data, which has to go through a migration process and follows a rigorous MDM (master data management) exercise to determine a unified enterprise-wide data schema. No, the problem lies in determining the final functionality and higher level business goals that need to be reached after the integration has been completed.

Traditional integration projects worry about ensuring that the back office systems are integrated and are based on sometimes justifiable needs to share data between systems. But the focal point of determining what data needs to be integrated from which back office systems has to happen at the customer level, in the front office to ensure that the business process flows correctly.

This is where Salesforce.com has it right. A company should determine what their sales people, customers or partners need on a smart phone or tablet to ensure that the business process flows faster and more efficiently. Based on that need, an integration project needs to be completed. The focal point of all data integration needs to be in the front office as SFDC proposes, not in the back office where we usually find it.

Last Night: I was invited to Advanced Screening of The Hobbit IMAX in 3D!

This was an amazing play on the 3D effects that only come to life when viewed in full IMAX technology. I have never seen anything like it and it gets a full 2 thumbs up!  It was simply amazing. This is as much a roller coaster ride as I have ever experience even in real life. The 3D makes this movie amazing, so that the experience is more than you physically fathom and I walked out drained of energy.

Go and see this, but don't waste your time with 2D or anything less than IMAX!

2013 Dreamforce: Lots to learn from results

I was unable to attend Dreamforce, which is the major event put on by Salesforce.com, but there were a lot of comments that could boil down some of the lessons for many of us who weren't there.

On this blog, I have spoken much about the Cloud transition and how the datacenter of the future will be in the cloud. Many companies offer real cloud solutions for companies ready to go, the leader in this industry is AWS. But for the mainstream customer with a data center the transition to the cloud is a major disruption and until things break, they are unwilling to move.

The other offering is the traditional SaaS model that companies can transition to, where a major business process transition that involves the use of a new SW package in the cloud can make the same transition. Here, there is often no choice. If the functionality you are searching for is in the cloud, then you will move to the cloud. 

In the case of Salesforce.com, many companies are finding the best of both worlds. Yes, following an improved sales strategy with a CRM model will lead you to their core CRM package and therefore to the cloud, since SFDC (=Salesforce.com) does not offer any other model. But what they have been able to create is a cloud based platform that can take the core of your customer data and expand it to include all other data your company needs or will need from here. SFDC can run your company's cloud transition completely from soup to nuts. Yes, AWS offers the cheapest computing and storage option and many different SW packages are becoming available in a SaaS model. But no where will you find the complete data landscape package that will allow you a transition on your time, piecemeal or complete, as you can with SFDC.

Here are some choice links that I learned a lot from:

Many new ideas
Idea is the customer
Take aways from Marissa Meyers

PaaS the Fastest Growing Cloud Segment

This is something I have written about before, but it seems that the market has finally caught on: PaaS will be the dominant segment throughout 2014. For more information on exactly what the market is and to understand what PaaS means, read this.

Why PaaS? Because the easiest business case for cloud adoption is the area of Software development and operations, called DevOps. The business case is simple and ask anyone who has ever run a data center why. Whenever a new version or release of a piece of code was ready for functional/regression/production test, there needed to be a new environment created for that test. The datacenter guys had to find servers, implement servers, install servers and create a complete environment to complete the test. Once the test was done, the hardware needed to stay untouched, so that a new test (yes, they did find bugs) could be implemented quicker.

The problem here is not only unused hardware and wasted hardware set up skills, but the worst was the time it takes to get a new release out the door.

Cloud and especially PaaS promises to change that forever. Users (SW development guys) can set up an environment on their own as a cloud account, test, and then only use the hardware environment when they need it. As soon as they don't need it, the resources no longer cost money. DevOps is the easiest to understand use of Cloud computing.

Is Risk Really Holding Back Big Data Projects?

In his WSJ article titled "The Risks of Big Data for Companies", Dr. Jordan explains a number of risks that companies should be aware of before jumping into this new and unchartered market.

He outlines that too much data could be dangerous for companies who might invoke inner company politics or make unjustified decisions too early with the results they make. I would imagine there are no risks in ignoring information that the data gives us and just keep moving along as a horse with blinders.

WHAT? If people seriously took the advice of this column in the WSJ, we would never advance in business. More importantly the #1 risk that is keeping up every CEO is not whether a plant in Canada has better output than the same in the US, thusly creating an "inner company rivalry". No, the #1 threat is the competition. The competition can close down your market and make your company's revenue tumble.

Listening to this sort of advice is not only dangerous, but plainly stupid. No company that I have helped with Big Data projects has decided to throw out human intuition and experience for decisions made based on the data alone. Instead all functioning companies want to see whether the data can complement decision making. That is what a pilot will show and the ensuing business case will prove it, or not. Pilot programs should be done with Big Data in small areas to show what we could know, without fear that we will uncover some terrible new insights that could cause inner company conflict.

If you are starting a new Big Data project, please do NOT read this article. Instead follow your own intuition and call me in as an adviser.

Big Data Projects: Who is the Organization's Big Data Leader

Deloitte Touche Tohmatsu Ltd.

The big question for any consulting organization trying sell Big Data projects is to find the Big Data leader in that organization. There is always one person in every organization who has spurred the company you are selling into to use the massive amounts of data that they have access to to create solid, decision driving data.

Who is that person? Is she the CIO? or is he the CMO? or could she be a Director-level person with a special charter? Tom Davenport, the Data Analytics Guru from Harvard discusses this well within his Wall Street Journal blog. You can read it here. Tom's opinion is pushing for a decision within a corporate organization for where the Big Data Leader should be. I agree with his overall complaint from the Deloitte poll shown below that the leader should be someone who works enterprise-wide, so that an overall Big Data strategy can evolve. In a perfect world that would be great.

But I think reality is different in the companies I have insight to when it comes to their Big Data adoption. In reality, a Big Data project comes out of a need to know more about your customers than your competitors know. It has to start with a dream somewhere in the organization that says, if we had this piece of information, we could sell more, grow market share or save the cost of creating products no one wants. This need could develop in an IT organization, but only if their goals are clearly aligned with the business goals, which sadly is not often the case. Most of the times, I have seen this need arise was in the business unit level, where goals are clearly tied to the success of selling products.

If 64% plan a Big Data project, why are so few IT consulting companies prepared?

-->

According to a Gartner survey, 64% of the companies surveyed plan to invest in some sort of Big Data project in 2013. Many of those surveyed will run some sort of pilot project, not investing in hardware as much as investing time with a singular goal: where will Big Data payoff? how much needs to be invested to get which results?

Despite this strong opportunity, few IT consulting organizations have strengthened their ranks with specialist who can guide their clients through the uncertainty of this new market. One indicator is the lack of career positions posted for Big Data sales specialists and consultants who can create a solid business assessment that is strong enough to justify further funding. There are some who are not only specialized in this area, but also host cloud environments, where their clients can start a real pilot without any data center investments. It seems that these few will do well in the upcoming boom as others scramble for Big Data market share.