Web 2.0 Can Be Dangerous...

by Jakob Nielsen on December 17, 2007

Summary: AJAX, rich Internet UIs, mashups, communities, and user-generated content often add more complexity than they're worth. They also divert design resources and prove (once again) that what's hyped is rarely what's most profitable.

... dangerous for your profits, that is. If you focus on over-hyped technology developments, you risk diverting resources from the high-ROI design issues that really matter to your users — and to your profits.

Unlike some older technologies (notably, Flash and PDF ), Web 2.0 ideas are not inherently bad for users. They can be highly effective; we sometimes see examples of usability-enhancing Web 2.0 designs in our studies. But it's more common to find Web 2.0 ideas that either hurt users or simply don't matter to users' core needs. While the latter case might seem innocent, irrelevant website "enhancements" diminish profits because they indicate a failure to focus on those simpler design issues that actually increase sales and leads.

While there's no single definition of the much-abused "Web 2.0" term, I'll look at four trends that are often considered its defining elements:

  • "Rich" Internet Applications (RIA)
  • Community features, social networks, and user-generated content
  • Mashups (using other sites' services as a development platform)
  • Advertising as the main or only business model

AJAX and "Rich" Internet UI: Too Much Complexity

There's no doubt that the pageview model of interaction provides a scaled-back UI. But this also means that it's a simple UI. When all they can do is click a link to get a new page, users know how to operate the UI . People are in control of their own user experience and thus focus on your content.

"Rich" Internet UIs highlight the more flexible GUI design options that we've enjoyed in personal computing since 1984. Such interfaces can work well, especially for actual applications that offer true functionality and thus require a full GUI. But if you're just designing a website , the more advanced UIs often confuse users more than they help. Why? Because users engage less with websites than with apps. (And many applications are ephemeral apps that also have low user engagement.)

Take the most famous example of rich UI: AJAX, which lets designers update part of a page, rather than taking users to an entirely new page. Because less data download is required, these smaller updates are typically faster, decreasing response times.

Only a fool would deny the importance of response time and download speeds for the Web user experience. After all, we've known since 1968 that speedy interfaces feel better and support flow.

So yes, faster is better. But only if users continue to understand what's happening. We recently tested about 100 e-commerce sites and found many problems with AJAX shopping carts . In particular, users often overlooked modest changes , such as when they added something to the cart and it updated only a small area in a corner of the screen.

It's deadly for e-commerce sites when users can't operate the shopping cart, so it's usually best to stick to simple shopping-cart designs that everybody understands.

To get the required response times, spend your money on bigger servers and better hosting providers. And stick fewer gadgets on each page : these days, slow response times are often caused by too many complex, dynamic design elements that eat up server time.

As illustrated in a sidebar, an AJAX feature can work well on a website. And our testing did find one usable AJAX shopping cart. As always, the real question is not technology, but usability. If you use technology right, it can help sales. Still, the risk is typically too high with new technology because best practices haven't jelled yet. You can't just emulate designs you see around the Web — they're likely to be bad because they were hacked together by geeks drunk on the newest and coolest tech. And, sadly, "newest and coolest" usually translates into "untried and unusable" — and thus money-losing.

Community and User-Generated Content: Too Few Users

User-generated content can be a great supplement to your own content. The most famous example is Amazon's book reviews, which date from 1996 (not exactly "2.0."). Communities, which were the main recommendation in the 1997 book Net.Gain , are also an old idea.

Community features are particularly useful on intranets , and many of the Intranet Design Annual winners offer them. The reasons communities work better on intranets also explains why they're often less useful on the open Internet:

  • A company's employees are an actual community with a crucial shared interest : succeeding in business.
  • Employees are pre-vetted: they've been hired and thus presumably have a minimum quality level . In contrast, on the Web, most people are bozos and not worth listening to.
  • Although some intranet communities — such as those around internal classified ads — are aimed at lightening up the workplace, most intranet communities are tightly focused on company projects. Discussions stay on topic rather than wandering all over the map.
  • Intranet users are accountable for their postings and care about their reputation among colleagues and bosses. As a result, postings aim to be productive instead of destructive or flaming.
  • Small groups of people who know each other are less susceptible to social loafing , so more users contribute to intranet community features. In contrast, Internet communities suffer from participation inequality , where most users never contribute and the most active 1% of people dominate the discussions.

Realistically, most business tasks are too boring to support community features. The fact that the city Sanitation Department will pick up Christmas trees sometime after December 25 isn't likely to inspire a longing to discuss shared experiences on the department's site. Users will visit the site to find the pick-up dates and rules. Nonetheless, the Christmas tree pick-up page is an example of how government websites can offer taxpayers great ROI : if done right, this one page will save the city from answering endless phone calls — each costing $10 or more. Often, such boring, workhorse stuff is where the money is.

Mashups: Co-Branded Confusion

One of the defining ideas of "Web as platform" is that it lets developers merge the features of different sites into a single service. If you're a business, doing this is dangerous for two reasons:

  • Co-branding confuses users, who find it much easier to understand the simpler model of one site = one company. Users are confused when other companies sell on Amazon. Similarly, our studies of investor-relations sites found that individual investors were confused when a corporate site's IR area linked to a third-party site for quarterly reports and the like.
  • Having part of your site effectively under another company's control means that you're at that company's mercy if they decide to change the terms of service. For example, the outside provider might decide to throw in advertisements from your competitors. Not material you want to promote to your hot prospects.

