Sincerity: The Currency You Can’t Counterfeit

“freedom from deceit, hypocrisy, or duplicity; probity in intention or in communicating; earnestness.” (sincerity definition from dictionary.com)

There’s a reason why many people prefer the earlier U2 records. Or why some don’t like Coldplay but listen to bands that sound similar. Or why Radiohead and The Flaming Lips continue to have a loyal following after two decades of risky – and sometimes flawed – experiments in the music business.

It’s often misattributed to snobbishness, but for many people taste – in music, film, art, fashion – is strongly influenced by our perception of the artists behind the art, and in particular their sincerity.

Art – including music, film, and design – is something that we buy in to emotionally. We don’t want to feel that the artist is taking us for a ride, like a smarmy confidence trickster in a dive bar picking up one-night stands. We need to believe in the honesty and sincerity of their work if we’re going to invest our deepest emotions into them. 

How can we trust Bono’s evolution from edgy rocker to hypocritical-enviro-loudmouth? Chris Martin’s musical whining in the daytime and Hollywood dining in the evening? Adam Duritz, who sang so beautifully of loneliness and heartbreak on the first Counting Crows album, couldn’t be taken seriously again with the same topics after serial-dating half the cast of Friends. 

Sincerity applies to web design, too.

The original introduction and adoption of the Web 2.0 design style – friendly, rounded, colourful, shiny – could be seen as a reaction to the insincere corporate styling that came before it. 

Most developers and small companies creating web applications weren’t mega-conglomerates with stock-photo-model employees: they were individuals with personality who enjoyed making things for the web. The Web 2.0 style could be seen as a reflection of putting this personality online, in a more honest representation.

The trouble is that as Web 2.0 applications became popular, the style was copied by a growing number of organisations that didn’t exhibit the personality traits implied by the simple, friendly interface. As these grew in number, so the Web 2.0 style became untrustworthy, or at least as ‘corporate’ as the style before it.

This is my argument for avoiding the Web 2.0 style for new apps you may be developing. Even if it does represent your personality accurately, stay away from it, because – like a brilliant musician who has been auto-tuned and produced by Timbaland – it’s difficult for the consumer to tell the difference.

Reputation: Context and Control

“the estimation in which a person or thing is held, esp. by the community or the public generally” (reputation definition from dictionary.com)

Given the democratic, equal opportunity publishing platform that is the web, it was inevitable that we’d see a shift online from authority (assigned by State, wealth or legacy) to reputation (earned through actions, output and behaviour).

What is more surprising is that we still don’t have anything near a final model or system for measuring or storing this important new metric. My personal reputation on the services I use – LinkedIn, Facebook, Twitter, Disqus, Hacker News, my blogs, others’ blogs, Flickr, and so on – are totally separate and invisible to one another. Every time I join a debate on a new site or service, I need to re-establish my reputation.

Worse still, some websites control the reputation of their users, when it is clearly in their best interest to do so. Companies like eBay or Amazon have an obvious self-interest in ensuring that their third-party sellers obtain and retain a high reputation (or ‘rating’). Although I had previously assumed that it was in their best interest to assign and display an honest rating, recent experiences have led me to believe that this is not necessarily always the case.

If there is a need to measure and display reputation, and this cannot be trusted to each self-interested website, we are left with two options:

1. A centralised Reputation Service
2. A decentralised Reputation Service

A current close contender for the first ‘centralised’ option is Disqus, where comments on disparate websites use the same profile. This could be taken a step further: perhaps Disqus could also be configured to access and store your reputation from other services.

For example, by using a Google Webmaster Tools style ‘claim token’ together with authenticated API access, Disqus might allow you to ‘claim’ your blogs, RSS feeds, analytics data, social connections, and other metrics that contribute towards your centralized reputation measurement. You could then direct people to your personal ‘reputation dashboard’, or it could even be included (via a Javascript widget) on third-party websites.

The Whuffie Bank almost replicates this model, but has nowhere near the sophistication or reach to fulfill the potential. In my opinion, Disqus is currently in a better position to achieve it.

For the second, ‘decentralised’ option, the closest we currently have is the PageRank model, where explicit connections between websites are used to imply a ‘trust’ for any website. Although this could be thought of as a ‘centralised’ system – owned by Google – the model itself, and the data it is based on, can be adopted by anyone.

The model needs to be evolved to accommodate not only websites, but also people, and potentially even individual items or sets of data. It also needs to change so that it is not so easy to game: the current PageRank system requires numerous special cases to deal with link-bait, link-farms and other scams for cheating website reputation.

A decentralised trust system has long been a goal (or, rather, a requirement) of the Semantic Web. The theory is that every person can publish a list of who or what he or she trusts. A simple implementation might be that you place an XML file on your website, or onto your homepage, that says something like:

<iTrust>
            <entity>person@organisation.com</entity>
            <entity>another@somewhereelse.com</entity>
            <entity>www.somekindofwebsite.com</entity>
</iTrust>

These ‘trust statements’ can be harvested and – like the PageRank model – used to aggregate an overall trust for anything that you can uniquely identify; in the example above, email addresses are used to identify people.

They could also include all kinds of additional information, such as the context in which you trust each entity:

<iTrust>
            <entity context="finance, law">person@organisation.com</entity>
            <entity context="technology">another@somewhereelse.com</entity>
            <entity context="health">www.somekindofwebsite.com</entity>
</iTrust>

The problem though, as mentioned earlier, is: how do you prevent this system from being gamed? It’s slightly ironic that the democratic nature of the web – that makes it easy to manufacture unlimited websites, email addresses, and online personas – makes reputation important, but also makes it incredibly difficult to securely declare or measure.

For this reason, I currently lean towards the centralised approach, even if it is more risky to rely on a single provider. Perhaps, to earn reputation themselves, the provider could ensure that all data was openly available for backup and re-use.