FLIRTing with the Crowds

The FLIRT model of Crowdsourcing (which I initially outlined here), has recently reached the point that can be reached with the case material and informal discussions I’ve gathered during the winter of 06/07. Before finalizing* it, it is still essential for the framework to go through a kind of validity and relevancy check through discussions and interviews with the very people that have had great influence in its inception. For this reason, I will in the following few posts describe in detail the different elements of the model, what they include and how they have come to form. I invite everybody to comment on and send further suggestions on advancing the model. I will also send invitations to selected people in the hope of a more formal interview on the subject. If you see yourself as possessing valuable insight to the subject and think you should be included in my thesis’ empirical section, drop me a line and I’ll be happy to interview you (e.g. instant messaging, approx. 1 hour), given that we can find the time together to do this.

In this first post, I will set the stage by outlining the updated model and shedding some light on its birth. The following posts during the next week or so will then delve deeper into individual elements of the model.

The FLIRT model of crowdsourcing

The model as it stands at present consists of five FLIRT elements that need to be considered and established as well as a view to the different levels of participation and how these different levels need to be taken into account in a crowdsourcing project.

The FLIRT elements are:

• Focus**
• Language
• Incentives
• Rules
• Tools

Although the model stresses continuous and open development through constant conversation and adjusting, the elements of the model connect to different levels of decision making, and can thus be, at least initially, thought of as sequential. The first task is to set up strategic Focus and goals for the project, after which tactical considerations, Language, Incentives and Rules can be considered. The most technical level is the Tools, and should be discussed only after the previous elements have been established to a sufficient extent. The following figure clarifies the division between the elements:

The levels of participation, on a rough level, are

• Creators,
• Critics & Connectors,
• Crowds, and

• Non-participating consumers

I decided to set out to depict a little more descriptive model of the various levels of participation than the common 1-9-90 model (1% generate content, 9% interact with the content, 90% simply consume the content) not least because, as it seems at present, the participation rate on the low-activity level can be anything from some percents to 80%, depending on the community in question. While it was not possible within the scope of this study to engage in a research studying all the various roles a person can have in crowdsourcing, I nevertheless felt it important to identify the most obvious ones in order to reflect on these the different FLIRT elements, most notably Incentives and Tools.

Origins, aim and methodology

A few words on the origins, aim and methodology of the thesis. The model has been the result of more than half a year of intensively studying the subject of crowdsourcing and related fields. Being a marketing major at the Helsinki School of Economics, the model is constructed and the problems viewed from the viewpoint of a marketer of goods and/or services wishing to engage its customers, existing or potential, in a collaborative effort, campaign-style or longer term, on a given field of business. Given the nascent and still forming nature of crowdsourcing both as a mode of business and field of study, various related fields needed careful examination in building the model. These other fields included lead user theory and customer participation in innovation, word-of-mouth and viral marketing, social media, user generated content and web 2.0, netnography and new tools for performance measurement in digital media, etc.

Outlined roughly, the research process consisted of

1. a general level information gathering and knowledge building on the subject of digital media and customer participation in the 21st century through a) scientific articles, b) magazines and newspapers and c) blogs and other online resources;
2. a cross-case study of a few successful crowdsourcing efforts/businesses;
3. netnographic research utilizing a set of widely read expert blogs, trusted newsblogs, and other online sources directly or indirectly linked to these primary sources (over 300 articles, all of which you can see here); and
4. expert interviews focusing on the validity and relevance of the model, which, as I already stated, are now in progress.

As the thesis was constructed using abductive research logic (in contrast to purely inductive or deductive), the research process was far from linear and the routes from the phenomenon itself to empirical research to scientific theory were numerous and directed in all the possible ways between the three.

In the following post I will dig deeper into the first element of the flirting with the customer - Focus.

*By ‘final’ I mean how it will be presented in my thesis, not as ‘in its ultimate form’, which according to perpertual beta -thinking is not even a desirable state. As the business models it aims to help build, the FLIRT framework retains a possibility to grow and develop through future insight as well.

**Those familiar with the first FLIRT model (If you don’t know it, better that way) probably noted that the first element in the model is changed from Facilitation to Focus. There is two reasons for this:

1. In order to sensibly establish an effort engage customers through crowdsourcing, strategic Focus of the effort, as described in the following post, clearly needs to be the first element to be considered and established before other considerations come to question.
2. The previously first element, Facilitation (e.g. platform for the community, findability, low barrier to participation) is, after all, a largely technical issue, and fits well as a subchapter of the model under Tools (together with tools for creation, tools for harvesting community input, etc.)