Just like the iceberg, it’s only too easy to concentrate entirely on the parts that the user can see. It may be obvious that you will need input questions, some logic and decision trees that make up the recommendations, but there is so much more to delivering an effective and engaging robo-advice process.
In this series of blogs, I set out some of the considerations you should be looking at if you want to deliver a truly robust and long term robo-advice service.
Who is taking responsibility?
Creating a robo-advice service does not mean that a robot writes the site, obtains FCA permission, researches the platform, sends in the compliance returns to the regulator and takes responsibility for the advice.
Perhaps this will happen one day but, for now, a real person is needed to perform each of these roles, with ultimate responsibility still resting with a nominated regulated financial adviser. As a result, the robo-advice process that is being followed should be taken as seriously as face to face advice.
Can you explain the client’s expectations consistently?
In a face to face meeting, the client’s risk and reward expectations can be discussed and set. When it comes to robo-advice, this conversation has to be replaced by something that the customer learns for themselves. A great way to do this is to show what could happen in the future to their investments by providing forecasts.
As well as the user interface explanation (the tip of our iceberg), the hidden issues for robo-advice are to ensure that these forecasts are fit for purpose and are consistent with the forecasts used in your face to face advice process.
At EValue, a lot of care and attention is taken to ensure that the basis on which our forecasts are set (Economic Scenario Generator stochastic modelling) is designed specifically for long term customer outcomes. It is also the basis for our adviser tools, to provide consistency. Does this apply equally to your own robo-advice offering and the face to face advice process you offer?
Is the risk questionnaire going to work with the wider population?
There are many possible risk questionnaires you could use in your process. However, the target market for robo-advice has the potential to be a much wider group covering knowledge, experience, age and existing wealth, than for face to face advice. Consequently, you will need to check that the method used to create the questionnaire still works across this wider range of clients.
Unfortunately, relative scoring in questionnaires has the potential to skew results if used in robo-advice. To see this in action, consider an exam where the bottom 25%, say, of all the students get a “D”. If the exam is then opened up to the general public, most of the students previously scoring a “D” would see an instant increase in their grade. This is because, although they themselves have not answered the questions any differently, the public as a whole will typically score lower than the students, having not done any study for the exam.
The same applies to attitude to risk questionnaires. If you consider how many customers (lots!) and what type of customers (probably a lot more cautious) will be using the questionnaire, then the chances of the results being skewed across the board is significant.
To overcome this, a questionnaire with an absolute scoring system, such as the one provided by EValue, will be needed. In the exam example above, those who hadn’t studied would simply score a low mark on the test whereas those who scored a “D” before would still be graded a “D”. That sounds a bit fairer all round to me, although I’m sure that those with the instantly improved grades would disagree!
If you like this blog, please spread the inspiration by sharing this blog with your business colleagues and friends on LinkedIn, Twitter, Facebook and Google+