(Editor’s note: David Stein is co-CEO of Rypple. He submitted this story to VentureBeat.)
The annual review is one of the most dreaded tasks on any employee’s to-do list. Companies won’t give raises and promotions without it. Human resource departments demand the forms, supposedly for legal reasons. And the process is fraught with anticipatory angst, defensiveness and dread.
But the worst thing about annual employee performance reviews? They simply don’t work.
The performance appraisal system’s greatest shortcoming is that it’s not really about improving performance. It’s about sticking to a schedule and complying with human resources guidelines. By emphasizing process over results, the performance appraisal system has fatally lost credibility.
Samuel Culbert, in his book, “Get Rid of the Performance Review: How Companies Can Stop Intimidating, Start Managing – and Focus on What Really Matters,” described the annual employee performance appraisal as a “pretentious, bogus practice . . . that produces absolutely nothing.” Workers require “evaluations dictated by need, not a date on the calendar.”
A Psychology Today article cites a 2005 national survey by consulting firm,People IQ, which reported “87 percent of employees and managers felt performance reviews were neither useful nor effective.”
In short, performance reviews are dangerously archaic. When the economy was about how many widgets an assembly line worker could solder in a day, maybe they made more sense. But the assumptions underpinning annual reviews—standardized job descriptions, strict hierarchies and (above all) stable market conditions that allow for annual planning—are completely out of date. Companies must change this process now. Here’s why:
The current high unemployment figures belie the tight labor markets in certain sectors, particularly knowledge industries. These workers want more than good pay; they expect a level of coaching and professional development that traditional performance appraisal systems don’t support.
Also, the so-called “Generation Y”—as many as 80 million in the US alone—will soon dominate the workplace. And this generation expects constant feedback. Significantly fewer baby boomers feel the same way.
Social software tools make it possible for supervisors to reclaim their role as employee coaches rather than strictly managers. They also make it possible for companies to accurately measure performance.
The best feedback-oriented systems have the following characteristics in common:
Regular coaching: Using tools like Rypple, among others, managers and employees can keep weekly notes of agreed objectives and goals. This lets both coach and employee quickly assess progress, say thanks, offer immediate feedback and reprioritize goals, as necessary.
Broad-based, real-time feedback: Managers can’t be everywhere. Particularly in today’s hyper-speed workplace, the contributions that matter may happen so fast and unexpectedly that they never make it onto a form. Most social software tools have a public feed where everyone on the team can comment and say thanks. Scan that feed periodically and see who gets called out for going above and beyond. It may not be who you think.
Regular review: Instead of filling out lengthy forms once a year about long-ago (and fuzzily remembered) activities, schedule a 30-minute session every quarter and use the tool to go over what actually happened during the past few months. The review should include not just formal goals from 1:1s, but comments and thanks from the public feed that round out the picture.
Bottom line: human memory is fallible. Evaluations based on actual data recorded in real-time can take the anxiety out of reviews and provide a better picture of employee performance.
By integrating feedback into everyday corporate culture, companies can change evaluation from a once-a-year chore to a valuable everyday tool for insight and learning. And the dreaded annual performance review can finally be laid to rest.
Web apps are helping storytelling evolve. Just look at how Storify is letting journalists pull together disparate snippets of social media into a coherent tale, while Memolane makes it easy for anyone to browse the photos, videos and status updates that they’ve shared online.
Now a new tool called Tiki-Toki has launched from a UK-based team, allowing users to build good-looking timelines that tell multimedia stories. Journalists and teachers will find this particularly useful, but anyone who wants to build a timeline to visualise a series of events should give it a go.
It’s important to note that unlike Memolane, Tiki-Toki isn’t automated – you have to do the work of manually inputting events on your timeline, although given the far wider range of uses Tiki-Toki has, that’s understandable. Offering a date range of 9999BC through to 9999AD (if you want to predict the far, far, future), you can add text, photos or videos at any point. Flickr integration makes picking photos stored there particularly easy, while YouTube and Vimeo embedding is supported.
Despite its gorgeous appearance, Tiki-Toki’s interface is a little frustrating at times. It’s the little things – like not starring out your password as you log in (really!) and the occasionally sluggish response times (even using Chrome 10) that annoy from time to time as you work with it.
That said, those kinks can be worked out, and the results of using Tiki-Toki can be incredibly satisfying. Just look at this example timeline tracing ‘The Fight for Democracy in the Middle East’ for example, or this one representing the life of Mary Kearns to see the quality of what can be produced.
