How HelloFresh embraced Hadoop

28

Nov
2016
Posted By : admin Comments are off
how HelloFresh embraced Hadoop
Categories :#AnalyticsNews

As businesses grow, it becomes more critical for them to have a solution that will effectively handle the increasing amounts of data they generate. However, one problem that many organisations find when they are expanding is that tools that were adequate when they were developed are not able to scale along with the company.

This was the problem facing Berlin-based home meal delivery firm HelloFresh. The five-year-old firm has expanded rapidly and now delivers more than 7.5 million meals a month to 800,000 subscribers in multiple countries. Therefore, it found itself quickly outgrowing the custom-made business intelligence system it had long relied on, and needed a new solution.

In a recent interview with InformationWeek, chief technology officer at the company Nuno Simaria explained how the company had been using a home-built business intelligence system based around PHP, using a mix of a relational database and key value storage for pre-calculated data. However, as the business grew, the limitations of this became clear.

One problem was it did not offer the flexibility or detail analysts needed. While it could track essential KPIs to provide details of what was happening within the business, it was unable to offer insight into the reasons behind any changes.

"It was definitely not a good idea, but at the time it was the technology we were most comfortable with," Mr Simaria said.

The system was also approaching the limits of its capacity, so it became obvious a change was required. The company looked at several options that would offer improved big data analytics performance, including MemSQL and SAP HANA, but ultimately, it was Apache Hadoop that won out.

Part of the reason for this was its low cost compared with competitors. Because the tools can offer high performance even on inexpensive commodity hardware, there was no need for HelloFresh to upgrade these areas. This made Hadoop a highly attractive option, even though the company's team did not have much familiarity with the technology.

This led to its own challenges. Mr Simaria explained that finding skilled engineers in the market was very difficult. Therefore, the firm's approach was to give two of its existing staff the time and resources they needed to learn about the tools.

"We'll give you the budget, and we'll give you the time," he said. "This is something we've done with other technologies as well. If it is not easy for us to access talent in the market in the short term, we will empower our developers and our engineers who are interested in problem solving, and we will let them discover the complexities of that technology."

At the end of this process, the engineers had to answer three questions: is Hadoop the right technology; how can the firm migrate existing resources to it; and what distribution should be used moving forward?

The result of the Hadoop deployment is that HelloFresh now has much faster insight into goings-on within the businesses, and is also able to delve much deeper into its data in order to uncover insight.

Mr Simaria said: "This technology has allowed us to spread data-driven decision-making to anyone in the organisation, from local teams to global finance to whoever needs to use data insights to make decisions."

Big data to help drive mobile app market past $100 billion

31

May
2016
Posted By : admin Comments are off
big data helping mobile app market
Categories :#AnalyticsNews

The global market for mobile apps is set to reach the $100 billion mark by 2020 as more developers seek ways of monetising their products – but this will only be possible with the help of effective big data analytics.

Prabhjot Singh, co-founder and president of business intelligence service Pyze, told TechCrunch that even though there are millions of entrepreneurs around the world building apps and offering them to the public via Apple's App Store and Android's Google Play, very few of these are financially successful.

Although the world's leading apps can rake in upwards of a million dollars a day, the vast majority have difficulty generating any revenue, with less than 50 per cent of developers making more than $500 a month.

However, this could be changed through effective use of big data analytics. One major reason for the failure of many apps to monetise at the moment is a lack of data and business intelligence that can help developers improve their decision-making and identify better prospects.

But as the technology for evaluating data and turning it into useful information becomes cheaper and more widely available, this presents a great opportunity for smaller app developers to improve their results.

"Companies that can build big data and analytics pipelines to learn about how their apps are being used and who uses them are in the best position to build a community of 'sticky users' who will continually use their apps," said Mr Singh.

He added these players are better able to link into media outlets and social services such as Facebook and LinkedIn to retarget their users and deliver the most relevant, interesting experience. This is something that could not be achieved by the majority of app developers, who may be made up of just one or two people.

