Google Wave was an island

There was no easy nor logical way to incorporate Wave into your ordinary workflow. If you weren’t opening wave.google.com on your browser, you were not “on the wave”. The most critical thing was that even though it looked almost exactly like a webmail client, and your user account had the form of username@googlewave.com, there was no email integration whatsoever. It’s bad enough that you couldn’t subscribe to any Wave updates to your inbox (later this feature was added without much fanfare), but the fact that a company hosting one of the largest email services in the world goes and assigns users pseudo email addresses you can’t send messages to is something that still boggles my mind. Ok, I understand that Google Wave was supposed to be something beyond email, but even emails could be printed on paper. How about some backward compatibility, eh?

Google Wave did not solve a specific problem

Most people just couldn’t quite figure out what exactly they were supposed to use Wave for. Pretty much everybody saw the potential of it for something useful, but were they able to picture themselves as a user in a specific use case where Wave was a natural fit? Let’s face it: there’s no point in releasing a technology demo and expect people to start using it straight away. What in fact was missing was the real product. Wave gave us the tools, but it would have probably taken an ecosystem around it to turn these tools into products that people could utilise for solving a specific problem (which Google did try to encourage through it’s API and federation protocol offering). Yes, collaboration challenges tend to be universal, but that doesn’t mean you could simply throw technology at them and expect people to take it from there.

Google Wave was not social

Given that the beta program with its invitation only approach initially stirred up quite a lot of interest towards the service, with invitations sold on eBay for $70, the approach quickly turned on its head. Let’s say that there were several groups of people in your company that were experimenting with Wave. How could you identify these particular people that were potential co-wavers and collaborators? Not through the service you wouldn’t. It happened on the corridors and water coolers, or in long email threads (sigh) circulating around the office. Come to think of it, for a company revolving around the concept of search, Google’s products can be incredibly hard to find. Just compare the findability of users and information on Twitter vs. Google Buzz.

Google Wave was developed by… Google

Huh? Why is that a bad thing? Well, let me explain. Google knows us all, but we still have no effective way of knowing each other through Google. This lack of the social dimension is perhaps the most significant barrier currently limiting Google’s growth beyond “just” indexing all the world’s data. 10 years ago that must have seemed like the ultimate goal you could have within the scope of Internet. Ever since the ‘net has evolved from a collection of documents to a platform for human interaction, that goal no longer sounds like the Holy Grail. Wave, just like Buzz or Orkut, relies on tapping into the social behaviour between people, not just a lone swordsman searching for a piece of information in the sea of data. Designing services to this new breed of customers requires a new way of thinking, which Google is only in the process of learning. Read this excellent post for further analysis on why Google cannot build social applications.

Just because the “surf’s up” for Wave, doesn’t mean the game’s over for Google’s initiative to change how people collaborate on information online. On the contrary: because they keep on trying, they can eventually succeed. Obviously there’s no other path for the corporation that’s built its product success on the concept of perpetual beta. I’ll hold up my board and wait for the next big wave.

Having said that, hosting your own WordPress site does require you to perform regular maintenance. For many of us this is not exactly the most rewarding part of runnign a website, but if you want to go beyond what hosted services offer you, then it’s just the price to be paid. I had been skipping payments for a while, so now when I finally had some well deserved time off in my hands, I knew the time had come for a summer cleaning effort.

WordPress 3.0 upgrade

With any application’s major version release, you may not want to be the very first user to install it. I certainly took a while to jump on version three and actually waited until 3.0.1 was out. The main reason was not the fear of native WP bugs as such but rather the plugin compatibility.

Before working on any plugin upgrades and installation, the natural first step is of course to patch up the core WordPress installation to the latest version. Before the upgrades, we are always instructed to take full backups of our precious data and other files, but let’s face it: how many of us really go through the trouble? Well, this time I thought I’d download a new full snapshot of my domain (0.5 GB of data), just to be on the safe side. I do have scheduled backup jobs running on the server, but there’s nothing quite like having the bits sitting on a local drive right next to you.

Luckily everything worked like a charm and WordPress 3.0.1 installed itself without any hick-ups. Had I actually needed to roll back to a previous version, things might have gotten interesting, since I’ve not once performed such an operation. I bet the 90-9-1 principle could be applied here: 90% never bother with backups, 9% take them and 1% know what to do with them.

