The Many Challenges of Widgets

Categories: Challenges, Social Networking, Widget StrategyPosted on February 28th, 2008

Before you invest time and money in a widget strategy, you better know the challenges. This applies to VCs, Web Strategists, Developers, and even Social network companies

An Objective View
In this blog, I strive to provide a balanced viewpoint of both the benefits and challenges of a web strategy, it’s easy for us to become over-hyped and then fall right into the pit of over exuberance. (See other posts tagged Challenge)

I’m moderating quite a few panels with widget developers (last week at Stanford, next week at Graphing Social, and in a few weeks at Ad:Tech) so I’ll be using many of these challenges to hold the vendors to their claims.

First, a few parameters:
Update: This list of challenges mainly applies to widgets within Social Networks, although many of the challenges afflict mobile, desktop and blog widgets.

Widgets are ‘mini-applications’ that can be embedded on other containers (such as social networks like Facebook, MySpace, Bebo, LinkedIn, and whoever decides to join the movement) The thing is, desktop widgets have been around for some time, so this really isn’t anything new, but for the purpose of this post, I’m just going to keep those on social networks in scope.

In regards to terminology, I’m just going to use widgets as a blanket term to also include “applications” and “canvas pages” terms that developers user on Facebook’s F8 platform. I’ll clarify that on another post in the future.

The Many Challenges of Widgets:
Each of the following hurdles can be overcome, but first, let’s identify them.

Difficult to Monetize
First and foremost, this has been the biggest challenge. For some widget developers, the money has come from investors or VCs on Sand Hill road. Secondly, I’m hearing that the CPI plague (see below) is becoming more common, and then lastly, advertising is not an effective way to monetize in social networks (read this for more).

Immature Market
Widget developers are mostly experimental, they are throwing gangly Spaghetti, Pasta, Rigatoni, and Jeremici (I just made that up) against the wall to see what will stick. In most cases, most of the half cooked pasta falls down from low users adoption, leaving a sticky residue of messy profile pages.

Overcrowding profile page
Have you seen my profile page on Facebook? It’s a mess, and with so much noise, who can compete? With there being thousands of widgets, only a few can survive on my profile page (I know there is tabbed segmentation) but really, how many do we need?

Low Barriers to Entry
A challenge for every web market, is that there are few ways to differentiate, and it’s easy for a young Russian developer or a Stanford student, or a team of Chinese engineers to quickly get in the game.

Metrics and Analytics Inconsistencies
Hardly anyone is measuring the success of their widgets in the same way, do we measure on install, activity, views, traffic, or clicks? As a result, other than Appsaholic, there’s very few industry ways to measure success.

Spammy
Sadly, I learned from a panel I managed that some of the most successful apps were the one that leaned on the social graph, no not the one that we all dream about, but in the context of email spam. Many containers are clamping down on this, as it’s best to preserve the user experience, but this could continue to be an issue.

Bastardization: Cost per Installation (CPI)
To me, this looks like one of the worst in our industry, to me, this is like ‘printing money’. Did you know some of the top developer companies sell to other developers the chance to let new widgets piggy back off successful ones by promoting them. The developer can then charge for cost per install.

Disposable and low value
Rodney Rumford first mentioned this term to me, he was describing that many widgets are simply not used more than once. These glamor widgets provide one time entertainment, or are used once and never reused –except for removing from ones profile page.

Recycled clones offer little uniqueness
Perhaps the worst plague is the “attack of the clones” in many cases, the code from widgets are created by one developer, ~~stolen~~ (I mean crowdsourced or collaborated) slightly modified or rebranded and then republished.

Low Utility
I’m trying to think of a widget that provides business utility, or one that improves my life other than casual communications or entertainment. Reminiscent of the web in the mid 90s, we’ve yet to see the business value.

Hard to build successfully
Specialized skill set are common among the developers, most traditional interactive firms, and most companies don’t have the skills or experience to create a widget. It’s a different game with a different mindset, the same strategies often don’t apply.

Multiple APIs strain developers
Most platforms or containers are offering their own API, although most are touting they are OpenSocial compliant (as I write this, OpenSocial is ~~not public~~, it is but in beta but should be soon) yet we’ve got to wonder is it too late for there to be a common industry API if it’s already fragmented? I spoke to the Evangelist of MindKey last night and he suggested that each platform has unique APIs (like news feed APIs) that the other doesn’t share –it’s already fragmented.

Ever changing platform APIs requires attentive team
When I hosted the widget roundtable, it became very clear that the APIs on platforms are quickly changing, and sometimes without notice. For the agile developer company, they’ll be able to quickly morph with their full time resources, but for the Interactive Firm or corporate web team, they’ll likely be too slow.

Pricing may vary, lack of standards
I have the privilege of being one of the few people in the world that gets to talk to the widget developers and find out what they are charging. In many cases they don’t know what to charge (as they have strong technical skills, but are just ramping up on the business side, and they are undercharging) This will clean up this year, and we’ll start to see some benchmarks.