Free accounts let you create one timeline, while premium accounts are available at prices starting at $5 per month, allowing you crate more timelines and embed them into your own website. A bespoke service allowing you to install Tiki-Toki onto your own server is also available. Details of premium accounts are currently only available once you;ve created a free account – something that probably needs fixing. An introductory video wouldn’t go amiss too.
Still, it’s early days for Tiki-Toki and it’s already looking like a highly useful tool with plenty of promise for the future. Tiki-Toki is available to try here.
The iPad 2’s new hardware and its popularity has led to speculation that tablet devices will soon replace consoles. Rovio Mobile’s Peter Vesterbacka is the most vocal champion of this position and recently claimed during a South by Southwest panel that console games are “dying.” Vesterbacka points to the lower price point of mobile games and their ability to be updated rapidly as two factors that will push mobile gaming ahead of consoles.
While Vesterbacka talks smack about consoles, former Sony executive Phil Harrison takes a much more balanced approach towards this subject. Harrison acknowledges that tablet hardware is evolving so rapidly that tablets could rival that of the console. The Apple tablet has more than doubled its performance in a year, while the Xbox 360 and PlayStation 3 remain unchanged after five years.
Though the hardware evolution may outpace the consoles, the gaming experience of a tablet will be different than that of the console. “I think while there are clearly some compromises in terms of user interface and user experience on a tablet, the touch screen interface will enable other forms of game experience and game design that you couldn’t do on a TV-based console,” said Harrison, “so I think there’s going to be a tradeoff. But I think the pads/tablets market is going to be a very vibrant market for games.”
Harrison points out the strength of the tablet gaming market comes from the diverse business models available to mobile game developers. Developers can choose to generate revenue from free, ad-supported applications, games with in-app micro transactions and paid games with value-added features like social networking. “These are 21st century business models that are going to prevail,” says Harrison.
I was recently re-reading an old story from a 1998 edition of The Onion mocking Compaq (remember them?) for introducing a $4995 computer that can – gasp! – play TV shows downloaded from the Internet. I remember reading it in 1998 as the Great Internet Bubble was in full swing. It was funny then – who would spend $4995 on a computer and install an unthinkably expensive high-speed 1.54Mbit/s T1 Internet line just to watch TV? – but it’s even funnier now because watching TV on our computers is exactly what we’re all doing. It's now so commonplace that the TV industry is running scared of the ‘cable cutters’ who are giving up TV to get their video from the web.
Probably the most exciting advance in what The Onion once mocked as outrageously expensive is the new category of low-cost computing device: the tablet. Apple has proved with the iPad that we can have a lightweight device with a long battery life that’s great for watching video.
Unfortunately, it’s not quite complete: the majority of web video (and nearly all the TV shows from that Onion story) is provided using Adobe Flash – famously excluded from the iPad by Steve Jobs in a public spat with Adobe that were dubbed “Flash Wars.” Apple has tried to move the web away from Flash and toward open standards like HTML5 and a format that the iPad can play. But YouTube has explained why it is sticking with Flash – video codec choice (there is no universal codec that will play on all PCs, unlike Flash) and Digital Rights Management (HTML5 doesn’t have any).
While it's true that YouTube is often accessed via a dedicated tablet app (although with restricted content), video on the web is fragmenting rather than consolidating: sites like Vimeo and Mpora are gaining viewers, and I think it’s safe to say that for several years to come there are going to be major web video sites that will be using Flash.
So why aren’t we seeing Flash-enabled browsers on tablets? Plenty of manufacturers have announced Flash capability but we’ve seen little evidence of it running well. Unfortunately, it turns out that Flash Wars is more than just a bit of industry politicking: there are some very tricky technical problems to implementing Flash on an embedded processor – and it’s not going to be solved by locking programmers into a room and yelling “fix it!” at them.
Flash was designed in the mid-90s for PCs and the new Internet era to display multimedia content in a web browser. It decodes video in software, combines the video and graphics in software, synchronizes the video and audio in software, handles Internet dropouts in software, and copies the results as fast as possible into the web browser. It relies on the CPU being fast and handling RAM accesses very quickly.
This is almost the complete opposite of how digital TV set-top boxes are built for video: they have slow CPUs hooked up to dedicated video decoders, which are directly linked to the display. The CPU basically is a manager for all the intensive video processing, which goes straight to the TV. Tablets are built from embedded processors that are designed for low power consumption and low cost: the Apple iPad (running its own ARM-based A4 processor) consumes about 1.7W including the screen. Compare this to the 8W for just Intel’s Atom 330 processor. The flip-side is that the Atom runs general software applications a lot faster.
Getting Flash to run on a tablet efficiently means that the video must be handled by a hardware video decoder, like the digital TV set-top box does. And indeed, Apple’s iPad can do 720p HD video decode without breaking a sweat (as anyone who has used an iPad to watch movies knows – it hardly dents the battery at all).
