Kayak, the meta-search travel site, have joined the trend for opening up and setting your data free by launching Kayak Trends. Trends allows you to mine through data on what’s actually being searched for on Kayak. You can view the most popular flight destinations (to or from any airport in the world) and the most searched hotels in popular destinations. The data is available on a daily, weekly, monthly or yearly basis and I think it’s really powerful data. Why is it powerful? Because it is a clear indication of consumer trends in travel, what are people looking for right now and how do those trends change across the year. That’s got to have some uses to other people than Kayak and their website visitors.
Travel sites have a huge amount of data on search trends and consumer preferences run through their booking paths and search forms. I can’t think of anyone else who has opened this up in the online travel world. Plenty of other travel companies have merchandising lists of their top destinations etc but they are based on their sales priorities as well as popularity so quite different to this almost realtime user data (if of course it is true data and not tainted). I know of online travel firms who feed this kind of data back into their web apps but not any who give it away in this manner.
So some powerful consumer intent data, I can think of a number of ways you could use this. This is great data for anyone (competitors, affiliates) who want to know what to target destination or hotel wise. Focus on these destinations, if they’re being searched for on Kayak users want product associated with them. Put them on your websites, marketing assets and in your ECRM shots. People must surely be looking at how they can monitor or scrape this data to inform their merchandising decisions, PPC/SEO campaigns (Bid more on the destinations and hotels that are hot on Kayak. Can you assume they will be searched for more on Google when they’re in the Kayak Trends? Long tail flight terms anyone?). You could use it to inform the biasing of your search results, or even to help you highlight areas where your inventory may be lacking.
Kayak will I’m sure have thought of all of that and probably won’t be bothered as their business model is to get as many users through the site and referred on to the agents, airlines and operators from their search results, which they do extremely well.
It’s a great move and I applaud Kayak for being so open. I hope that something positive comes from this; that other online travel sites see this data and realise the power of it, and so start to collect and utilise the data their own websites generate. Too many don’t do that (you may be surprised to hear).
Why don’t businesses measure ROI on social media?
October 2, 2009
Babson Executive Education and Mzinga have published the results of a survey into businesses use of social media. The report shows that while many businesses are spending both money and resources on social media the number of people measuring ROI is low.
The graph below shows that 40% of respondents spend budget on social media and 57% expend resources on it.
So you’d think that they’d be measuring its effectiveness. Wouldn’t you? The pie chart below shows the percentage of respondents who said yes/no to the question ‘Do you currently measure ROI for your social media program?’. Only 16% said yes…
That’s low! It seems to me that there’s a real lack of understanding of online measurement and when it comes to social media that gets even more apparent. Social media measurement requires an understanding of more than just web analytics. You need to grok user behavior, interaction and engagement while also living brand equity, sentiment and what people are saying about you on all mediums. You need to understand the online/offline journey your audience is taking when it interacts with your brand, as an online social campaign could manifest itself as increased (or decreased) offline engagement. It’s a tough mix and there’s no one way to measure all of that right now. The best advice I could offer to businesses who trial social campaigns would be to distill it down to the things that are measurable (there’s always something even if only referrals and changes in rates of sales) and at the same time monitor sentiment like a hawk while listening to your audience (they will tell you a lot about how well you’re doing).
It’s getting better, there are many sentiment analysis tools out there now and you can even create your own dashboards with RSS and something like Netvibes. At the same time smart folk like @willmcinnes and the Measurement Camp crowd are trying to come up with answers to the social media measurement question and there’s a lot of information available on the subject.
So, in my opinion, measuring nothing is inexcusable, measuring or monitoring something is just fine if you’re dabbling in social but if you’re spending a lot of budget and manpower on your social exploits you either need to learn to measure it or find someone who can help you do that.
How to solve problems
October 1, 2009
I found this amusing so thought I’d share. Obviously I don’t suggest any of you use this to help you solve your problems. I can think of a number of people I have worked with in the past who may well have used this (if you’re reading you will know who you are). I can’t remember where I came across it but it tickled me.