Until now, the problem with mobile analytics has been that the solutions that were available did not scale particularly well to provide developers with the type of answers they needed. The tools struggled to effectively analyse millions of data points and then draw out the key information that could give an insight into user behaviour and desires.

Dickey Singh, co-founder and CEO of Pyze, explained that this problem could be solved by using machine learning tools that can automatically cluster groups of users into segments, making it easier to offer more personalised services.

This can enable developers to better monitor usage habits to identify different patterns of app use. 

"From here, an app can begin to develop personalized messages for given users based upon how they use the app," TechCrunch noted. "Nonprofit and for-profit uses of the app can be identified; this enables app developers to more clearly see who is using their apps in a premium, pay-for mode, and where they should be investing their efforts to further monetise their products."

In a study of 12 companies using such tools conducted by Pyze, customer engagement has increased by an average of 35 per cent, while revenue has gone up by 20 per cent.

Prabhjot Singh noted this can help level playing fields between independent designers and large app companies, while also improving rates of engagement for large enterprises that are not primarily in the app business, but that want to use such tools to improve relationships with customers.

Gartner: Global business intelligence and analytics market rises in value

10

Feb
2016
Posted By : admin Comments are off
Gartner: Global business intelligence and analytics market rises in value
Categories :#AnalyticsNews

The global business intelligence and analytics market has seen a large amount of activity in recent years, a new study has indicated. 

Research from Gartner found that global revenue in the business intelligence (BI) and analytics market is forecast to reach $16.9 billion (£11.73 billion) in 2016, representing a 5.2 per cent rise from 2015. 

The study suggested that the BI and analytics market has now reached the final stages of a multi-year transition from IT-led, system-of-record reporting to business-led, self-service analytics. 

Because of this, the modern business intelligence and analytics (BI&A) platform is set to meet new requirements for accessibility and analytical insight. This will help to improve the amount of data and detail available to businesses and improve their efficiency. 

How important is analytics to modern businesses?

Gartner believes that analytics has become an increasingly strategic technology for many businesses, turning into a central element for the majority of roles. The organisation believes that every business is now focused on analytics, with company processes and employees now closely tied to analytics. 

Ian Bertram, managing vice president at Gartner, said: “The shift to the modern BI and analytics platform has now reached a tipping point. Organisations must transition to easy-to-use, fast and agile modern BI platforms to create business value from deeper insights into diverse data sources.”

In order to meet the time-to-insight demanded in today’s competitive business environment, many organisations now want to democratise analytics capabilities by using self-service technology and the movement could be set to continue rising in popularity in the coming years. 

What is different in BI and analytics platforms?

However, there is one key difference between modern BI and analytics platforms and traditional, IT-centric reporting: the amount of upfront modelling required to build analytics content. Creating analytics content via IT-centric reporting includes using IT consolidating and modelling data ahead of time. 

In comparison, a modern BI&A platform accommodates the IT-enabled development of analytics content. Analytics is also now moving into more applications and has now developed to the point where appropriate and inappropriate actions can be put in place for specific industry departments. 

However, BI is where companies truly determine the full value of their data. Without the right ways of interpreting information, it can be challenging for analytics to truly inform business managers about developments in their company. 

BI is designed in order to notice business patterns and trends and allow officers to report on findings, so it is understandable why its value has risen so much in recent years. 

Where could BI&A move in the next few years?

The BI&A market has already gone through significant changes in recent years, causing businesses to reconsider how they think about data management. 

Along with the growth of BI&A, companies need to rethink how to tackle a variety of cyber threats in the industry, ranging from phishing attacks to general common sense approaches to data management. 

With important information now kept on a variety of devices, it is vital that employees are aware of the dangers that come with leaving information unprotected. Encryption should be used on all folders and different passwords must be assigned to key information. Otherwise, companies could be at risk of disclosing vital information to criminals. 

Facebook

Twitter

LinkedId