On July 31 2010, Twitter passed the number of 20 billion tweets sent since the service was created in 2006. This is quite remarkable, considering that the threshold of 10 billion tweets had only been reached 5 months before, in March 2010. Currently, Twitter states that there are about 750 tweets sent per second and 65 million sent per day. From these figures it is easy to see that Twitter is even more quickly becoming a force to be reckoned with than we anticipated in our podcast on the business uses of Twitter.

Among the uses we identified was employing Twitter in PR and product announcements. The goal must be to get your message across to as many people as possible. So, having many followers is good, having many more followers is even better, meaning the more followers your Twitter-Account has, the more return on your announcements you will see. But this is only half of the truth. Even more important is the question whether your followers re-tweet, i.e. redistribute, your tweets – and in doing so spread your announcement further and attract new followers to your account. This is exactly what HP Labs Research tried to analyse in a recent study called “Influence and Passivity in Social Media”.

HP-Researchers determined that the average Twitter-user only re-tweets one out of 318 tweets he receives. However, this average does not tell the whole story, because the vast majority of users almost never redistribute messages, while a select few are very active in doing so. Now, seeing that in order to make your tweets as widespread as possible it is important to get your followers to re-tweet them, you could just measure the total amount of re-tweets you get. But if you consider the difference in activity across users, as mentioned before, such a total would be biased depending on which users follow you and their general likeliness to re-tweet messages they receive. HP’s study tries to amend this by introducing their IP-algorithm, a way to assign relative influence and passivity scores to every user. In this model, influence depends on the quantity and quality of the audience a user influences, and passivity is a measure of how difficult it is for other users to influence him. In short, HP-Researchers try to determine the degree to which a Twitter-user can get his followers to re-tweet his tweets and visit the URLs he links in those tweets – given that most users are passive by nature and not easily motivated to re-distribute or visit URLs they receive in the first place. In that sense, the attention a user gets from normally passive followers is even more valuable than that of generally active ones.

After analysing 22 million tweets with this method, HP Labs Research determined that the Twitter-account “Mashable” is the most influential one. This, as you probably know, is the account of Pete Cashmore, CEO and founder of Mashable.com, currently rated second on Technorati.com’s Top 100 blogs worldwide. Pete founded his blog in 2005 at the age of 19 and has since risen to “must-read-status” on all topics concerning technology and social media in particular.

What is his secret then? It is very good content. In a world of social communications, wisdom of the crowd and the long tail it is not enough to simply have good content. Aside from being interesting to readers, very good content not only sparks the interest of people but is also wrapped in a form that stimulates reflection and comments on the topic – and motivates readers to tell their friends about it. As a business user, you need to keep this in mind. It is not enough to send out PR and marketing material clearly identifiable as such. Instead, you need to try to talk to your customers on a personal level, engage them in an open conversation. In so doing, you will not only develop a favourable reputation with customers but will achieve referrals, too, bringing your customer’s contacts and their contacts’ contacts into the conversation. Making the information you distribute viral, as the term goes. Aside from referrals, reputation gains and ultimately ROI, there is another use in engaging your customers (and your partners and employees, for that matter): In the true spirit of crowd sourcing, it could very well be that you will be able to gain additional insight into the mind of your stakeholders, harness their knowledge and experience and ultimately develop better services and products for your customers – all based on talking to them as equals.

The classic Everett Rogers graph of product diffusion, including ideas such as innovators and early adopters, is well known even the world over, and the terms used can be found in every magazine article about new market trends. But is it right?

Rogers stated that adopters of any new innovation or idea could be categorized as innovators (2.5%), early adopters (13.5%), early majority (34%), late majority (34%) and laggards (16%), based on a bell curve. These ideas about diffusion of innovation are among the standard vocabulary of product managers and marketers. But in a Marketing Bulletin, 1995 article I was shown recently, Malcolm Wright and Don Charlett raise some big questions about the Rogers Model, stating that the Bass Model, also from the 1960s, is more accurate.

Examples they quote to show how the Bass model has given good predictions include consumer durables like televisions and clothes driers, but also more complex projects such as diffusion of cocoa-spraying chemicals among Nigerian farmers, spread of an educational innovation in the US, and purchase of photovoltaic home energy systems in South-West US.

So, what is wrong with Rogers’ model?

I had assumed like many that complex products need a little market testing by innovators and early adopters before the mass market will adopt them. However, some of the examples quoted above in support of the Bass Model instead are pretty complex. Wright and Charlett question two key assumptions:

1. That some individuals are “venturesome”, as a personality trait that is consistent and correlates with length of time in education; however, the evidence for this trait is weak.
2. That the early phase of marketing is dominated by media advertising, and word of mouth becomes important as the market begins to mature.

The Bass Model stresses the influence of interpersonal communication, including nonverbal observation, right from the start.

So why has the Rogers Model been so popular?

My guess is that it probably worked quite well when applied to buggy software that needed a period of beta testing or of being in stealth mode, but then the idea became over-generalized.

If Bass works best, what does that mean for marketers?

Before answering that question, I would pose another one. Why might it be even more important now? The key lies in network effects. Social media creates powerful network effects, so if the power of interpersonal communication was important before, it is now even more so. If the Bass Model is really more accurate, focusing on mass advertising as products are launched, or concentrating mostly on early adopters could waste valuable time and make the difference between product success or failure. The key is to realise that network effects are the best friend a marketer can have, and should be aimed for as early as possible.