Tag: data analytics

The Data Analytics Trends that Will Shape 2018

Posted by on January 10, 2018

2018-data-analytics-trends

As a field, data analytics is only growing. Not only has the industry of data science broadened substantially, but many companies are finding themselves devoting large amounts of resources towards understanding data analytics and trying to identify new trends. This reliance upon data is only going to grow through 2018, as companies are finding that the data that they collected may contain even more useful information than they previously believed.

2018 is going to find many companies making better use of the data that they already have, and fine-tuning their existing data collection and analysis methods.

Better Personalization Metrics

Industry leaders are hard at work creating incredibly detailed profiles of their customers. Companies don’t need to develop this information themselves. Google, for instance, has fine-tuned its customer profiles and made these customer profiles accessible to those using its analytics and advertising services. Social media platforms have been able to capture customer information similarly, from Facebook to LinkedIn.

The result of this is that advertising is likely to become hyper-personalized to each customer. Not only will companies know the demographics of each customer (age, gender, location), but also their buying habits, how much money they make, and which locations they frequent. Businesses will be able to increasingly target customers and anticipate their needs, ideally creating a situation in which advertising becomes more valuable to the customer.

Augmented Reality Systems

Augmented reality has been kicked around for the last decade, held back by issues regarding processing speed and (perhaps more importantly) battery life. Augmented reality feeds digital information about an individual’s location directly to them, often through a visible “heads up” display.

Not only is this going to change the way individuals interact with the world, but it’s also going to change the data collected. How often do users spend looking at a specific product? Which products or locations do they display further interest in? These will all create incredibly valuable data points that will again be used to create a realistic model of what customers want and need.

Streamlined Data Solutions

Companies have built up their data caches. Now they’re looking for streamlined, agile solutions that can help them make use of the data. In the past, companies were satisfied with collecting as much data as possible and then mining them for as many insights as they could find. Now, companies are more focused on fine-tuning their systems, generating and using the minimal amount of data they need for effective results.

This will create a rise in agile data science, whereby companies will be able to quickly create data sets, respond to and modify these data sets, and produce tangible results from their data sets. In this, the emphasis will be less on the data itself and more what the data can do for the company.

The Science of the Customer Journey

Buyer personas have led to further exploration of the customer journey, a science that attempts to identify the stages that customers go through when investigating and making a purchase. Customer journeys are an incredibly effective way to understand customers and their unique needs.

Data science is likely to be integrated into further understanding of the customer journey. What drives a customer to seek a product? How often does a customer generally research a product? What types of research are most effective and most compelling? What makes a customer more or less likely to find a company and engage in a purchase?

Customer journeys are designed to model customer behavior, so that companies are able to more accurately give customers the information and the prompts they need to continue their journey. In the coming year, this will evolve into a science of its own, and marketers will likely be collecting more customer behavior-related data than ever.

Machine Intelligence Continues to Advance

Alongside all of this, machine intelligence and machine learning will continue to advance. Many businesses have large volumes of data, but it is actually identifying patterns within that data that has become difficult. More advanced machine algorithms will be developed to clean usable, actionable insights from the data that is stored. Machine intelligence will increasingly be used for tasks such as scoring leads, identifying keywords, and targeting specific demographics.

More advanced, learning algorithms are being developed that can, within their parameters, work to improve their own functions. With the right data sets and the right code, marketing algorithms will be able to fine-tune themselves and optimize their own performance. This will be especially useful in A/B testing or split testing, as algorithms will be able to test out different marketing functions and determine the optimal configuration on their own.

Small amounts of this are already cropping up in apps and social media platforms, such as the ability of an algorithm to determine what is most likely to get a profile clicked on, or which photos and posts are most engaging. This can be used in a marketing sense to determine not only which products are most attractive to customers, but which photos they prefer to see, and what descriptions they’re most interested in.

For the most part, 2018 is going to see a maturation of data analytics and data science, as companies invest more money into both collecting and understanding their data. But technology itself is going to play a significant role as well, as the technology behind machine learning and AI is becoming more sophisticated and complex. Either way, companies are going to have to invest more in their data if they want to understand their customers and continue to market directly to them.

DXP Series, Part III: DXP and Data-Driven Decision-Making

Posted by on December 06, 2017

Business team meeting analysis financial chart together at cafe.

Think about how you make important business decisions. Decision-making begins at the point where intuition takes over from analyzing the data.  If your data analysis carries far less weight than intuition, your decision process may not be taking full advantage of available data.

If so, you are not alone. Bi-Survey.com surveyed over 720 businesses. The survey found that 58 percent of respondents based about half of their regular business decisions “on gut feel or experience.” On the other hand, over 67 percent of those businesses “highly valued” information for decision-making, and 61 percent considered information “as an asset.”

