So you want to do a Big Data project?

2014

So, you want to do a Big Data project…

22014

…OK, we’re here to help!

But before you do anything else, ask yourself: why?

32014

A. Everybody else is doing Big Data!

B. You want to be a pioneer in your field and maybe get publicity through your project

C. You have defined a business goal that can be furthered through Big Data

this is not a good reason! And it

isn’t quite true…

this is OK, if you have the

resources for exploration

this is the best way to start

42014

When your goal is sorted, you can go for the data bit:

Remember to consider different types and sources of data.

Data is much more than numbers (it can, for example, be text, voice or video).

It’s also more than your company-internal data. The right source for your needs might be public (e.g. web content),

commercial (e.g. survey responses), or open (e.g. demographic data)

52014

Some ideas to get you started on data:

Your goal:Examples of useful data

sources:

Increase sales

Improve customer experience

Enhance production

Create new products

Social media /user generated content

Records of customer behaviour

Production logs, sensors

Market stats & forecasts

62014

Don’t be shy, talk to people about your project, and most importantly:

Get management buy-in.

You need executive advocates for your project in order for your results to be acknowledged, distributed and acted upon within the organization.

Also, data and analytics isn’t the only way of working towards your goal

so be sure to collaborate with other insight-generating projects for maximal results.

72014

Some things are better done sooner rather than later:

Consider privacy issues before you start.

If you’re working with consumer transactions or other sensitive data, don’t spook out your customers by your insights, and make sure your internal

data doesn’t leak to outsiders.

In some cases you might want to consider a ”mydata” approach – letting customers freely access and use their own data.

82014

And finally, for a bit of counter-intuitive advice:

Start small.

You don’t want your first project to get out of hand.

Remember, what matters is not the size of the data and the number-crunching power of the tools you use. What matters is reaching your business goal.

92014

1. Start with a clearly defined business goal2. Remember the many types and sources of

data 3. Obtain management buy-in4. Plan for privacy & security issues 5. Start small

Recap: Five pieces of advice for a successful first Big Data project

10

Questions? Comments? Need more advice?Contact Camilla Magnusson: camilla@instanssi.fi

Twitter: @InstanssiFiwww.instanssi.fi