Poor User Experience
Dennis McDonald (via comments) suggests that the usability for widgets is often poor. With there being little standardization on the inteaction design, I’ll agree, he states: “It’s hard enough to keep track of multiple incoming data streams representing different people, sources, relationships, keywords, etc. When you try to cram too much though a widget you have a real usability problem because of the variety.”

Performance Issues
Pravda (via comments) suggests that some widgets have poor performance, and thereby causes a disrupted experience. Since widgets are hosted on third party servers, some laggards can hinder the rest of the social network experience, he states: “This is because they require additional HTTP request, and in some cases, this request delays the rest of the page. This is the reason that I am not using Meebo widget in my blog”

Lack of Brand Control
Len Kendall (via comments) suggests that some brands may be concerned where their sponsored widget may appear. In many traditional advertising deals, Coke will never want it’s advertisement near Pepsi, but with widgets, that’s unavoidable.

Add your own in comments
What other challenges of widgets did I miss? Please leave a comment and credit and link to you.

Caveat: While I’m highlighting the challenges, it doesn’t mean they can’t be overcome, and it most cases, the value is higher than the challenge. I’m just suggesting, we shouldn’t only look at the beautiful side, but set yourself up for success by knowing what you’re in store for.

I could write a solution or a fix for all of these challenges, but that’s a post for another time, or one I’d be happy to answer for clients.

Share This Post:

This entry was posted on Thursday, February 28th, 2008 at 7:02 am and is filed under Challenges, Social Networking, Widget Strategy. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.

How about “usability”?

I’m thinking of the increasing number of widgets that provide access to some mix of multiple feeds from multiple types of sources.

It’s hard enough to keep track of multiple incoming data streams representing different people, sources, relationships, keywords, etc. When you try to cram too much though a widget you have a real usability problem because of the variety.

Posted by Dennis McDonald on February 28th, 2008 at 7:14 am

When looking at blog widgets, I found that most of those widgets slow down page loading time. This is because they require additional HTTP request, and in some cases, this request delays the rest of the page. This is the reason that I am not using Meebo widget in my blog.
This issue increases the challenge, as now widget developers need to integrate better with blogging platforms in order to overcome these problems.

Posted by Kfir Pravda on February 28th, 2008 at 7:18 am

I was very confused, you’re talking about Facebook applications for most of this, not like, standalone, JS or flash-type widgets, right?

Posted by Dan Zarrella on February 28th, 2008 at 7:20 am

Brand Relevancy? Not all brands have a “sexy” tangible product that can easily translate to widgets. EG. Insurance, Certain Packaged Goods, etc.

Posted by Len Kendall on February 28th, 2008 at 7:31 am

Dan

These are mainly social network widgets I’m referring to (although many of these impact other types of widgets). I’ll update the post to reflect that.

Posted by jeremiah_owyang on February 28th, 2008 at 7:32 am

Some really good challenges being listed here. I’ve added a few.

Posted by jeremiah_owyang on February 28th, 2008 at 7:34 am

Great post Jeremiah,
Thanks for helping potential brands and businesses understand the widget/app landscape a little more clearly. There great challenges & opportunities.

But clearly the “build it they will come” mentality days are behind us. A clearly defined strategy will go a long way.

Note to Len @4 “Brand Relevancy? Not all brands have a “sexy” tangible product that can easily translate to widgets. EG. Insurance, Certain Packaged Goods, etc.”

A brand does not need to be sexy to work. here are a few ideas off the top of my head.

Insurance: A widget that brings to the surface connections about accident occurrences, risks, etc. Make it a engaging game where people can see who has the riskiest behavior.

Consumer packaged good: Let’s use Monster Energy drink as an example. You could have an app that challenges people to share extreme behavior and compete for status in a lifestyle behavior type of app.

The opportunities are many. It just takes several hours of brainstorming to flush out the best ideas on a high level.

Cheers!

Rodney Rumford
http://www.facereviews.com

Posted by Rodney Rumford on February 28th, 2008 at 10:12 am

Another point is privacy. Look at Twitter for example. I communicate both business and private issues using this system. So, when I am looking into the option of putting Twitter widget in my blog, I ask myself - should all my blog readers see my tweets? This could have been solved if privacy settings were in place.

Posted by Kfir Pravda on February 28th, 2008 at 10:26 am

Great post!

I don’t see why we have to install facebook apps on our own page (a tedious operation) and then remove it after. Why can’t there just be an apps directory which you can access right away, and then choose which ones you really want on your profile. In a way that’s reducing the disposability factor. I would think psychology plays a huge role. Say you want to take a quiz right. You install the app, take the quiz and then remove it. But if you want to take the quiz again after having implemented some changes in your life to see how you do this time, you probably won’t install the same app again, because you’ll want to be consistent in your behavior.