What’s your UX (user experience) end game?
September 21, 2009
I dug out and dusted off an old set of slides recently to present to some colleagues on the benefits of taking a truly user centered approach to design and development. The slide below in it’s original form dated from a presentation I gave to a small conference back in 1998 (in ’98 it was Photoshop drawn). Brilliant to dust it off and find it’s still totally relevant (would it ever not be?). It describes the end game, where I strive to get to by adopting a user centered approach.
I get really unhappy with business people who want me to help them deliver business goals with no regard for the UX! I’ve come across this in almost every business I’ve worked in (which is why this slide get’s dusted off every now and then and I talk them through the benefits of UX).
It’s also a great way to get clients or the business to understand that taking a UX centered approach and delivering a top user experience doesn’t have to compromise their business objectives and is often the best way to deliver them!
Of course, everyone’s approach is different and their aim may differ from mine, so what does your UX end game look like?
Is this the future of consuming news online?
September 16, 2009
There’s so much talk about news at the moment with the moves to bring back paywalls (as I’ve written about previously here) and the general nervousness among old media houses that the likes of Google are stealing their thunder (and their ad dollars). Everyone is musing about how this will play out, who will be the winners and what is actually the best way to deliver news to web users. A couple of things have struck me in the last couple of days that deserved a blog post.
Firstly, Google have launched a new product from its labs department. Google Fast Flip is being touted as a new way to read the news, that brings the web experience more in line with newspapers. Basically Google has taken screenshots of web pages containing news stories and put them into a neat user interface that allows you to flip between stories and zoom in on them before deciding whether to click through to the original article.
Google Fast Flip
It’s a really nice way to browse and a great UX to find serendipitous content as you might stumble across something you’re interested in. Of course there’s also Google’s search power under the hood so you can narrow down the content available, isn’t there?
No, it seems that Google hasn’t put it’s search power to good use by letting you search the full content of the articles available (I’ve tested it by searching for specific content within the stories). It is actually just a nice UI to flip and browse through news stories, it doesn’t put that UI on top of the power of Google News though.
Secondly, Microsoft recently unveiled a vision of a ‘Next-Generation Newspaper’ as part of a response to a request from the Newspaper Association of America asking for ideas about ‘monetizing digital content’. The next-generation newspaper delivered as part of their response is remarkably similar to Tweetdeck, as noted by Nieman Journalism Lab today.
Microsofts idea for a 'next-generation newspaper'
The news-deck delivers content in an RSS reader kind of way using a stack of Microsoft technology, content and advertising products. It looks good, and the promise of semantic search, personalisation and contextual awareness is promising.
So, are either of these a ground breaking new way to consume news and do they hold enough promise to herald the future of web based news reading? I don’t think so. They both offer nice solutions to different types of user scenario in my opinion.
Google Fast Flip is a great UI that allows for casual browsing of the news in a similar kind of way to which I read the Sunday newspapers. It’s a coffee and bacon sandwich type situation, but not a serious trawl for information experience for me. It’s too much detail upfront with not enough ability to refine for anyone serious (or used to) discovering content online or even using RSS readers. At least with an RSS reader you can make your choices of who to subscribe to and read articles based on headlines or quick glances at the content. Yes, this is a quick way to thumb the pages but it’s a little gimicky for me and doesn’t fulfill my needs for consuming information. Google partnered with top newspapers to create this but I don’t think it helps them. Digital news content is consumed in a different way to paper content and I think this needs more work before it could ever become the norm for online news reading. Must admit it’s nice on an iPhone though!
Microsoft’s news-deck aproach suits my needs much better. It has the search and refine type features I’d demand and a familiar UI that works for dashboard type experiences. Is it the future? I don’t think so. I actually think it’s a little lame of Microsoft to propose a solution like this to the challenge from the NAA, it’s too similar to RSS readers and looks too like a cross between Netvibes and Tweetdeck. In fact if you read the details of their response (available from the Nieman Labs post) it really does just sound like next-gen Netvibes to me. It’s not really the future, more what we have now mk2.
