2 items tagged "succes"

  • How Big Data is changing the business landscape

    jpgBig Data is increasingly being used by prominent companies to outpace the competition. Be it established companies or start-ups, they are embracing data-focussed strategies to outpace the competition.

    In healthcare, clinical data can be reviewed treatment decisions based on big data algorithms that work on aggregate individual data sets to detect nuances in subpopulations that are so rare that they are not readily apparent in small samples.

    Banking and retail have been early adopters of Big Data-based strategies. Increasingly, other industries are utilizing Big Data like that from sensors embedded in their products to determine how they are actually used in the real world.

    Big Data is useful not just for its scale but also for its real-time and high-frequency nature that enables real-time testing of business strategies. While creating new growth opportunities for existing companies, it is also creating entirely new categories of companies that capture and analyse industry data about products and services, buyers and suppliers, consumer preferences and intent.

     

    What can Big Data analytics do for you?

    *Optimise Operations

    The advent of advanced analytics, coupled with high-end computing hardware, has made it possible for organizations to analyse data more comprehensively and frequently.

    Analytics can help organisations answer new questions about business operations and advance decision-making, mitigate risks and uncover insights that may prove to be valuable to the organisation. Most organisations are sitting upon heaps of transactional data. Increasingly, they are discovering and developing the capability to collect and utilise this mass of data to conduct controlled experiments to make better management decisions.

    * React faster

    Big Data analytics allows organisations to make and execute better business decisions in very little time. Big Data and analytics tools allow users to work with data without going through complicated technical steps. This kind of abstraction allows data to be mined for specific purposes.

    * Improve the quality of services

    Big Data analytics leads to generation of real business value by combining analysis, data and processing. The ability to include more data, run deeper analysis on it and deliver faster answers has the potential to improve services. Big Data allows ever-narrower segmentation of customers and, therefore, much more precisely tailored products or services. Big Data analytics helps organizations capitalize on a wider array of new data sources, capture data in flight, analyse all the data instead of sample subsets, apply more sophisticated analytics to it and get answers in minutes that formerly took hours or days.

    * Deliver relevant, focussed customer communications

    Mobile technologies tracks can now track where customers are at any point of time, if they're surfing mobile websites and what they're looking at or buying. Marketers can now serve customised messaging to their customers. They can also inform just a sample of people who responded to an ad in the past or run test strategies on a small sample.

    Where is the gap?

    Data is more than merely figures in a database. Data in the form of text, audio and video files can deliver valuable insights when analysed with the right tools. Much of this happens using natural language processing tools, which are vital to text mining, sentiment analysis, clinical language and name entity recognition efforts. As Big Data analytics tools continue to mature, more and more organisations are realizing the competitive advantage of being a data-driven enterprise.

    Social media sites have identified opportunities to generate revenue from the data they collect by selling ads based on an individual user's interests. This lets companies target specific sets of individuals that fit an ideal client or prospect profile. The breakthrough technology of our time is undeniably Big Data and building a data science and analytics capability is imperative for every enterprise.

    A successful Big Data initiative, then, can require a significant cultural transformation in an organisation. In addition to building the right infrastructure, recruiting the right talent ranks among the most important investments an organization can make in its Big Data initiative. Having the right people in place will ensure that the right questions are asked - and that the right insights are extracted from the data that's available. Data professionals are in short supply and are being quickly snapped up by top firms.

    Source: The Economic Times

  • Recommending with SUCCES as a data scientist

    Recommending with SUCCES as a data scientist

    Have you ever walked an audience through your recommendations only to have them go nowhere? If you’re like most data scientists, chances are that you’ve been in this situation before.

    Part of the work of a data scientist is being able to translate your work into actionable recommendations and insights for stakeholders. This means making your ideas memorable, easy to understand and impactful.

    In this article, we’ll explore the principles behind the book 'Made To Stick' by Chip Heath and Dan Heath, and apply it within the context of data science. This book suggests that the best ideas follow six main principles: Simplicity, Unexpectedness, Concreteness, Credibility, Emotions, and Stories (SUCCES). After reading this article, you’ll be able to integrate these principles into your work and increase the impact of your recommendations and insights.

    Simple

    Making an idea simple is all about stripping the idea to its core. It’s not about dumbing down, but about creating something elegant. This means that you should avoid overwhelming your audience with ideas. When you try to say too many things, you don’t say anything at all. Another key component to making ideas simple is to avoid burying the lead. If during your analysis you find that 10% of customers contribute to 80% of revenues, lead with that key insight! You should follow an inverted pyramid approach where the first few minutes convey the most information, and as you get further you can get more nuanced. Analogies and metaphors are also a great way to get your ideas across simply and succinctly. Being able to use schemas that your audience can understand and relate to, will make it a lot more digestible. For example, a one-sentence analogy like Uber for X can capture the core message of what you’re trying to convey.

    Unexpected

    An unexpected idea is one that violates people’s expectations and takes advantage of surprise. You can do this in several ways, one of which is making people commit to an answer, then falsifying it. For example, asking to guess how much time employees spend doing a task you’re looking to automate before revealing the real answer. Another way to generate interest and leverage the unexpected principle is to use mysteries since they lead to aha moments. This might take the form of starting your presentation with a short story that you don’t resolve until the end, for example.

    Concrete

    Abstractness is the enemy of understanding for non-experts. It’s your job as the data scientist to make your recommendations and insights more concrete. A key to understanding is using concrete images and explaining ideas in terms of human actions and senses. The natural enemy of concreteness is the curse of knowledge. As data scientists, we need to fight the urge to overwhelm our audiences with unnecessary technical information. For example, reporting on the Root Mean Squared Error of a model, may not be as helpful as breaking up the language into more concrete terms that anyone can understand.

    Credible

    Adding credibility to your recommendations can take three forms. The first is the most common one when we think of credibility, which is leveraging experts to back up claims or assertions. Another way is using anti-authorities who are real people with powerful stories. For example, if you’re talking about the dangers of smoking, the story of someone who suffers from lung cancer will be a lot more impactful than a sterile statistic. The third way of adding credibility to your story is by outsourcing the credibility of your point to your audience. This means creating a testable claim that the audience can try out. For example making the claim that customers from region X take up 80% more customer support time than any other region. In posing this claim, your audience can confirm this claim which can make it easier to lead to your recommendation.

    Emotions

    Weaving an emotional component to your ideas is all about getting people to care. Humans are naturally wired to feel for humans, not for abstractions. As a result, one individual often trumps a composite statistic. Another component of emotions is tapping into the group identities that your audience conforms to. By keeping those identities in mind, you can tie in the relevant associations and evoke certain schemas that your audience will be most receptive to. For example, if you know one of your audience members is a stickler for numbers and wants to see a detailed breakdown of how you arrived at certain conclusions, adding an appendix may be helpful.

    Stories

    Humans have been telling stories for centuries and they have proven to be one of the most effective teaching methods. If you reflect on the books you’ve read in the past 5 years, you’re more likely to remember the interesting stories rather than objective facts. When weaving stories into your recommendations, make sure to build tension and don’t give everything away all at once. Another useful tactic is telling stories which act as springboards to other ideas. Creating open-ended stories that your audience can build on is a great way for them to get a sense of ownership.

    Next time you’re tasked with distilling your insights or pitching recommendations, keep in mind these six principles and you’ll be creating simple, unexpected, concrete, credentialed emotional stories in no time!

    Author: Andrei Lyskov

    Source: Towards Data Science

EasyTagCloud v2.8