But here we hit the first problem with implementing Flash: video codecs. The H.264 standard is very common. It’s used in all kinds of places, from digital HDTV to Blu-ray to iTunes. It’s widely supported in embedded processors and it’s one of the three video codecs used by Flash (the other two being VP6 from On2 – the company recently acquired by Google – and Sorenson Spark). The most common format for Flash video is H.264. So what’s the problem? Why not get the Flash software to call up drivers for video decode hardware and get that to do the work? The problem is that there’s no such thing as just ‘H.264’ – the standard was drawn up to satisfy contradictory requirements. Some wanted it to be easy to decode (for cheap and low power video hardware) but others wanted sophisticated compression (for efficient use of bandwidth).
The H.264 committee decided to please everyone by producing ‘profiles’: Baseline Profile (BP), Main Profile (MP) and High Profile (HiP). A chip salesman may well say “my chip does 1080p HD H.264” but what he mumbles behind his hand is crucial: “for Baseline Profile video”. Adobe’s Flash software for the PC supports H.264 HiP, and lots of web sites encode their video for this profile (after all, if Flash decodes it and the bandwidth is lower, why not?). Unfortunately, it means that lots of web video won’t decode on H.264 BP.
So these embedded processors have to use software to decode the video – soaking up CPU time and battery power, the very thing they shouldn’t be doing. In many cases, the processor won’t be able to decode the video fast enough and the video won’t play at all. So why don’t they just support H.264 HiP in hardware? The answer is that most of the embedded processors used in tablets today were designed a couple of years ago for the then smartphone market, and this market wasn’t asking for H.264 HiP video: everyone was expecting to sell music videos for $5 a pop from me-too online music stores that had H.264 BP video encoded for small screens.
One solution to the video codec problem is of course to get everyone in the world with a web video site to re-encode the video using H.264 BP. This I call the ‘Field Of Dreams’ approach (“if we build it, the content will come”) and as a commercial model it’s been a miserable failure countless times before (remember WAP?). I’m not holding my breath. But even if we assume the video decode problem magically goes away (maybe the ghosts of famous baseball players will help out with the re-encoding) there’s still another problem: memory bandwidth.
Graphics and video have to be merged together for display (usually so that text and icons appear over the top of video) and then synchronized with the sound (a process called A/V sync). In digital TV this is pretty straightforward and is done in hardware: the output from the video decode is mixed with the output from the graphics chip to form the picture (a particular color is chosen as transparent, and video shows through: this is the ‘green screen’ technique for superimposing one video source on top of another). It’s very efficient and the hardware to do this is very cheap.
In Flash, designed to run on PCs, this is of course done in software: each decoded video frame is copied to a piece of RAM, the graphics and text are overlaid and the result copied again into the web browser and finally the visible part of the web page is copied into the display hardware for output on the screen. This whole process has to be done 30 times a second to give good quality video. On a PC this isn’t too bad: they are designed for doing this kind of thing. For embedded processors, it’s a problem: again, it’s CPU time and power consumption that makes things run sluggishly and runs the battery down. And the bigger the screen, the more data to be handled. The solution is again to get Flash to use specially-designed hardware for doing all this work: to have dedicated hardware pathways for converting video, for layering graphics and for scaling the outputs so that the huge amount of data being handled doesn't hit the CPU and soak up memory bandwidth. Unfortunately, we are only just now seeing embedded processors with this kind of video handling hardware coming to market and Flash needs to be adapted to use this hardware.
I can’t see how we are going to see tablets that can do a good job of showing HD video from Flash web sites until we get tablets with embedded processors that are up to the job and the system software developed to use that hardware properly. If anyone tells you they have a tablet that does Flash video really well, arch your eyebrow in a Spock-like skeptical manner and ask to see it.
The incident took place during a game that Spooner called between the Minnesota and Houston on Jan. 24. Timberwovles coach Kurt Rambis got upset with Spooner over a foul called on one of his players.
Spooner told Rambis he would review the call at halftime, prompting Rambis to ask how he would get those two points back.
Spooner says he ignored that question, but Krawczynski tweeted during that game that Spooner “told Rambis he’d ‘get it back’ after a bad call. Then he made an even worse call on Rockets. That’s NBA officiating folks.”
Krawczynski works for the AP (who stands by his reporting) and has been critical of the NBA’s officiating problems. Spooner wants an apology, plus $75,000 in damages.
It’s possible that Spooner meant “you’ll get it” back in a sarcastic manner and then simply just made another terrible call. In other words, referees make enough unintentionally bad calls in every game that it always evens out.