Sharing links

Next in line was doing something visible for the sites. It has become pretty obvious that our methods of consuming web content have been transformed by the social media revolution, which in turn has an effect on how people reach your site. Search engines are still cool and SEO does matter, but referrals are what’s truly worth gold in the new social online economy. Therefore, you certainly want to make sharing content from your sites as simple as possible.

I’ve noticed myself being far more likely to retweet or share links if there was a convenient button available right there on the website to do so. This is one of those cases where your initial reaction from a developer perspective (“who needs these annoying external scripts and pop-ups, when copy-pasting text is trivial”) is quite different from your behaviour when acting as a user (“can I quickly get this link out there so I can move on to the next site I already have in mind? Hmm, can’t see a button, so I’ll rather skip it”). At the end of the day, it’s a usability improvement that’s difficult to argue against.

Looking at what WP content sharing plugins there were available, AddToAny seemed to be the most up to date offering. However, it turned out to have someundesireable tracking features built in. After trying out some smaller, less commercial options like Sociable, I came to the conclusion that it’s just better to bite the bullet. AddThis ended up being my weapon of choice, as it seemed to provide a good selection of services and customizability(not too fond of the header.php modification for Twitter message configuration, though) with an acceptable level of privacy breach (for example, flash cookies). Pre-selecting what sharing services to support would be a futile attempt, so it’s much better to outsource this burden to a development team rather than worrying about service links and icons myself.

How about the Facebook like button then? Well, let’s  just say that I haven’t yet fully figured out how what specific problem the Likes on FB are solving. Therefore I’ll rather stick to well known sharing methods and reduce the amount of clutter on my sites. Enabling it through AddThis can be done quickly if I change my mind.

Caching

One thing the information overload is definitely doing to our brains is shortening the attention span for any single piece of content. At worst, our attention may already be diverted elsewhere before we even get to the content, if the delivery channel is not responding fast enough. Also Google knows this, which is why the page load times are now a factor used in determining PageRank.

I know, the web traffic to my sites will surely never have such peaks that I’d need to be worried about them. What I am concerned about is the basic ability of my web hosting providerto serve the dynamic content from the WordPress application fast enough. With a site that has only a limited amount of visitors per week, chances are there’s not going to be much data cached anywhere, resulting in expensive PHP calls made to compile all the details you see here on this page.

This is where WP Super Cachesteps in. It creates static HTML versions of the otherwise dynamic pages, which are then served to all users who are not logged in (meaning almost everyone). The benefits are obvious and caveats fairly limited. However, there are plenty of settings needed to configure the cache, so you’ll want to spend a few moments making sure the caching really works before forgetting about the plugin.

Keep in mind that content caching is not the only setting that affects page loading times, so when choosing to include functionality on your site that calls external Javascripts (such as the social link sharing services), it’s important to keep an eye on what their performance impact will be.

Mobile version

I used to think that there were only selected few sites in the world with a real need for providing a mobile optimized version. Useful everyday sites like news portals, TV listings, public transport timetables etc. would certainly have a market for on the road browsing, but who in their right mind would try to access a blog site through their mobile device?

Enter social networks and Twitter apps in your smartphone – suddenly the scenario of a visit from a mobile device becomes much more likely. If you’ve gone through the trouble of making your content easy to share across different networks, chances are that the links may spread to people who are consuming the network content through their mobile phone. Such a big share of tweets consist of short links to web content that running a site which is not accessible on a modern smartphone really doesn’t make much sense. Short attention span + ubiquitous social apps = content must adapt.

The nice thing about hosted WordPress.com is that you get the mobile themes by default, but in your own installation you’ll need to take care of the theme configuration yourself. There are naturally plugins readily available for handling the task of identifying how the user is accessing the content and providing an alternative mobile optimized theme for those who need it. My choice was WordPress Mobile Pack, which offers an attractive set of features for mobile-enabling your blog. Removing Flash, shrinking images, limiting page size and specifying mobile widgets can all be achieved from plugin settings, while the package contains a set of mobile themes courtesy of Forum Nokia.