The survey showed that when businesses were not using information as the basis for decision-making, it was because the information was not available or reliable. They were either not collecting it or were not using what they had.

KPIs are there, but not the data to read them

Another significant finding involved the role of key performance indicators. There is an important connection between KPIs and the data that measure and drive them. Here is where another disconnect stood out like a beacon: Nearly 80 percent of the companies had defined and standard sets of KPIs, but only 36 percent were using them “pervasively across the organization.”

So there was an obvious disconnect between valuing the information and a willingness to use it. In this post we shall address that contradiction and explore ways to close the gap between valuing the data and using it for data-driven decisions.

How DXP leverages data analytics

The road to data-driven decisions must go through data analytics.  In a previous blog, we discussed how data analytics and other tools plug into the realm of DXP. Data analytics are what help you find meaning in the data you generate and collect.

Those meanings are what drive the decisions and strategies that focus on efficiency and excellent customer service. In terms of business decisions, the ones based on verifiable and quality data are the most beneficial to the business. They are data-driven.

So, data-driven decision management is a way to gain advantage over competitors. One MIT study found that companies who stressed data-based decisions achieved productivity and profit increases of 4% and 6%, respectively.

Two “how-tos” to get on the road to data-driven decision management

#1. How to head towards a data-driven business culture (and benefit from it)

The survey showed that respondents were operating at half capacity when it came to using data-driven decision methods. To unlock the process as well as the data, businesses need to do the following:

  • Focus on and improve data quality.
  • Ease and lower the cost of information access. Break down those proprietary silos and use the best data-extraction tools available.
  • Improve the way the organization presents its information. There are many outstanding presentation products on the market.
  • Make the information easier to find, and speed up the process where users can access the information.
  • Get senior management on board and aware of the value of business intelligence and data-based decision making. Promote a culture of collaborative decision-making.

#2. How to improve internal data management

Data governance (where the data comes from, who collects and controls it) is a major obstacle to taking advantage of data-driven decision benefits. Survey recommendations were that companies should take the following steps:

1. Build an IT architecture that is agile and which can integrate the growing number of data sources required for decision-making. Plug into external big-data sources and start harvesting them.

2. Look for ways to break down barriers to promote cross-departmental cooperation and data alignment. A business intelligence competency center (BICC) can play a major role in achieving that goal.

3. Re-define and use KPIs across the organization and align those measures of success with a focus on data governance.

A strategy for applying data-based decision-making

Bernard Marr in his Forbes online piece, provides the following suggestions for any business to for applying data to decision making:

1. Start simply.

To overcome the overload of big data and its endless possibilities, design a simplified strategy. Cut to what your business is looking to achieve.  Rather than starting with the data you need, start with what your business goals are.

2. Focus on the important.

Concentrate on the business areas that are most important to achieving the foregoing strategy. “For most businesses,” says Marr, “the customer, finance, and operations areas are key ones to look at.”

3. Identify the unanswered questions.

Determine which questions you need to answer to achieve the above focus. Marr points out that when you move from “collect everything just in case” to “collect and measure x and y to answer question z,” you can massively reduce your cost and stress levels.

4. Zero in on the data that is best for you.

Find the ideal data for you: the data that will answer the most important questions and fulfill your strategic objectives. Marr stresses that no type of data is more valuable or inherently better than any other type.

5. Take a look at the data you already have.

Your internal data is everything your business currently has or can access. You are probably sitting on much of the information you know you need. If the data has not been collected, put a data collection system in place or go for external resources.

6. Make sure the costs and effort are justified.

Marr suggests treating data like any other business investment. To justify the cost and effort, you need to demonstrate that the data has value to your long-term business strategy. It is crucial to focus only on the data you need. If the costs outweigh the benefits, look for alternative data sources.

7. Set up the processes and put the people in place to gather and collect the data.

You may be subscribing to or buying access to a data set that is ready to analyze, in which case your data collection efforts are easier. However, most data projects require some data collection to get them moving.

8. Analyze the data to get meaningful and useful business insights.

To extract those insights, you need to plug into the analytics platforms that show you something new. Look for platforms that squeeze out the reports, analysis, and switchboard displays that tell you what you need to know.

9. Show your insights to the right people at the right time.

Do your data presentation in a way that overcomes the size and sophistication of the data set. The insights you present must inform decision-making and improve business performance. Go for style, and substance will take care of itself.

10. Incorporate what you learned from the data into the business.

Here is where you turn data into action. When you apply the insights to decision making, you transform your business for the better. That is the crux if data-driven decision-making. It is also the most rewarding part of the venture.

Summary and Conclusions

1. Business decision-making based on data results in greater reliability, efficiency, and profitability. DXP leverages data analytics towards the goal of more data-based decision making and achieving a competitive advantage.

