Building Data-Driven Apps: 5 Best Practices

Five key tips for building the data-driven customer apps that will define the next wave of the customer experience.

By Allen Bonde, VP of Product Marketing and Innovation, Actuate

What's the best way to deliver data-driven apps? These are apps that give consumers what they want but that are also highly scalable and enterprise class. Based on my years in the industry, I think there are five central principles that really will help us get there.

There's a healthy appetite to get these data-driven apps out the door, and there's a huge amount of interest, verging on hype, in big data; Forrester Research recently estimated the potential size of what it calls "smart" computing -- that is, the big- or smart- or small-data market -- at more than $48 billion. It's thought that 90 percent of the Fortune 500 have some sort of big data projects either starting or established.

No wonder IDC sees 2014 as the year of data-driven sales, with marketing really taking off. Although we want data-driven apps, we may be less sure how to build them and, as I will show, how to scale them.

Here, then, are my five "Best Practices" tips for building data-driven applications.

Best Practice #1: Recognize how data impacts the customer journey

Brands and, increasingly, even governments want to work with the public in ways that offer a high level of personalization and responsiveness. Consumers -- the smart, connected, and demanding consumers of today -- will buy such services if they feel they can genuinely inform, connect, or motivate them (in the case of fitness apps, for example).

A fantastic service that brings all this together in one place is the Kayak travel search engine. This is very much a big data system that delivers what I'll soon be calling small data. By looking at the bigger trend lines of holiday action, it can tell me that the time to book is now, or to hold off because I'll likely get better savings later. The key term here is 'helpful,' in other words. Other great services you might want to look at include UK retailer Tesco's virtual shop in South Korea, which connects buyers with brands and offers in a compelling way. The upshot is that we need to offer data-driven apps that help people.

Best Practice #2: Focus on the "last mile" of big data

Ultimately, people don't care how much data you are working with on their behalf -- they just want to see results! We need to make apps that work for people at the sharp end; apps that work for real business users and shoppers and travelers -- not just for data scientists. To do that, never forget that you need to see things from the user's perspective -- that's the only way you're going to deliver all the potential of big data to the mass market.

Let's break it down. When we say data, what do we really mean? Do we mean real-time, instant data or do we mean historical trend data? That is an important question, as they have different business applications (and toolsets needed to deliver them). A key aspect of delivering that value is immediacy. How much latency is important to make this data useful to allow the app user to make a decision? How easy have you made it for them to grasp the insight the data has opened up -- how visual and easily accessible is it?

You might want to conceive of a kind of funnel workflow here, with big data going in at one end, being worked on by a collection and composition process and emerging as useful, targeted small data at the other -- manifesting as those alerts, flash-sale offers, and management dashboards your users actually want/need. These are all great ways for you to focus your development effort.

Best Practice #3: Think about how to build to scale in term of your sources, formats, and devices

Your applications need to securely manage, access, and deliver data from a wide variety of sources, in a variety of formats, and be accessible intuitively from a large number of different devices. In fact, in Actuate's 20-plus years of experience, we've found the bigger the application, the higher the associated complexity and, therefore, the bigger the need for a secure and scalable platform to access, manage, and deliver personalized content. Bottom line: You need to have the best stable, scalable platform at the basis of your data-driven push and the right tools to help you develop apps using it.

Best Practice #4: Follow the crowd

"Open" is better. You can get tremendous value out of opening up your app development process to "the crowd" -- in other words, the Open Source community. We've been doing that for 10 years at Actuate with the BIRT process. The numbers are impressive. We now have 3.5 million developers worldwide using BIRT and 13 million BIRT downloads so far. BIRT has also been of huge importance in developing the commercial products, including the BIRT iHub deployment platform.

You can grow your opportunity by taking advantage of not just open source but the power of iHub -- which is already the engine behind a range of interesting, early-day, data-driven apps, from the telco to the government market.

Best Practice #5: Start small but don't be afraid to think big

This best practice is the most pragmatic of the group. You need to get to market quickly but in a way that will deliver results for you and your customers. Experience has shown time and time again that great projects start by attacking the low-hanging fruit and then expanding up and out. You should be doing the same with your data-driven app projects. I would take this a bit further, though, with that concept I noted earlier: small data, which came out of some work I did with the Data Clarity Group.

What small data is about, ultimately, is giving focus and agility to your data work: Your ideas need to be as simple and smart (simultaneously) as you can make them. This is the way you're going to get the kind of compelling apps capable of delivering timely, meaningful insights that will provide the kind of accessible and actionable help users need to make their everyday consumer or business tasks easier or more efficent.

Finally, the "start small but think big" plan is the way to grasp the real aim here: you need to build applications at scale that literally millions of people can use but that are so personalized and responsive that each individual user always feels "this app was built just for me."

You can't do that unless you think about the customer at all times. That vision is what will make you able to give users the data-driven apps they want and you want to give them. Good luck!

Allen Bonde is vice president of product marketing and innovation at business analytics software specialist Actuate Corporation, the firm behind the BIRT iHub, Actuate Customer Communications Suite, and BIRT Analytics solutions. Follow the author on Twitter at @abonde or contact him at .

TDWI Membership

Get immediate access to training discounts, video library, BI Teams, Skills, Budget Report, and more

Individual, Student, & Team memberships available.