You’ll want to ensure that also the caching solution is compatible with your mobile site (see settings for WP Super Cache). This was one thing I struggled with while testing the sites on my mobile devices. Speaking of which, content testing in the mobile environment can be a much bigger pain than your usual multi-browser setup on a PC. Proper emulators tend to be either parts of an SDK or then commercial solutions aimed at professional developers. You can try out a few sites like the Opera Mini simulator, or face the really inconvenient truth and have mobiReadyshow you how bad the site will look like on a Motorola Razr, how long it will take to load on a GPRS connection and what the end user data cost will approximately be. All I can say is: yuck! Thank god those kind of handsets are gradually becoming a part of our mobile history.

With the explosion of sites and services where many of us register with our real name and create a real profile, the number of potential hits to be found is growing. This makes it all more interesting to see how Google ranks different sites that publish your name, because they are very likely to be not something you would have expected. Anyway, that’s just a side track of the topic I’m trying to get at, so let’s see what the results for a search on Jukka Niiranen looked like today on Google.com:

The first hit goes to a namesake of mine. Nothing unexpected there, as there’s plenty of us. My personal domain niiranen.eu comes in on places 3 and 4, after that we move to the usual LinkedIn and Facebook profiles. But wait: what the —- is Bantam Live, and why does Google think it’s the second most likely page that a person would be looking for when searching my name?

Ok, here’s the story. Bantam Live is a hosted CRM service, which is nothing like the traditional Enterprise Software that CRM used to be like in the 90′s and early 00′s. It’s in fact more like the social network services that people like you and me are actively using when not in the office. I tried the service, loved the concept and decided to show some respect by writing a tweet about it.

So what, there’s a million tweets every minute, right? Well, it just so happens that smart companies nowadays are listening to the conversation that takes place in Twitter. So was John Rourke, the CEO of Bantam Networks. Not only did he retweet it, but the company decided to quote the tweet on their own site’s front page. Nothing much I can or should do about that, since it’s a piece of text I have intentionally broadcasted to the world to see.

The end result is that now Google thinks I have something to do with the service. But is that actually such a big mistake at all? Isn’t it rather a lucky shot that the search engine has managed to surface such a connection, instead of blindly following hyperlinks and counting words? Sure, there’s absolutely nothing dynamic about the tweet text ending up on the page in this case, but it gives a glimpse of what could be possible with the data that’s already out there in the social network sites. Without having to build a proper semantic web to replace the web we have right now, the mere association of Twitter identities to people and companies could already have a huge impact on how the search engines might begin to understand the online content in a whole new way.

The real revolution of course is already taking place. Whether you call it Social CRM or something else, the fact that the customers now have a voice to speak with is creating a big demand for solutions and strategies that allow the companies to listen to the conversation that is taking place out there, and most importantly to be able to react to it. It might be a topic that I should rather write about in my CRM blog, but it’s becoming increasingly difficult to draw the line between fun web apps and serious business platforms. Which is exactly the whole point of the revolution.

I felt like such a stalker, but was I really stalking on the poor guy? That is a question I was left pondering as we parted our ways and I moved on to the next list of tweets. Unlike in Facebook and some other networks that are repeatedly making headlines for alleged privacy violations, Twitter truly represents the raw power of untamed social networking applications. There is no privacy, period. The name of the game is in the public broadcasting of your thoughts to an unspecified audience. You don’t need to worry about the concept of a “friend”, as there are no friends in Twitter. You can of course follow other users, but this doesn’t have any impact on what they can see and know about you. It’s all out there and that’s why we love it. That’s what makes it the ultimate sharing platform.

Let’s take a look into the future for a minute. Layar is a great mobile app for demonstarting the concept of augmented reality. How it works is you launch the AR browser in your mobile phone, point the camera to any direction and Layar will start to append the image with location based information. The usual stuff like restaurants and points of interest are of course available, but you can also view things like geotagged tweets. With the kind of devices we are carrying around in our pockets, it is not at all far fetched to envision a time when you can pull up an augmented reality browser that shows you not just the buildings around you but the names of the people. Think of avatars and @username’s floating on top of the commuters in the traffic jams. The ultimate nude scanner for your mobile?

In the tech or media industry, or any knowledge work intensive line of business, it can no longer be considered bizarre behaviour to be constantly revealing yourself to the world through various social media sites and services. It is rather becoming the norm of what is expected. You better be active on Twitter and Foursquare, otherwise there’s a risk of people thinking you don’t “get it”. Ok, I’m perfectly fine with that trend, with my active sign-up policy to new and exciting web apps.

