Blog article
See all stories »

Regulation Chatbots: Knowledge First, Data Second

Data is the enabler for any digital strategy. Organizations need timely and accurate data from human interactions. This is an imperative for sensing early and responding quickly to ever changing circumstances.

We all collect data from customers, employees and suppliers using forms, checklists and surveys. This data must be of high quality otherwise it compromises productivity, performance and decision-making.   

Data input is prone to error when there is dependency upon contextual understanding. This error rate rises when the required knowledge involves regulation.

By regulation we mean the fabric of related knowledge. This includes national laws, statutes, guidelines, regulators handbooks, and standards such as ISO. The regulation fabric also includes deployment in practice through governance, policies and procedures.   

As regulation increases in complexity, the greater the risk of data input misinterpretation, bias, distortion and deception.

Poor quality data input related to regulation is highly problematic as it:

  • Delays visibility of non-compliance.
  • Lacks full transparency and traceability.  
  • Weakens tactical and strategic decisions.  

The implications are profound.

The current method of data input is deeply entrenched everywhere. This means the design and deployment of forms, checklists and surveys are formulaic.

Resistance to a new framework of thinking is a challenge. However, a tipping point has been reached. This is best understood through examples and evidence. 

Let’s take treating customers fairly.

The tick box is the most simplistic data input. But then again, the tick box is a binary decision, which means it is prone to misinterpretation, bias, distortion and deception. The European Union GDPR (General Data Protection Regulation) explicitly states that the tick box alone is unacceptable for customer consent. There needs to be clear proof of the customer understanding the context of each way their personal data could be used.   

The FCA (Financial Conduct Authority) in the UK has defined six principles for treating customers fairly. These principles govern the life cycle of any financial product. In practice, this requires data input across each event such as selling to a customer or a customer making a complaint.

Take in retrospect the misselling of PPI (Payment Protection Insurance). This involved the full spectrum of data input: forms, checklists and surveys. The solution deployed was formulaic across financial institutions. This included checks and balances using supervision, training, compliance assessments and audit reviews.

Nevertheless, systemic weaknesses were not sensed early. The tick box combined with all the other evidence failed to prove beyond reasonable doubt that the customer understood the product. The misselling of PPI has led to non-compliance pay-outs of more than £44bn.

Why is this of strategic importance? A recent study found that non-compliance is 2.7 times operationally more expensive than being compliant. This cost includes business disruption, declines in productivity, fees, penalties and other legal and non-legal settlement costs.

Many more examples can be found in the UK.  

  • The NHS has over 20,000 forms, checklists and surveys, which is a fertile ground for systemic risks. They set aside a £65bn provision for clinical negligence in 2017.
  • Health and Safety has many forms, checklists and surveys. Again, there are systemic weaknesses. The costs of workplace injuries and work-related ill health is circa £15bn per annum.

 It is time to think differently. A new sense making framework is needed.  

The mantra everywhere is to be data driven. But is this right for data input related to regulation?

Compliance within institutions is at a breakpoint. There were 56,321 regulatory alerts worldwide in 2017. This is an increase of 547% since 2008. The implications are profound for related data input.  

Regulation Chatbots is a new approach powered by knowledge. This knowledge is used to influence decision choices, which in turn determines the pathway that leads to a distinctive outcome. Data input is collected in context to the conversational pathway.

The user decision journey is automatically captured as composite data, which is a combination of dialogue data and contextual data input.

Knowledge first, data second is the mantra for Regulation Chatbots.

A Regulation Chatbot simplifies and streamlines knowledge processing and contextual data input. This capability provides timely and accurate data assets from front-line activities.

The ability to sense early and respond quickly to ever changing circumstances is now attainable. This new benchmark of transparency and traceability, safeguards the brand, balance sheet and people.

This is the beginning of a seismic change for both the public and private sector.

 

  

 

3460
External | what does this mean?
This content is provided by an external author without editing by Finextra. It expresses the views and opinions of the author.

Comments: (0)

Freddie McMahon

Freddie McMahon

Chatbot Thought Leader

DF2020 Ltd

Member since

04 Aug 2017

Location

London

Blog posts

57

Comments

2

This post is from a series of posts in the group:

Banking Strategy, Digital and Transformation

Latest thinking in respect to Banking Strategy, Digital and Transformation. Harnessing our collective wisdom to make banking better. Ambrish Parmar


See all