Finally, a "mashed" service will never have usability as good as one that's designed specifically for your needs . In testing store finders and other locators , we found that the most usable maps were custom drawn to highlight a specific store and included surrounding landmarks, recommended parking facilities, and information on how to best access the location using public transportation. A general mapping service doesn't know your customers' needs and thus can't draw the map that will bring in the most customers.

Of course, if you're a small company, borrowing features from an outside service can help you add functionality to your site. But if you're big, the profits from an optimal user experience usually beat the cost of its creation.

Advertising-Funded Business Models: Bubble 2.0

The number of companies that chase the same advertising dollars as their only business model is a sure sign that we're at the peak of Bubble 2.0. It would be much more sustainable if companies aimed to create services that users valued enough to pay for.

Right now, considerable advertising money is sloshing through the Web because most marketing managers remain clueless about how it works. They think that because search advertisements generate lots of business, other Web ads must work just as well. What a fallacy — brought on by ignorance of the basic Web user experience. People go to search engines when they're explicitly looking for a place to do business. This is why search engines profit from sucking up the work of content sites (where users exhibit strong banner blindness ).

Marketing managers won't remain clueless forever. Sooner or later they'll discover that Web advertising offers almost no ROI. Only two forms of Web ads actually work: search ads and classified ads (such as eBay and real estate listings). A third type of Internet advertising that might work are video ads, because video is a linear media form (in contrast to nonlinear website navigation). At this point, we don't have enough user research about Internet video to say for sure.

Hyped Websites: Unrepresentative for Business Sites

In 1997, I said that atypical examples are poor indicators of what you should do with your site. In 2006, I said this again, using examples that were widely hyped that year. Sadly, many Internet managers continue to make this mistake and ask their team to emulate approaches highlighted in the news stories they read. So, I'll say it for the third time:

  • If you're a mainstream business site (including government and non-profit sites), your user experience needs are very different than those of the few hot sites that attract all the attention.

As an example, a smaller e-commerce site should not emulate Amazon.com's design . While they do many things right, there are also many ways in which Amazon deviates from the mainstream guidelines for e-commerce user experience . They can (and probably should ) deviate because of their unique position (which you don't have).

By definition, any website that gets extensive press coverage is unrepresentative for the vast majority of sites. The media covers only "exciting" stories — not everyday business.

The most-hyped site right now, Facebook, is the "Iron Chef" of the Internet . The Iron Chef competition makes for great TV, but has nothing to do with running a restaurant as a successful business. After all, chefs aren't typically assigned a "mystery ingredient" shortly before dinnertime that they must feature throughout a multi-course meal. Broccoli ice cream? Not if you want to make money and bring the diners back.

Like Iron Chef, Facebook has much drama that makes for good press coverage, but most of its features are worthless for a B2B site that, say, is trying to sell forklift trucks to 50-year-old warehouse managers. Instead of adding Facebook-like features that let users "bite" other users and turn them into zombies, the B2B site would get more sales by offering clear prices, good product photos, detailed specs, convincing whitepapers, an easily navigable information architecture , and an email newsletter .

Adapt a Few Web 2.0 Features, but Focus on Core Services

As I've noted, there are many reasons not to get caught up in the Web 2.0 hype. Still, Web 2.0 has some good ideas that can benefit mainstream websites. The problem is that different sites need different subsets of the Web 2.0 features.

As an extremely rough guideline, here's the percentage of Web 2.0 infusion that might benefit different types of user experience:

  • Informational/Marketing website (whether corporate, government, or non-profit): 10%
  • E-commerce site: 20%
  • Media site: 30%
  • Intranets: 40%
  • Applications: 50%

Applications score so high because users perform actions repeatedly and thus truly benefit from rich UI possibilities. In contrast, mainstream websites have very few repeated actions that justify the added complexity of a full GUI's shortcuts.

For website usability, the problem is not whether a specific operation takes 1 second or 10 seconds; people typically perform each operation only once or twice. The problem for websites is the 5–10 minutes users lose when they do something wrong because the site is too complicated. (After such an experience, they usually leave — and you lose the business.) Simplicity is more important than efficiency for done-once actions.

Take drag-and-drop : when used right, it can expedite application interactions. But on a corporate site, the lack of perceived affordances can cause users to overlook important options. (And, of course, when used wrong , fancy interaction techniques also doom applications, which is why it's important to follow application usability guidelines if that's your game. The fact that a technique is "2.0" is more likely to harm than help your users if implemented with poor usability.)

The bottom line? While a modest 2.0 infusion can be beneficial, advanced features are rarely the most important contributor to good user experience or profitable websites. If you get caught up in the hype, you divert attention and resources from the simpler things that really matter. This opportunity cost is the real reason to take it easy on Web 2.0.

Before throwing spending money at "2.0" features, make sure that you have all the "1.0" requirements working to perfection. Of the 149,784,002 sites on the Web, maybe a handful can make this claim. Most sites don't even use the customers' terminology in headlines and page titles — if you want one quick action item to improve site profitability through better SEO ranking, more clickthroughs, and better understanding of your services, rewriting the first two words of your microcontent will beat any technology any day.

Share this article: Twitter | LinkedIn | Google+ | Email