It is only when the virtual world meets the physical world that things can start to feel ackward. When you meet a familiar avatar in flesh and blood, there cab be a sudden sensation of “OMG, I know too much about you, yet you don’t know anything about me”. The unilateral nature of the relationship can play tricks with your head. People you’ve never met but who you’ve followed through Twitter can start to feel like pseudo celebrities, even though they are likely to be far more average Joe’s in reality than you are, with nothing better to do than posting stuff online 24/7.

I think we’re reaching the uncanny valley of social networks. This concept was originally introduced for describing how in the field of robotics there is a point in which the machines can begin to look too human, thus causing a natural feeling of revulsion in us human beings. In the world of social networks, this same sensation may be achieved by simply knowing too much about the stranger standing next to you. Something that is perfectly cool when sitting in front of your monitor at home can suddenly feel just plain “wrong” when meeting face-to-face. Sharing your life and thoughts is great, but just don’t do it when I’m around. God, us human beings can be such weird creatures at times.

Is there going to be a moment when we simply get enough of revealing ourselves to others? Will the new sociality trend reach its peak and make way for the ultimate privacy backlash, where people simply refuse to give out any personal details to any online service? I’d say that’s an unlikely scenario. Certainly we’ll need to go through the emotions and find the right balance, time and time again, but eventually we’ll have to make it across the valley. With social media and robots alike.

Now you have the chance to take a trip down memory lane and revisit the Internet from the 90′s, courtesy of the Geocities-izer by Wonder-Tonic. For example, you can remove all the Web 2.0 nonsense from this blog of mine by viewing it through Geocities-izer instead:

Ah, the good old times. More great examples available here.

I used to be tracking my sites with a free account from Mon.itor.Us for over a year and was quite satisfied with getting a weekly report delivered to my inbox, with the average uptime and response time stats for the week. However, trying to do anything within the service’s web interface, like examining the performance statistics or configuring monitors was just really painful, thanks to the poor usability of the service UI.

Enter Pingdom. Wow! Talk about the difference that a good UI design can make to the same underlying service, which at the end of the day consists of pinging different URL’s. It’s never  just about how you collect and store the data, it’s the presentation layer that really counts. Pingdom does an awesome job on this front, delivering a Google Analytics type of a user experience. They offer a free account for one single monitor, so trying out the features is a breeze. 5 checks is already $9.95/month, but if you’re actually making money with your sites, then I’m sure you could make worst investments.

So how’s the uptime then? Here you go:

Uptime report, last 30 days

Response time report, last 30 days

In addition to the widgets above, you can also access the monthly summaries over here. The reports inside Pingdom are much more detailed, one particularly nice feature being the possibility of filtering data based on different tracking locations, ranging from Stockholm to Los Angeles.

I’ve got a dirt cheap hosting package from Suncomet, which gives 2 GB of disk space with unlimited traffic & DB’s for just €30 per year. It’s not a lightning fast service and the P2 server does have it’s share of downtime, so every now and then I think about if I should bother to switch to another service provider. With a monitoring service like Pingdom I won’t need to make guesses on what I’m getting for my next-to-nothing investment. As for knowing the performance of other potential providers, well, that’s still going to be to some extent guesswork and leaps of faith. At least until we get more universal hosting service monitors like HostPeek.

Facebook has such bad usability all-around that nowadays I tend to only navigate to Bejeweled Blitz and follow status updates throug TweetDeck. When do I then go to Facebook? When I get an email from the service, telling me that someone has commented my stuff or sent me a message. Rumor has it that Facebook is in fact working on developing a full webmail service, where you could receive messages to your vanityurl@facebook.com.

Google is the current king of email with Gmail. Nothing comes close, except the threat of people’s increasing usage of media other than traditional email for their messaging needs. Google isn’t standing still, instead they are trying to incorporate more and more social features into Gmail, like the recent announcement of Google Buzz. And where does the Buzz exist in terms of UI? In Gmail. Where do the comments to your status updates come to? Your inbox. ¡Viva la email revolución!

