Archive | Data Integration RSS for this section

How can Cloud Data Quality Lead to MDM?

I recently attended a joint panel webinar between Informatica Cloud, StrikeIron, and Address Doctor (a subsidiary of Informatica Cloud) discussing Informatica Cloud’s latest service from its Winter ’12 offering, the Contact Validation Service.

Cloud data integration has really taken off in the past couple of years with the proliferation of SaaS applications, and the need for these applications to access and interact with data stored in on-premise systems. But data integration alone only solves part of the problem. In order to maximize your “Return on Data”, you need to ensure that you’re dealing with high quality and trustworthy data. As Ted Friedman, a Distinguished Analyst from Gartner aptly points out,

“Organizations cannot be successful in their data integration work unless they have a very strong focus on data quality built in. That’s because it’s not only about delivering stuff from here to there. You also have to make sure you’re delivering the right stuff.”

I also learned of the 1-10-100 rule – a Bloor Research whitepaper that states that it takes $1 to verify a record as it is entered, $10 to cleanse and de-dupe it and $100 if nothing is done, as the ramifications of the mistakes are felt over and over and over again.

Contact Validation ensures an abundance of benefits for all departments and industries. Marketing functions can benefit from higher campaign response rates by filtering out leads with incorrect contact information – in fact, a BtoB Online article claims that validating and managing data in the earliest stages of collection can lead to better lead scoring and lift conversion rates by about 25% between the customer inquiry stage and the point where marketing qualifies the leads.

In addition to marketing, warehouse managers can calculate correct shipping costs and legal departments can ensure regulatory compliance. Industries such as Insurance can use geocoding information to check whether properties are near a fault line, flood plain, or landslide zone, while banks can use contact validation in their deduping process.

With 30% of people changing their email every year, invalid emails can trigger bulk email flags, causing future emails to be filtered as spam, and result in the sender being blacklisted by ISPs and spam filters. This is why email validation is so important. The $16,000 per violation fine from the FTC also underscores the importance of phone validation against the Do Not Call registry.

I also learned that Contact Validation fits into the Data Quality spectrum, and is only a small part of it. But before even embarking upon a data quality project, it’s important to first of all integrate all of your data. Then, the next step is to cleanse, enrich, & augment contact data as a first step towards achieiving higher data quality. Once this is done, parsing and standardization, deduping, and finally monitoring and reporting round out the rest of the data quality activities. Moreover, these data quality activities must be done for all data types, and not just contact data.

With data quality processes properly implemented, a company can then finally take it to the next level and develop data governance policies, and hierarchy management for a full-fledged Master Data Management Solution. You can view the full replay of the webinar below.

Advertisements

Best Practices Around Cloud Data Integration

A couple of weeks ago, I witnessed a very interesting webinar panel discussion around people’s experience around data integration in the cloud. The panel was moderated by Jeff Kaplan of THINKstrategies, an analyst firm and the panel members included Andrew Bartels of PSA Insurance & Financial Services, Tom Carlock of Dun & Bradstreet, Aaron Mieswinkel of Astadia, and Darren Cunningham of Informatica.
I found this collection of panelists especially interesting because they represented constituents from several ends of the data integration spectrum – PSA was a consumer of cloud data integration services, Dun & Bradstreet was a provider of data quality services, Astadia was a System Integrator with deep experience in salesforce.com cloud integration, and Informatica was an ISV that specialized in out of the box data integration and application integration software.
Amongst the other interesting predictions I learned about on the webinar:
  • IDC predicts that 80% of new software offerings will be available as cloud  services and that by 2014, over one-third of software purchases will be via the cloud
  • More than 85% of Fortune 500 organizations will fail to effectively exploit big data for competitive advantage through 2015
  • 33% executives in a survey conducted by Saugatuck Technology identified integration as their top concern regarding SaaS deployment and use
  • On the SaaS vendor side, a THINKstrategies survey found that 88% of SaaS companies identify integration as important in winning new customers and a common sales hurdle.
These predictions seem quite telling about the main obstacle towards moving to the cloud – data integration. Many CIOs are still hesitant towards moving to the cloud because they believe that doing so would result in a massive strain on IT resources when it came to data integration, that the cost savings for moving to the cloud itself would be shot.
In the first wave of cloud computing adoption, businesses merely used companies such as Amazon Web Services for testing and development of their applications. CRM, and email then followed in the second and most recent wave of cloud adoption. However, when it comes to the third wave of cloud adoption, of moving mission critical systems such as ERP or proprietary financial applications to the cloud, there is still resistance to the idea because of the complex nature of these deployments.
This is where cloud data integration plays a huge part in taking away the complexity of integrating these various on-premise and cloud applications. When you have your systems on separate islands, it is extremely hard to benefit from the real-time BI that cloud computing can offer. Cloud data integration that can be done without the aid of IT personnel, while still following the procedures and rules around IT governance is the solution towards achieving real-time insight into your company’s operations. 
Cloud data integration also serves as a first step towards eventually achieving MDM nirvana – that state where your data workflows, procedures, and processes move seamlessly from system to system, with all personnel having insights into data that’s relevant to them. Cloud data integration can also serve as a catalyst
for cloud analytics, data-as-a-service (a.k.a. DaaS), and PaaS. 
Accelerating PaaS adoption is an area I’m especially interested in, given some work I’ve done in this space with past companies. A lot of application development in the cloud still takes place on two ends of the spectrum – on one end, there are developers who use a cloud application (such as salesforce.com) and need certain customizations built into it – they then use force.com to perform those customizations. In other instances, you have indie developers who use Windows Azure, or VMware’s Cloud Foundry to develop “nice to have” applications from scratch. But by and large, mission critical application development still occurs on-premise. 
I strongly believe that cloud data integration that can be performed by non-technical business users, will leave more time for IT personnel to develop mission critical applications in the cloud. With the PaaS space gaining more importance, and consolidating further, the time is ripe to take PaaS development to the next level. This in-turn will lead to cloud data integration morphing into “iPaaS” (or Integration Platform-as-a-Service). This interdependence between iPaas and PaaS will eventually lead to true SOA where we have open systems working hand-in-hand with each other, all in the cloud.
You can view the entire webinar below, and also check out Hollis Tibbetts blog on cloud data integration


%d bloggers like this: