Find the 'Move 37 Moments'- they are everywhere!

The core of why we setup MV37, was that we wanted people to leverage emerging technologies to really make a difference to a business.

The essence of choosing the name MV37 (“Move 37”) was to embody the idea that data in particular, but more widely technology can and should be used to look at our world differently. We have seen many opportunities for change that have been ignored, due to our clients innate comfort at looking at problems through a human lense e.g. our social norms, trusting our gut / intuition / monkey brain or just the way things have always been done.

Game 2 : Move 37 in AlphaGo V Lee Sedol, is just one example of many where, when you apply technology and take the passion or emotion out of decision making, one can find a wholly new understanding of the problem space and potential solution. AlphaGo found a solution to a problem that a human couldn’t due to the human lens we apply. But once a human sees the solution, it is like a shroud being lifted from one's eyes. I am sure many of you will have had situations where once you realise something, it becomes absolutely bleeding obvious and you can’t imagine how you didn’t see it before. This is what we aim to help our clients do, we call it a ‘Move 37 Moment’

But ‘Move 37 Moments’ are not just about decision making. It is about using technology to find fundamentally different approaches to doing business.  In the last 30 years businesses have thought and re-thought manufacturing processes and approaches using Robotisation, Lean Manufacturing, IoT etc. But outside of the introduction of the personal computer, the white collared jobs have generally not been wholly re-thought. However, a recent McKinsey study found that 45% of all current jobs could be automated today with the current level of technology, that is a significant finding. It means we don’t need to perfect quantum computing or build loads of Large Hadron Colliders, the majority of the benefits are possible now, if we can look at our current business approaches differently and remove our shrouds.

I want to provide some examples of the sort of re-thinking that needs to happen, from the simple to the more complicated to give you a better sense of what we think of as Move37 Moments. These are instances where there are now wholly better ways of doing something, if you can lift the shroud of the human lens.  

MV37 Moment Example 1: A simple application of Natural Language Processing (NLP) and Machine Learning (ML)

This is an example from an Investment bank, but is relevant to many industries.

Trading is a complex undertaking and over the years many systems have been put in place to handle all aspects of trading different asset classes. When one of those systems has a problem with processing a specific trade they create an error alert and that trade gets stuck. Many banks have armies of people whose job it is is to monitor these alert conditions, reading what has broken and forwarding the alert to the relevant department to solve the problem. In one small area of a bank (Equity Derivatives) there are ~70 such systems supporting that trading activity. When a trade breaks in one of those systems it sends an email alert, and there is a team of 100 people reading the alert emails and forwarding them to the right department to fix the trade break. They are acting as human workflow engines forwarding emails around the bank. This is repeated across the bank in many pockets managing the 000’s of trading systems.

If one applies some simple Move37 thinking, you see that this is a waste.  With the application of an NLP engine, to read and process the alerts from the systems and a simple ML model, trained on how to route these alerts, this process can be automated. This is not hard to do. It does not need some massive development team, and will make the business significantly more efficient for every year going forward. A small number of people (<10) will be required to monitor the model and correct processes when the systems change their alerts. In one bank alone, we estimate this would free up +1000 people who could be employed elsewhere in the bank.

This is a very simple solution and anyone could have come up with it, but the bank in question didn’t.  When asked the reasons were varied...1) because the operational team has always been there, 2) they are a standard expense that the bank carries, it is an OPEX cost that is considered as a sort of the base-load of what it costs to do business, 3) it has been outsourced so very few people know how or why the team does what it does, 4) it works.

Fundamentally the managers had not sat back and looked for the Move37 Moment, why do we have 100s of people reading an email from a system and forwarding it the right department in a company. That is dumb. But it isn’t until someone lifts the shroud and asks the question; when they do, it becomes obvious and somewhat embarrassing.

MV37 Moment Example 2: Execution Efficiency - Machine Learning and Credit Rating

This comes from Financial Services again and is a more complicated example, so the Move37 Moment is not as obvious. But it shows the power of emerging technology techniques to do something that was not possible when the original process was defined.

When an institution is deciding if it will lend an asset to a counter-party, they need to assess the risk of that counter-party going bust in the period they are lending it. This is so that they can price in that risk and cover their exposure. This has been happening for centuries. But when this trading started, there was a lack of easily accessible public data so it was hard for an institution to have trusted information on all their counterparties. Institutions would only lend to people they could trust, or had a track record. But this doesn’t scale, so back in 1841 the first credit agencies were setup to provide credible information on different counterparties to the institutions. This function is still performed today by the ratings agencies, such as Moody’s, S&P and Fitch. They provide a rating on the risk that the institution can use in its calculations.

However, they can differ in their judgement of the risk of a certain counterparty. So, the financial institution employs a team of people to make a decision on which risk to go with, using models and their judgement/intuition.

However, with the many diverse data sets that are now available which an institution is able to get access to.  They can augment the rating agency data with significant public digitised information about companies and their own data about sector performance. Using this data and Machine Learning it was able to create a model that would give a percentage likelihood of default, which was more accurate than the human intuition and ratings agencies.

This approach was run in parallel for 6 months alongside the existing process, and back tested against previous defaults. It was found to be 89% more accurate at predicting a default then the old system. It is now running across their $500BN book.

The institution has raised its confidence in risk assessment. Meaning they are now able to price more keenly, make their capital work harder and take on less trades that go bad. This is a Move37 Moment because it uses emerging technology to effect a significant change in the current process which has existed for centuries, leading to a material increase in performance. By looking to maximise the execution of a process they have increased revenue opportunities and reduced their risk.

There are many other examples I could quote, including the much discussed opportunity for Distributed Ledger Technology (DLT), or Blockchain - as it is erroneously labelled - to provide many Move37 Moments, such as completely rethinking the contracting process across all industries; integrating with IoT in Industry 4.0 and generally removing manual effort and standardising on approaches. Lots of these examples will make all participants more productive and can be applied internally or across an industry to make major gains.

It is proven that the combination of AI, Distributed Ledger Technology and commodity computing (“Cloud”) enables us all to make many Move 37 Moments across our businesses. The main challenge for our clients is taking off the shroud and looking for them.

You need to be self critical of your business. Look at processes that have existed for years or even centuries and challenge yourself on if they could be a candidate for a Move37 Moment. Then start some experimentation to see what is possible.

Don’t worry about the technology, nearly everything you can think of is possible. Just let your imagination run a bit. Remember if 45% of your company’s processes can be automated, you should find Move37 Moments everywhere!

If you want to find some inspiration, contact us. We’ll go for a coffee and explore your ideas.