Google Wave is trying to go beyond email, but the current preview version (notice: not even beta!) has one severe limitation: it doesn’t act like webmail, meaning you can’t actually send emails to your @wave.com address. Yeah, what do you call and @address that’s not an email address? It’s hard to see the adoption rate picking up until Wave embraces email.

Numerous Enterprise 2.0 application providers are keeping themselves busy by building wonderful collaboration environments for office workers, to make them more productive in their daily tasks and teamwork. But still they can’t come anywhere near email. Everyone uses it and it is the lowest common denominator that every information worker loves to hate, but couldn’t live without. As Jacob Uckelson writes about the enterprise collaboration paradox:

So even though almost every enterprise has special purpose solutions available for collaboration and process management, good old e-mail always ends up being the primary method for both collaboration and processes in the enterprise. This can be called the “enterprise collaboration and process paradox,” and is the “dirty little secret” of both collaboration and process execution in the enterprise. Realistically, there doesn’t seem to be any way to displace e-mail as the king of collaboration and processes

Everybody used to be stressed about the growing amount of spam email a few years back. Today the rate of spam that avoids my junk email filters is probably 0,001%. Email used to be confined into desktop applications like Eudora or the omnipresent Outlook. Now it’s in the browser and in your mobile phone, meaning in practice everywhere.

Email has come a long way and it’s not going away anytime soon. It’s more likely that we’ll just be getting more of it.

Recently I haven’t been touching iTunes much at all, since my online music consumpion has transformed from files to streaming, thanks to Spotify. Many times there are still tracks that are not available through the Spotify subscription service, which is when I turn to see if iTunes is providing them available for purchase. Much to my surprise, I’ve started to increasingly dislike the iTunes experience. Looking at the results from the new 2010 Customer Experience Index by Forrester, it looks like I’m not the only one.

Let’s compare what happens when I search for an artists in both services. iTunes goes first:

With the current version of iTunes 9, there’s just too much going on in the user interface. Even with the UI chrome excluded from the picture, there’s still way too many items fighting over my attention. Furthermore, I don’t get the immediate Google Experience ™, where the first search result is big and bold, waiting to be clicked. In iTunes the items are too small and don’t appear clickable, maybe because there’s just so much I could click on. Overall, the navigation just feels like too much work, not enough fun.

What about in Spotify then:

Lots of small text here as well, but it’s a lot more bearable, since the structure is so clear. I know the purpose of each UI element and they are grouped in a logical manner, so that none of the information I’m not interested in feels like any kind of distraction. Even without any previous experience of using the application, I think I would be quite at home in navigating in Spotify, since it follows the univesal language of audio library software. What’s funny is that I think much of that language has evolved from the most popular audio library out there: iTunes.

Why is iTunes starting to become more cluttered? Why is it going down the path of Altavista, which used to be the top search engine of its time, before transforming into a messy portal with too much ads and features going on, then consequently losign the game to Google? My theory is that both iTunes and Altavista have (or had, in the case of AV) the same problem, which is the need to be constantly selling to the user. iTunes Store does not make any money until the user clicks “buy”, and it needs to achieve this same behaviour time and time again. In a similar fashion, the portal fever that Altavista was infected with consisted of presenting as many ad banners to the visitor as possible, whereas Google cleared away the clutter and developed a way to show only relevant text ads in predetermined sections.

The Spotify model doesn’t have the hard sell built into it. Its model lures in new users with free accounts, to explore the simple functionaliy built over the vast library of music available for streaming by just clicking on it, building up an engaging first experience of the service. Sure, the users are greeted with audio ads between tracks, unless they purchase the premium account. However, that’s a handicap that you can disable with giving them money, and I’m betting people can easily understand the trade-off there. In iTunes Store there is no subscription service that would make the catalog browsing free from the hard sell. Not until Apple starts offering the similar streaming scheme as Spotify, that is.

Let’s take one final look at another UI for searching a particular artist. With the introduction of Google Music giving some glimpse of things to come, I’m betting that the future UI design patterns for audio libraries will most likely be coming from the masters of search.

The issue

Today I got a promotional email from Amazon.co.uk, saying that you now deliver thousands of additional products to Finland.

I reviewed the delivery restrictions, saw that you ship watches to Finland, and therefore proceeded with trying to place order for the following item (Diesel DZ1132).