There should just be one place you can goto and be able to access all these apps. It woulnd’t hurt the developers too much because their app would still be getting used, and you could even have like a sequence of say, quiz apps, that you can take and form an overall opinion. Or you could be part of a gaming network and see your overall ranking with combined stats from all games on that network, sort of like xbox live. And for the important apps like productivity etc, you could just label them as important, because having something on your profile doesn’t mean its more important to you. Plus if you go to someones profile and want to interact with that app, you can’t even do it without having to install it, so it’s like a stupid teaser that just pisses one off. Instead if I came to your profile and saw what apps you use, I could click on it and start using it right away and just enter your name to have that app interact with you, so ‘installs’ aren’t needed. I guess a system of impressions and CPA would need to be devised.

Posted by Azam Khan on February 28th, 2008 at 11:10 am

There are some real utilities being realized if you check out what people are creating and sharing in the forums over at http://www.sproutbuilder.com check the forums out. You can build widgets that contain multiple pages. It is up to you to make the content compelling, captivating, and utilitarian though. http://sproutbuilder.com/forum/share-your-sprout

People are combining yahoo pipes with RSS feeds and mashing up other technologies to create more interactive and dynamic content.

I think http://www.gydget.com has built a container that is the most easy for the end user to create that keeps the widget from becoming disposable. At this point I have so many placements of it and it performs so many tasks from selling tickets, to mp3’s, to linking to our e-store that I am invested in it. It tells me what my clicks are on all the purchase buttons too. It take me very little time to update, it updates live, and it monetizes. If it made the band I work with 1,000 bucks in ticket sales and 1,000 bucks in mp3 sales, it is doing it;s jobs because it cost me nothing to make.

Posted by Adam Gershenabuam on February 28th, 2008 at 11:50 am

Outstanding post.

The key to resolving several of these issues (in my mind) is to provide updated content within the widget via RSS. You wouldn’t need to update everything, just information that is related to the content that spurred the initial “widget grab” by the end user in the first place. This would improve utility, value and (to some extent) end user experience. Widgets that deliver headlines and info related to a specific interest provide continuosly refreshed content relative to the end user experience.

The brand that has its logo on that widgets’ “skin” is going to reap the benefits of sitting in front of the end user for an extended period of time. The key is to ensure that the content doesn’t conflict too much with the brand message.

While there are indeed many problems, widgets are truly in their infancy. On the upside, the potential rewards are astronomical. And with the money and resources being poured into the widget ecosystem, I believe solutions and innovations will move swiftly in this emerging industry, particularly as it is becoming clear that RSS and widgets pose a very real threat to the enormously lucrative Google Adsense advertising model.

Posted by Jughead on February 28th, 2008 at 11:54 am

The key with widgets is their simplicity. Need to be fast loading, intuitive and meaningful usage.

And don’t always expect it to be a mainstay in someones page or world.

Widgets should be simple and disposable.

Posted by Kin Lane on February 28th, 2008 at 12:05 pm

Hi Jeremiah,

Great post. I would add that a big problem for serious applications is getting seen above the noise of the spammy applications you talk about.

In the long term, applications must engage and give value, and spread by real word of mouth, not just spammy ‘invite your friends to get a badge’ “viral” techniques.

Luckily, platforms like Facebook are recognising this and cracking down very heavily on the spammy applications, which will make a real difference. At the moment however the vast majority of invites that are sent have no real thought gone into them, and a lot are simply ignored - so much so that Facebook has had to allow users to mass-ignore invites. This is a real problem for good applications that aren’t spammy, as the relevant invites and notifications get lost in the irrelevant.

It will be interesting to see how the application landscape of Facebook changes over the next six months or so, and also whether the other platforms, like MySpace, take a similar hard line.

Josh

Posted by Joshua March on February 28th, 2008 at 1:14 pm

Still can’t understand what is widget:(

Posted by Nina.Gan on February 28th, 2008 at 5:01 pm

Nina

No worries. Are you on Facebook? Go to people’s profiles and you’ll see widgets like “X me, or places I’ve been, or vampires” or other mini-applications.

Posted by jeremiah_owyang on February 28th, 2008 at 5:40 pm

HI，Jeremiah:
Thanks for your help.
Now, I get it.

Posted by Nina.Gan on February 28th, 2008 at 6:17 pm

[...] can be installed on a web page, and run independent of that web page. Jeremiah Owyang writes about The Many Challenges of Widgets, while GigaOm believes that Companies Can Money with [...]

Posted by Slide, RockYou = Live By The Hit, Die By The Hit « I’m Not Actually a Geek on February 29th, 2008 at 12:02 am

[...] challenges of widgets” — Forrester analyst and new media marketer Jeremiah Owyang posts a laundry list of problems facing widget-makers, from poor metrics to rule-changes on developer platforms, to weak [...]

Posted by Roundup: IPhone SKD rumors, Bischof at Meritech, and more » VentureBeat on February 29th, 2008 at 2:59 pm

[...] Owyang writes about the many challenges of a widget strategy. He lists about 18 different problems with using widgets to market. Among them: high costs to [...]

Posted by Give Widgets a Chance : Selling to Small Businesses on March 3rd, 2008 at 8:25 pm

[...] Jeremiah outlines the challenges of widgets as a whole and helps bring more definition to the category. I [...]

Posted by Marketing.fm » Welcome to the Widget Wars on April 2nd, 2008 at 5:03 am