细节决定成败

A key live streaming event just concluded in which your corporation’s CEO gave a presentation to a number of employees and stakeholders. The management team wants to know details about the viewership, so they ask you to provide these details.

With an open-ended request like this, what level of detail do you provide? Better yet, what type of details can you provide? Welcome to the world of reporting and analytics, which has as many options as a high-end encoding box has formats, 编解码器, 和比特率.

在光谱的一端, you may think that the CEO only wants to know how many people clicked on the live streaming link. 如果是这样的话, 一个有用的, 简单的总结概述是有序的, at the level of URL reporting such as 谷歌分析 (which is based on the Urchin web analytics strategic business unit)—or Webtrends might provide.另一方面, management may want to know how many people from each satellite office watched the stream in its entirety, what percentage of viewers dropped off before a key point was made, whether viewership rose throughout the event, or whether social media usage rose during those key points. In other words, you need expert analysis to identify trends not easily seen from a summary overview.

在这个买家指南中, we’re going to consider five key questions that may help determine what kind of reporting and analytics package to use.

At the outset, 虽然, you need to know what you want to measureand why you want to measure it. Simply having reporting and analytics will not be helpful unless you know how to use them; just having the raw data can actually be counterproductive without the business intelligence to turn the data into actionable 信息.

免费与付费
This may be the biggest question raised during the initial search for reporting and analytics tools. There are numerous reasonsto choose a free option. 但有时候，免费也会变得昂贵. One of the biggest limitations of free software is the amount ofdetailed 信息 that can be gathered and shared. Most analytics programs are geared towards summary 信息 aboutwebpages, not streaming content.

也, while many free solutions have expanded their graphical capabilities, moving from raw text-based logs to representational graphs and diagrams, these same programs are provided on a best-effort basis and have no service-level agreement, 可靠性, 或者质量保证. Finally, free products frequently do not have very compelling privacy policies. 谷歌分析, 例如, uses aggregated web analytics to optimise its own search engine model.

支付解决方案, 另一方面, often have much more detailed 信息, including demographic cross-references down to the ZIP code level, a guaranteed quality of service level associated with the subscription, and an explicit privacy statement that identifies the data being gathered as belonging to the subscriber rather than to the reporting service.

实时与延迟
A subset of the discussion around any reporting service is the question of how quickly reports and analyses must be delivered.关于这个话题已经写了很多, so we’ll just touch on the highlights of the debate of real-time reporting versus delayed reports. 仅用于流分析服务, especially for those who may do subscription-based entertainment streaming, the answer may be that a 24-hour turnaround is acceptable since the entertainment industry has long been accustomed to receiving its viewership numbers the following day. Yet even for these companies, the ability to track viewership trends in real time may be beneficial.

Are real-time statistics worth paying a premium for? 在某些情况下，确实如此. 例如, many ecommerce sites need detailed reporting of inventory levels to determine if a daylong sale’s video advertising is having the best impact or whether there’s a need to switch back to more traditional advertising. 即使是现场活动, real-time data means that critical decisions can be made in the moment, 甚至在活动结束之前.

最后, 虽然, all the discussions seem to boil down to one adage: I want my reports quickly, 我希望我的报告准确无误. I’m willing to wait a reasonable amount of time to have accurate numbers, but I’d rather have accurate numbers immediately.

The other concern that comes into play is which 信息 needs to be available in real time. The next comparison addresses that question.