Upon trying to check out, I was granted with the familiar statement of “This item can’t be shipped to your selected destination”. Apparently it was an Amazon Marketplace item, where different delivery restrictions apply.

The suggestions

Since you keep a record of what my account’s delivery destination is, I would like to have the option of filtering out products that I cannot purchase. At the moment, there is no way for me to search for watches that are sold by Amazon.co.uk itself and which could be delivered to my country of residence. The logical result is that I must consider all watches to be not available for me to purchase, as there’s no way for me to validate this outside the order placing process.

If you ever implement such a feature, be sure to remember to promote that in your newsletters as well. Previously I requested the option of not receiving the electronics related email advertisements from you, as the products promoted in them were not available for me to purchase. At the same time, I did not want to give up your book related recommendations emails, so opting out from all email communications did not seem sensible to me. I was told then that no such option unfortunately existed.

The final positive words

Now when you have increased the coverage of your delivery destinations, we’re “almost there but not quite”. Since the shopping experience at Amazon is in general superior compared to other online retailers, I’m looking forward to having the last few detailes sorted out one day, to truly deliver a shop that works the way the customer would expect it to work.

Best regards,
Jukka

After a few days, I got a reply from Amazon.co.uk customer service. Nothing out of the ordinary there, typical corporate lingo.

The reply

…I do understand your concern regarding the Marketplace delivery restrictions. Although we are aware that all sellers will not be able to ship all the items to international destinations, we are always working to serve our international customers better.

Please know that all our customers, international and at home, are very important to us. Also it is always important for us to hear how customers react to all aspects of shopping at Amazon.co.uk. Strong customer feedback like yours helps us continue to improve the selection and service we provide.

I have forwarded your message on to the appropriate department who will be glad to review your suggestion.

The surprise

Today when I started writing this blog post, I visited Amazon.co.uk to recreate the bad shopping basket experience, in order to get a screenshot of it. Surprisingly, there was no such experience, as the order went through! Apparently many of the Diesel watches have now been moved to the category “dispatched from and sold by Amazon.co.uk”, which means I will get my Xmas present after all.

The lesson

Negative customer experiences, when offset with a positive surprise, can turn into moments where the customer relationship is actually solidified. For a more theoretical reference, see this post by Frank Capek on the importance of the “element of surprise”.

Months went by and along came Spotify. Instead of being free first, then going paid, there was always a solid freemium pricing model set up for the service. After listening to the ads on Spotify for a couple of weeks, I was more than happy to pay €10 per month for the no-ads, higher quality service. Lesson number 1: never ask the users to start paying for something that used to be free for them. Paying must give you more, not just more of the same.

Spotify does a fairly good job in the basic music player functionality, but otherwise the service is shallow. The recommendations and new additions provide a very rudimentary interface to explore the content of the Spotify database. This means that most of the time you will need to come up with the artist or track name you want to search for in some other context, then try your luck on possibly finding the right music on Spotify. You can’t collect a usable library of the albums you have listened to or want to in the future (which is what I use Tumblr for as a temporary solution). Furthermore, there’s no community aspect to the service whatsoever, which in the year 2009 cannot be considered just a nice-to-have feature anymore.

Re-enter Last.fm. I didn’t initially pay any attention to the Spotify settings menu, since there are not too many things to configure in the player anyway. Nevertheless, it turns out there’s been a native Last.fm scrobbling feature built into the Spotify client for almost a year already. So, just by entering the account information, my up-to-date playlists were back  in Last.fm. With the addition of this nice little Greasemonkey script, launching the Last.fm tracks right from Firefox is also a breeze. Even more apps/mashups can be found from here.

Last.fm’s CFO has stated that they don’t view Spotify as a rival but rather a collaborator. While that might sound like the typical corporate talk where everyone’s always a leading provider in their own little niche market as defined by themselves, there’s a lot more wisdom behind this particular statement. Lesson 2: Don’t reinvent the wheel, unless you plan on learning more about wheels. (Ok, I ripped that from Jeff Atwood, but I’m just practicing what he’s preaching.) Sure it would be perfectly possible for Spotify to build their own community features into their client, but why would they bother? It’s all there already, just integrate the two services and let them both focus on their strengths (and come up with a fair revenue sharing model, which is easier said than done). I’ll be a happy user of them both.