Now I’m not sure what the future of news consumption is but I’m pretty sure this isn’t it. If you could combine the two you might be getting somewhere, if Fast Flip allowed me to search and read all the content and not click off to the news source it might be better and if Microsofts dashboard wasn’t just a rehashed Netdeck/Tweetvibe it might get me more excited. Personally I’m seeking something more intelligent and with better data mining possibilities to satisfy my search for news (and I wouldn’t limit it to news, any solution should be my medium for information consumption as a whole).
One thing this does make apparent though is that as much as the newspapers and old media want to pull their content in and keep it close, they are going to have to relinquish their hold and let their content free as whatever the user experience of the future is it will demand that.
Are bit.ly statistics accurate?
September 10, 2009
I’ve never really looked at the bit.ly stats that you get for every link you shorten using the service (handy tip, add a + to the end of any shortened bit.ly URL to see the stats). So this evening I did a little experiment. I had a need to create four shortened links each directing users to a different page. Each page has Google Analytics tracking installed and also another well known analytics package (not free). To each link I also added a click tracking parameter for the ‘not free’ analytics package. All four links were distributed only via tweets on a Twitter account (not my personal one, and not a well followed one which is why the numbers are low, but it suited the purpose) using the Twitter web interface. So at the end of this little test, I should have three sets of figures. One from bit.ly, another from Google Analytics (counting referrals from Twitter to each page) and a third from the other analytics package.
The results:
| bit.ly | Google Analytics | Other analytics | |
| Link 1 | 5 | 10 | 11 |
| Link 2 | 8 | 16 | 18 |
| Link 3 | 2 | 12 | 12 |
| Link 4 | 1 | 20 | 24 |
There’s nothing particularly scientific about this, I just checked out the bit.ly stats and noticed the discrepancy. Both Google Analytics and the other stats package use cookies to track users while bit.ly (I assume) can count directly any clicks on its shortened URLs on its servers. There may be a good reason for this difference, perhaps bit.ly has been having some server issues today, but the difference is enough to make me think twice about trusting bit.ly.
I’ve always used bit.ly for shortening links, for no reason other than it’s conveniently embedded in tools such as Tweetdeck, but I may consider trying another one of the multitude of services that do the same job. Obviously bit.ly shouldn’t be relied on for your main traffic analytics, it’s not their main business, but should we expect its stats to be more accurate (or at least closer to other analytics packages)? These are low traffic links but on high traffic ones how wide would the gap be then? Would be interested to hear what others think, especially if you know of a reason for the difference.
Digital facade
September 6, 2009
Adam Greenfield (or @agpublic) spoke at dConstruct 2009 on Friday about Networked Urbanism (which was great) and mentioned a developing trend for architects to put digital skins on buildings. I’d seen some of this kind of thing in Tokyo last time I visited and it’s an interesting idea. So I’ve been having a look around and found some amazing examples on realities:united website. They describe themselves as a studio for art, architecture and technology.
They’re latest work is Crystal Mesh which is on a building in Singapore. They say ‘ILUMA “Urban Entertainment Centre” building the ability to dynamically express itself and to dynamically react to changing activities within and to the urban surrounding. The installation, which is identical with the building skin aims to extend the architecture and the building’s inner activities. Thus it does not represent a defined screen nor does it aim at the broadcasting of conventional (TV style) content.‘
It looks awesome!
So far most of the examples I’ve found are essentially large, permanent art installations. This could get really interesting though when the digital skin starts to react to the building or local environment through the use of sensors, that would really make buildings blend into their surroundings and almost feel organic. Having buildings which could interact with each other would be cool.
Here’s another crazy example from realities:united on the Kunsthaus Graz art museum in Graz, Austria. Amazing building with an amazing facade.
Kunsthaus Graz art museum
