Posts in Strategy

chief data officer

6 Tips for A New Chief Data Officer

September 29th, 2017 Posted by Executive, Strategy 0 thoughts on “6 Tips for A New Chief Data Officer”

A guy walks into a bar. The bartender asks, “What’ll it be today?” The guy answers: “I don’t know.”

Pretty boring joke, right?

The same letdown occurs when a new Chief Data Officer has no clear strategy for the future.

The title of Chief Data Officer has only emerged in the past few years and most CDOs today are the very first to be filling that role at their company. And because they have no shoes to fill, they tend to walk in barefoot. That is, they come in without any overarching strategy for how the company’s data should be stored, shared and exploited. BIG MISTAKE, because the earlier a company develops its data strategy, the more valuable its data will become in the long-term.

Here are 6 tips for an incoming CDO:

1. Think outside the box. Every day, companies are finding new ways to slice and dice data in a valuable way. For instance, internal data like inventory are now helping companies become more efficient. The CDO should make sure to survey all the departments to find out what kind of data is out there, just waiting to be activated.

2. Develop a working plan to define the ways in which data can unlock business outcomes. Don’t set a 1 year plan, let alone a 5- or 10-year plan as this will only get changed multiple times within the first year.

3. Determine who is in charge of what data. That could be several teams – for instance, IT could be in charge of securing it, and Data Analytics in charge of building apps over the data lake.

4. Ensure the company has a logical map for the data. The data lake is an incredibly rich resource, and why it needs to be used smartly. A clear data architecture will save time and money and facilitate the creation of apps.

5. Define a strategy for apps. This should be a central concern for the CDO. After all, the data on its own is like silver in the mine. It is useless until it gets activated. From the outset, the CDO should have a strategy on the types of apps the company should develop versus buy.

6. Start small, scale fast. Find a dataset that is easy enough to pull insights from, such as clickstream (or event) data. The company can leverage this kind of data very quickly, especially with predictive behavioral analytics.

 

If any of this resonates with you, tweet at me or email me to share your thoughts and experience with, or as, a CDO.

 

Data That Stays Together, Works Together

July 31st, 2017 Posted by Analytics, Behavioral Data, Strategy 0 thoughts on “Data That Stays Together, Works Together”

When it comes to data analytics, marketers are missing the forest for the trees. If you think your company’s most marketable data source lies in your enterprise data, think again. Your company is sitting on a gold mine of customer data, siloed in different departments, just waiting to be integrated and activated.

According to eConsultancy’s 2017 Digital Intelligence Briefing on Digital Trends, 59% of marketers who have an intermediate or advanced understanding of the customer journey stated that they had trouble unifying different data sources.

On the front end, you may have your clickstream data, which can include activity from ad displays, social media and email campaigns. Some companies even have data on the voices of their customers and that’s a real trove for piecing together customer demographic profiles. Companies also have loads of data on the back end, waiting to be mined, and this includes margin data, CRM product data, and enterprise resource planning, among others. Combined, both front end and back end data can turbocharge your data analytics system.

But in order for this to happen, the data needs to be removed from its silo and made accessible in a central behavioral data repository. Everything under one roof and one program to rule them all.

A predictive behavioral analytics platform can take almost any type of data sitting in your data lake and turn it into gold. It does not require you to manually identify every single data point across different departments because machine learning algorithms do the work for you. An individual data point teaches the model something completely new, regardless if it is tied to other data points in the dataset or not. As the number of data points coming in to the central behavioral data repository grows, the algorithm’s predictions on user behaviors become more and more accurate. Therefore, companies which have started activating their data are gaining the edge needed to secure their spot as a market leader for tomorrow. The sooner you “compound”, the greater the benefit.

Are you already doing something similar? Tweet at me or email me to share your experiences.

 

Are Homemade Predictive Behavioral Analytics Applications Better?

July 17th, 2017 Posted by Analytics, Strategy 0 thoughts on “Are Homemade Predictive Behavioral Analytics Applications Better?”

Homemade apple pie beats the Entenmann’s variety, right? Well, only if there’s a good cook at the helm. The same principle holds true when building a data application. Your team may be perfectly capable of building an infrastructure and enterprise application. But what happens when you have to build for a more specialized function, such as predictive behavioral analytics? With an advanced analytics solution, IT teams can, for example, tailor their websites to a visitor in-real-time, delivering the right content at the right time. However, predictive behavioral applications require specialized skill sets that may fall outside the scope of your team.

You’ll know at the time of deployment just how tall this order will be for your company’s IT coder. They will declare victory after they have created a number of scripts, yet still there is no product. That’s why large organizations – including government agencies – get an advanced behavioral analytics solution to sit on top of their open-source big data stack. It’s that lattice top that every homemade apple pie needs. In this way, organizations have complete control over their data, proprietary models and implementation. There is no confidential or proprietary data that ever needs to be shared with an outside vendor. And with an open-source framework, your data scientists have the freedom to build models on top of it if they so wish.

So, the question is which applications should you build in-house and which you should buy. Enterprise analytics applications are designed for enterprise IT to build in-house, whereas predictive behavioral analytics applications are acquired, as long as it is the right one – one that can sit on your data lake and is not a “black box” of proprietary technology.

Do you have a similar experience? Tweet at me or email me.

SHARE THIS ON: Twitter     LinkedIn     Facebook     Email

Copyright 2017 SYNTASA®. All Rights Reserved.