It's no secret that most fin-tech apps are rarely used at all although a handful, the authors of which, have done the necessary research and development, become the rock stars of FIN-TECH. How bad is the spurn rate for the ignored apps? You can go Google the various market research reports on the topic and draw your own conclusions but the magnitude of the issue was brought into sharp relief for me when I examined an app produced by a wealth management firm for its clientele. It launched with 10,000 invites that generated 5,000 opens and after 12 months now has the grand total of <10 regular users (twice or more per month). That's a frightening spurn rate in anybody's book. It's my understanding this is not atypical. What a waste of $250-400K. Spending 10-12% of that on a behavioural systems engineer may have been a smarter investment. Clearly, there is room for improvement but where do you start? How about with a concerted effort to complete research due diligence from the outset and engaging in an honest, learned appraisal of the available evidence. From this point, at least the participants may get to a realistic point of go versus no-go to build with any go-decision based on real world validated data and perhaps some helpful and relevant models.
It's not so much a question of which guiding theory or model would-be developers should be encouraged to select as a product development foundation but rather it's more a case of actually using something with a reasonable evidence base. I'd point aspirants to a broad brush of approaches spanning a variant of the Action-Control Theory, application of the Transtheoretical Model, and the delights of digital nudging; (Mirsch, Lehrer and Jung, 2017) and (Weinmann, Schneider and Brocke, 2016). From experience, it's often best to pick the best of a number of models before embarking on analytic activities. I'm a fan of the Financial Management Behaviour Scale (FMBS), (Dew and Xiao, 2011) and (Xiao et al., 2004), although I have seen its results butchered on a major financial services website. Certainly, the validation work from the researchers is a good read and the admitted limitations and constraints touched on in the discussion are worth bearing in mind before trying to make use of the scale. It does however reveal that increased FBMS scores reflect increased levels of saving and greater debt reduction. Look, its not the be-all and end-all but it's a reasonable start point.
Of great interest to me and the methodology we use is the Personal Savings Orientation (PSO), (Dholakia et al., 2016) based on the Action-Control Theory. The researchers here were adamant that the control of actions in saving money and monitoring spending activities over time tends to be based on the motivational tendency of consumers that presents as a trait. They set out to show why and how some consumers save and accumulate and others fail. The work may provide part of the basis for designing interventions that transform money saving actions into habits and a shift in the lifestyle choices that may ameliorate this. Again it's only part of the picture and requires widespread testing and re-validation "in the wild". And herein lay the rub. The fact is not only is there a lack of suitable practitioners to design and test appropriate scales but also a reluctance to invest in them. Yet, unless you know for sure what is affecting the financial behaviour of your customers and why it is difficult for them to save and reduce debt, how do you expect to build anything that may help steer them in an affirmative behavioural direction using nothing other than guesswork?
Still, occasionally the right people in the right place at the right time work out they may need help. As a practitioner, it's not enough to sit down and design the survey and generate the models for them, send the invoice and swan off. If, as a behavioural systems engineer you want to make a difference then you need to dovetail your output into agile engineering practices and be first amongst equals at the table when it comes to round tripping the original findings in your models. This is the space we work in on a daily basis and whilst accumulating a wealth of hard-won experience, I'm not sure we have answered more questions than we have asked. What we have learned though is you must work closely and consistently, right through to user acceptance testing with the UX team. There is a natural fit between your models and their interaction design practices.
Emotions & personality
It's not all methodology of course, if your team is serious about making a difference with the app you need to consider how you can not only get a better handle on UX and manifest that through adaptive UI but also how you may be able to tap into emotional and personality-based datasets to potentially pre-empt impulsive behavioural incidents. There are a small number of kick-arse API's and SDKs for emotion tracking and analysis that are worth considering. In the next 6-8 weeks I will delve down into the newest and most interesting of these but for now at least make a slot in your architecture for a RESTful API that helps track and support emotional undulations in your user populations that may otherwise derail affirmative behavioural intentions for your app.
Dew, J. and Xiao, J. (2011). The Financial Management Behavior Scale:Development and Validation. Journal of Financial Counseling and Planning, 22(1), pp.43-59.
Dholakia, U., Tam, L., Yoon, S. and Wong, N. (2016). The Ant and the Grasshopper: Understanding Personal Saving Orientation of Consumers. Journal of Consumer Research, 43(1), pp.134-155.
Mirsch, T., Lehrer, C. and Jung, R. (2017). Digital Nudging:Altering User Behavior in Digital Environments. In: 13th Annual Conference on Wirtschaftsinformatik. St.Gallen, pp.634-648.
Weinmann, M., Schneider, C. and Brocke, J. (2016). Digital Nudging. Business & Information Systems Engineering, 58(6), pp.433-436.
Xiao, J., Newman, B., Prochaska, J., Bassett, R. and Johnson, J. (2004). Applying the Transtheoretical Model of Change to Consumer Debt Behavior. Journal of Financial Counselling and Planning, 15(2), pp.89-100.