Карта сайта
Версия для печати

Аналитика для всех: Возможности для работы с данными в режиме «real-time»

5 марта 2013 Раньше аналитикой занимались только люди, наделенные сверхъестественными способностями к усидчивости и терпению. Сегодня же инструменты бизнес-аналитики настолько просты в использовании, что работать с массивами данных стало возможным с молниеносной скоростью и без особых талантов к математическим исчислениям. Более того, благодаря стремительному развитию технологий компании могут анализировать данные в режиме «real-time». (Материал опубликован на английском языке)
With analytical tools that are more user-friendly than ever and breakthroughs in in-memory computing, analytics is moving out of the back room and into the board room — as well as the sales representative’s iPads and the call center agent’s cross-selling prompts. Today, analytics can provide more than just quantifiable insights into what happened in the past. Analytics can provide suggestions, refined in real time, about what we might want to do next — in the next quarter, in the next sales visit, even in the next statement we make to a customer.

If you liked that…

Anyone who has browsed for a book on Amazon knows about its recommendations. Behind the scenes, analytical engines are busily identifying books with similar characteristics, books that customers who’ve bought this book have also bought, even books that other customers have purchased after looking at a given book on the website (even if they have not actually purchased that book). It’s all useful information for customers who, once upon a time, would simply have asked their local bookseller, “What’s new that I’d like?”

These approaches to analysis and recommendation are finding their ways into all aspects of business today. Sales and marketing teams are looking at what their customers have purchased in the past to discover what they might want to buy in the future. They’re doing the white space analysis to see what customers A, B, and C have bought that similar customers X, Y, and Z have not yet purchased (and vice versa).

Service teams are performing predictive analytics, too: They’re looking at trend lines for product returns and repairs and then proactively reaching out to customers who have not yet encountered problems with these products to sell services.

Adding real-time to the formula

While the use of analytics in these business contexts represents a real departure from the back-office analytics of the past, they do share one aspect in common with those Nerds-Only analytical efforts: most of these analytical efforts involve a significant amount computing resources and a significant amount of time. The simple physics of performing an analysis on a huge data store — which may cross hundreds of disk spindles – precludes the possibility of performing these analyses in real time. Using traditional approaches, you can perform the analysis you want, capture the details, and then use the details in real time as you make your plans, engage with your customers, or offer recommendations on your website to drive customers to products they might find interesting.

But if you could perform these complex analyses in real time, that would open whole new possibilities. You could work with real-time data, not data that was old as soon as the analytical snapshot was saved. You could potentially even refine the nature of your analysis while you were engaging with the customer – factoring new information or new needs into the analysis to come up with recommendations that are even more tailored and more personalized to the customer with whom you are engaging.

That’s where today’s in-memory computing technologies come in. With the ability to process and analyze huge volumes of data in memory, in real time, opportunities for analytically-enabled engagements blossom in ways we have never known before. We can monitor trends in real-time, learn from emerging insights, and refine our interactions based on those insights – all on an ongoing basis.

That has significant ramifications for every aspect of business — from the ability to make automated micro-adjustments to your supply chain to the ability to understand just what and when to up-sell a customer who’s calling into the call center, visiting a webpage, or using a location-aware mobile application. Instead of executing inflexibly against a marketing and sales plan that you hope will work, you can execute flexibly against a marketing and sales plan that is being reviewed and refined in real time.

Source:  sap.com