Signal Detection Theory Innovation

A common theme in innovation is the importance of identifying the most promising innovation projects as opposed to those that are less likely to succeed (see Paul Sloane’s excellent posting on killing weak innovation projects to protect the strong projects). Other innovation practitioners may refer to this as separating the wheat from the chaff, managing the signal to noise ratio, or following an Integrated Product Development process with a managed pipeline and phase gates. While I agree that there is value in applying strict and objective scrutiny to the assessment of ongoing innovation projects, I noticed a concept recently that suggests an alternative approach to the process of investigating ideas as part of an innovation pipeline. In an article titled “Why We Should Scam the Scammers” in the Wall Street Journal, Christopher Chabris and Daniel Simon highlight an interesting, counter-intuitive approach to scam emails originally suggested by Microsoft Researcher Cormac Herley. This approach, in which there are benefits derived from engaging in a process that at first glance should be avoided, provides interesting implications for approaches to assessing the value of innovation ideas.

It would be difficult to find a regular user of email who has not at some point received an email from an exiled member of the royal family from an African country claiming to need a bank account to temporarily house some of his great wealth, for which the cooperative person will receive a generous commission payment. Herley, in a paper delivered at a security conference in Berlin, applies signal detection theory to the scam to develop an alternative approach to how to handle these scammers. Signal detection theory focuses on the importance of discerning the signal from the noise in any analysis.

In the case of the email scammers, Herley turned the analysis around and looked at the situation from their perspective. The scammers send thousands and thousands of emails hoping to find the small handful of individuals who could actually be duped into providing bank account information (or sending a payment, in the myriad variations of the scam). The scammer must weed out any false positives from the very small subset of individuals who respond because the scammer has a finite amount of time to parse through any responses to determine which ones are worth investigating. Sending out thousands of emails takes little time, but deciphering the responses, and deciding which ones are worthy of a follow-up email, is much more challenging.

This is where Herley provides a fascinating alternative approach. Today, when we receivethese emails, our immediate response is to delete them and/or mark them as spam, hoping that we can block them in the future. The scammer is not impacted by this action as we are doing part of his job for him by not sending back a false positive. In essence, we become part of his work force. Blocking his email address is only a temporary setback, as he can easily change domains for future email blasts. Herley’s recommendation is counter-intuitive in that he instructs us to engage the scammer and overwhelm the person back with a deluge of false positives that would require a great deal of labor on his part to investigate. This bombardment of replies could be done manually with quick notes from individuals or, more promisingly, through a computer-generated reply that mimics the varied nature of human responses, including typographical errors, the use of colloquialisms, and even the use of realistic-looking bank account or credit card numbers. Just as the scammer seeks to draw us into his web ever more deeply with each exchange of information, we, too, could ensnare the scammers by dragging them down a path that inevitably leads to a dead end, but not before the scammer has invested large amounts of time and energy, thus limiting the damage that person could do against a real potential mark.

Reading about this approach to counteracting scammers made me think of the process we typically follow in innovation of following up on an initial idea proposed by a participant in an innovation workshop or jam session. The normal approach is to acknowledge an idea proposed by a participant by repeating it verbally and then writing it on a flipchart or whiteboard to record it. In some cases, the innovation session facilitator will identify an idea that has potential merit and engage in a more thorough discussion about that concept with the initiator of the idea and the workshop team as a whole, driving out multiple threads and new concepts to explore along the way. While this approach has been proven useful time and again in innovation, reading about Herley’s counter-intuitive approach to scam emails led me to wonder if there is value in looking at less-promising innovation workshop or jam ideas differently. Just as the normal response to a scam email is to mark it as spam and ignore it, so, too, is our normal response to a less-promising idea proposed in an innovation session to log the idea and essentially ignore it.

I am particularly guilty of leaping to conclusions about some ideas without exploring them more deeply, as would be the case for an idea that does not match the theme of the innovation session. For instance, we might be in the midst of an innovation jam dedicated to exploring the theme of social media and a participant introduces an idea on the topic of mobile device management. The challenge for the innovation facilitator is how to explore that thread to derive value from the new idea without derailing the overall momentum of the session. My tendency in these scenarios is to acknowledge, record, and move on to another participant, but we may want to consider the benefits from driving into every idea, at least a few layers deep, to explore the concept. This approach is based on the assumption that every individual has unique value to contribute to a session, and somehow that individual made the connection between the topic under discussion and the concept that he or she introduced. Although it could be that the person was not paying attention to the session and introduced the idea without a valid reason, there may be subconscious factors at play in that individual’s mind that are worth examining to see if it can lead to a better overall outcome. I do not advocate getting bogged down exploring every thread that comes up in a workshop, but I believe there is value to exploring counter-intuitive concepts as a way of avoiding self-imposed constraints. Such an approach could increase the creative aspects of the workshop and open up the discussion to new ways of thinking. The innovation facilitator will need to work harder to keep the overall session on track, but the improved output could be worth the effort.

Source: Christopher Chabris and Daniel Simons, “Why We Should Scam the Scammers,” Wall Street Journal (August 4, 2012). image credit: bluedish & wikicommons

Don’t miss an article (4,500+) – Subscribe to our RSS feed and join our Innovation Excellence group!

Scott Bowden works on Innovation Programs for IBM Global Services.