Seduction and advertising

-- sidebar to Lecture notes on use-oriented web design --


The British architect Denys Lasdun said around 1972:

"Our job is not to give the clients what they want. Instead, we must give them something the never imagined. And when they see it, they should recognize it as what they wanted all along."

Lasdun is clearly not talking about task analysis and user testing. Another interesting angle is the study by Tractinsky (CHI'97 Proceedings, 115-122) where it was shown that a system that people find aesthetically pleasing is also perceived a priori as easier to use. Something other than task efficiency is involved here.

But what are the ways to go beyond web design as usability engineering? We take a brief look at the concept of seductive technology and the theories of advertising as applied to the web.


Seductive technology  

Seductive means to have alluring or tempting qualities. A promise of extraordinary experiences is involved, as well as the qualities required to make good on the promise.

Tetris is clearly a seductive game. People keep playing, developing a relationship that is something other than the simple fulfillment of task-related goals. And it is quite obvious that stunning graphics or beautiful visual design are not involved (at least not in the original version for the 9-inch Mac).

Dynamic queries (Ahlberg et al., CHI'94 Proceedings, 313-317) is another example of a seductive, albeit less well-known approach. The concept of turning boring old database queries inside out and shaping the data with your fingertips is immediately attractive and enticing to many people.

Khaslavsky and Shedroff (Communications of the ACM, 42(5):45-49, 1999) describe seduction as a process of

Enticement: Grabbing attention and making an emotional promise.

Relationship: Making progress with small fulfillments and more promises. This step can continue for a long time.

Fulfillment: Fulfilling the final promises and ending the experience in a memorable and positive way.

The checklist for identifying seductive qualities goes as follows. Ask yourself whether the product under scrutiny:

Entices you by diverting your attention?

Surprises you with something novel?

Goes beyond obvious needs and expectations?

Creates an instinctive emotional response?

Espouses values and connects to personal goals?

Makes inherent promises to fulfill these goals?

Leads you to discover something deeper than you expected?

Fulfills small promises related to your values and aspirations?


Web sites as advertising  

Singh and Dalal (Communications of the ACM 42(8):91-98, 1999) apply basic communication theory from advertising to web design. The main idea is to view the web page, like any other advertising message, as the vehicle for a two-step process.

Inform: Make the user aware that the specific web site exists and provides certain information. create the belief that the web site is 'cool'.

Persuade: Generate positive attitudes, such as favorable impressions about a site and the idea being promoted. Generate positive behaviors, including revisiting a site, bookmarking or printing the page, buying the service promoted at the site, emailing the web master of the site.

They present an experiment where people browsed a number of commercial sites, then classified them along the rational-emotional dimension. They also rated their attitude towards the site and its sponsor. Finally, they indicated the likelihood of further exploring the site.

The results show that emotional-rational classification is very consistent. More importantly, the emotional dimension had a significantly greater impact on the persuasion measures than the rational dimension.

The authors suggest thinking about web design quality in communication terms: Does the message create intended beliefs and persuasive intent toward the message, the promoted idea and the sponsor?

Such notions of quality can be turned into metrics just like usability. (A month after launch, 20% of the target audience should have visited the site and be aware of the sponsor as well as major products and services. At least 50% of visitors should be encouraged to explore at least one more page of the site. And so on.)

Experimental results always reflect the intentions of the experiment. Lohse and Spiller (CHI'98 Proceedings, pp. 211-218) studied 28 online retail stores in August -96 to determine what features best predicted traffic and sales. They found that

Improved product lists and store navigation features by far had the strongest effect on sales.

Large stores attract more traffic but not more sales.

Appetizers, store presentation, number of levels, and number of featured products along the path had no significant effect.

Is this a contradiction? Hardly. Lohse and Spiller are not looking at the emotional variables that concern Singh and Dalal. Both sets of conclusions are relevant in order to design persuasive communication and successful commerce on the web.