2. Migrating towards a data-driven business culture requires unlocking the 50 percent of the decision-making and data currently not being used. It requires improved internal data management and governance and breaking down barriers to internal communication.

3. Finally, when those barriers are down, you can begin a strategy for applying data-based decision making. Start simple and focus on what business areas you need to improve and determine what data you need. No data is better or more valuable than any other; the key is to find the data that meets your objective, analyze it, and translate it into actionable decisions and improvement.

Machine Learning is State-of-the-Art AI, and It Can Help Enhance the Customer Experience

Posted by on October 05, 2017

connecting-people

Is artificial intelligence the same as machine learning? Machine learning is really a subset of artificial intelligence, and a more precise way to view it is that it is state-of-the-art AI. Machine learning is a “current application of AI” and is centered around the notion that “we should…give machines access to data and let them learn for themselves.” There is no limit to that data (or Big Data).  The challenge is harnessing it for useful purposes.

In his Forbes Magazine piece, contributor Bernard Marr, describes AI as the “broader concept of machines being able to carry out tasks in a way we would consider ‘smart.’” So, AI is any technique that allows computers to imitate human intelligence through logic, “if-then” rules, decisions trees and its crucial component, machine learning.  Machine learning, as an application of AI, employs abstruse (i.e., difficult to understand) statistical techniques, which improve machine performance through exposure to Big Data.

AI has broad applications…

Companies around the world use AI in information technology, marketing, finance and accounting, and customer service. According to  a Harvard Business Review article, IT garners the lion’s share of popularity in AI activities, ranging in applications that detect and deflect security intrusions, to automating production management work. Beyond security and industry, AI has broad applications in improving customer experiences with automatic ticketing, voice- and face-activated chat bots, and much more.

Machine learning is data analysis on steroids…

AI’s subset, machine learning, automates its own model building. Programmers design and use algorithms that are iterative, in that the models learn by repeated exposure to data. As the models encounter new data, they adapt and learn from previous computations. The repeatable decisions and results are based on experience, and the learning grows exponentially.

The return of machine learning

Having experienced somewhat of a slump in popularity, AI and machine learning have, according to one software industry commentator, Jnan Dash, seen “a sudden rise” in their deployment. Dash points to an acceleration in AI/machine learning technology and a market value jump “from $8B this year to $47B by 2020.”

Machine learning, according to one Baidu scientist will be the “new electricity,” which will transform technology. In other words, AI and machine learning will be to our future economy what electricity was to 20th century industry.

The big players are pushing AI and machine learning. Apple, Google, IBM, Microsoft and social media giants Facebook and Twitter are accelerating promoting machine learning. One Google spokesman, for example, recognizes machine learning as “a core transformative way in which we are rethinking everything we are doing.”

How Machine learning has transformed General Electric…

A striking example of how AI and machine learning are transforming one of the oldest American industries, General Electric, is highlighted in this Forbes piece. Fueled by the power of Big Data, GE has leveraged AI and machine learning in a remarkable—and ongoing—migration from an industrial, consumer products, and financial services firm “to a ‘digital industrial’ company” focusing on the “Industrial Internet.” As a result, GE realized $7 billion in software sales in 2016.

GE cashed in on data analytics and AI “to make sense of the massive volumes of Big Data” captured by its own industrial devices.  Their insights on how the “Internet of Things” and machine connectivity were only the first steps in digital transformation led them to the realization that “making machines smarter required embedding AI into the machines and devices.”

After acquiring the necessary start-up expertise, GE figured out the best ways to collect all that data, analyze it, and generate the insights to make equipment run more efficiently. That, in turn, optimized every operation from supply chain to consumers.

5 ways machine learning can also enhance the customer experience…

Machine learning can integrate business data to achieve big savings and efficiency to enhance customer experiences, by:

  1. Reading a customer’s note and figure out whether the note is a complaint or a compliment
  2. Aggregating the customer’s personal and census information to predict buying preferences
  3. Evaluating a customer’s online shopping history or social media participation and place a new product offering in an email, webpage visit, or social media activity
  4. Intuitively segmenting customers through mass customer data gathering, grouping, and targeting ads for each customer segment
  5. Automating customer contact points with voice- or face-activated “chat bots”

How Rivet Logic can make you future-ready and customer friendly

Your business may be nowhere near the size of General Electric. You do, however, have a level playing field when it comes to leveraging Big Data and machine learning products to a winning strategy. What we do is help you plan that strategy by:

  • Aligning your business goals with technology—What are the sources of your own data and how can they harness the power of NoSQL databases, for example?
  • Designing your user experience—What do you need? A custom user interface, or a mobile app with intuitively simple user interfaces?

We can do that and much more. Contact us and we’ll help make your business future-ready to collect, harvest, and leverage all the great things you are doing now.