Financial Advisory firms are arguing about furniture while the house is on fire.

A recent article in the business times – Talent war, consumer confusion worry smaller financial advisory firms (1) - makes for interesting reading. To me it’s like arguing about rearranging furniture when the house is on fire.

The article highlights 2 major issues that are affecting Financial Advisory firms:

1. There is a lot of poaching among FAs and Insurance companies, and such poaching may not add value to customers nor indeed to the organisations themselves. This is exacerbated by ‘traditional agents’ being metamorphosed overnight into ‘advisors’.
2.  There are many types of Financial Advisories with different incentive structures and this can muddy waters.

            According to the business times, the 5 different business models are:

1                     Exclusive FA - fully owned by an insurer, mainly selling their product

2                     Hybrid FA - wholly owned by an insurer but also offering others’ products

3                     Dominant FA – insurer has majority stake, but offers others’ products

4                     Independent FA – offers products of many insurers

5                     Exempt FIs – banks, insurers, insurance brokers 

What is the common thread? Or rather, what is missing? 

The customer is an afterthought instead of being at the centre.

The way insurance companies, brokers and FAs have been going about their business has not changed. The focus is on the product, on the distribution channels, not on the customer. That is at the core of the problems for existing players in this space, and why this industry is more than ripe for disruption.

Unless the players change their focus, they might not be around for very long.

Let me first tackle the simple problem of poaching.

Most organisation in this space try to be everything to everyone, hence they create many products with slight variations to address slightly different perceived needs customers; this leads to a huge mess of products that do not make it easier for their distribution channels to identify the best option for each customer (2). This is exacerbated by the incentive structures that are also product focused: sell my latest product and get more commission/fees... This is even true within say insurance companies where the different product lines feel they are competing for the limited customer dollar. (Again, note the emphasis is on flogging my product, not necessarily providing what the customer needs.)

Players who at least focus on customer segments are at a huge advantage over the jack of all trades types simply because they have to understand the needs of the segments they choose to go after, and are thus able to design products that suit these segments, and get the distribution channel that is suitable. The suitable distribution channel could be a mix of agency/advisors of the right skill and demographics and remote channels whether web or phone for example.

An organisation that is able to really focus on specific segments will have a natural barrier to poaching. Imagine you are a distributor, why would you, who are at ease serving the segment you are focusing on move to another organisation where the focus is not there, where the customers may not be from the segment you are good at serving, where the products might not be as suited to your preferred segment? After all, a sign-on bonus can only get you that far and you have to start generating revenue sooner rather than later.

Furthermore, as a distributor for an organisation suited for the segment you are good at dealing with you do not have to worry about your customers churning, decreasing persistency, or even avoiding you because the products you offer them have been designed for them. Customers are not stupid.

The advantages of an organisation being focused increase when the organisation becomes truly data driven and able to execute on a segment of 1 basis, total personalisation. In such a case, the needs of the customer could even be anticipated and a combination of modular features put together to create a personalised product at a personalised price. Why would any distributor want to leave on organisation that enables him/her to reach that level?

Poaching is a direct result of organisations that are not focused of specific customer segments, let alone customer centric, but rather focus on products or distribution channels, and this is not a new problem (3). Apparently self-regulation will not be enough this time and there are rumours of MAS intervening as reported in the business times article.

How about the second issue, that of different business models?

The answer is again the same; these different business models exist because they are designed to reward behaviour that favours the organisations or the distributors, not the clients. For example, FAs owned by traditional insurance companies are likely to pay higher commission/fees to the distribution channel if the product of the insurer is sold. Please note that this incentive to the distribution channel is usually paid irrespective of the degree of suitability of the product to the customer.

Basically, any organisation that becomes customer centric will, along its journey, adjust its incentives to reward customer centricity. From my past experience in the industry, I have seen superstar agencies that are showcased and feted. Basically they attained their status via sales volume, irrespective of what was sold. In fact many of them focused almost exclusively on a very narrow set of products. Unless their customers really need only that narrow set of products, they were not serving their customers like superstars should, on the contrary.

Analytics can help design modular features than can be put together to personalise products to the needs of individuals, but unless the organisation is able to consume this data and act on it, unless the distribution channels are incentivised to make use of these functionalities, then the analytics would have been in vain. This is why the Analytical Maturity of an organisation is critical.

Not that many organisations are even ready to start making the move to being customer segment driven, let alone be truly data driven and customer centric. However those who lag behind will be the first to get overtaken by aggressive disruptors.

While the house in on fire, deal with the fire and take steps to ensure no fire can start again, rather than rearranging the furniture, it will burn with your house too.

1                  http://www.businesstimes.com.sg/companies-markets/talent-war-consumer-confusion-worry-smaller-financial-advisory-firms

2                 In fact they might not need to find the best solution for their customers (that’s if they have fiduciary responsibilities, and what supposedly distinguished advisors from agents), they sometimes just need to find a suitable solution (suitability) https://www.assuredretirementgroup.com/investment-advisor-vs-insurance-agent-vs-financial-advisor/

3                  http://investmentmoats.com/budgeting/insurance/aia-taking-steps-to-stop-poaching-of-agents/

      

NTUC Fair Price's extremely creative use of “data science”, or the crappiest I have seen in a while (1)

I really cannot decide, either it is a stroke of genius that is beyond my comprehension, something a machine speaking its own language has decided (2) (and hence humans will be trained accordingly) or simply a very bad example of application of “data science”.

Let me first give some background. Supermarkets are basically landlords who try to maximise the yield they can get on their space. Hence what they put on the shelves, where the put their shelves, what they keep in stock, when and how they replenish stocks, what items they have promotions on are all very valid questions where the use of analytics and “data science” can be very helpful.

I am sure you would have noticed that supermarkets in different areas of town carry different stuff, this is usually based on what their customers buy and are likely to buy. For example, supermarkets in upper market parts of town tend to carry more premium brands (“King Oscar” rather than “Ayam Brand” sardines, “Bob’s Red Mill Flour” rather than “Bake King” flour).

Where you put the different shelves also matters; for example, you can optimise the route customers take through the supermarket based on the common baskets purchased, and creatively put high profit alternative brands along the most common path.

Also, you must ensure that the stocks keep moving; and that doesn’t apply just to perishables. Hence anticipating when you are likely to run out of stocks of particular items and getting them replenished just-in-time can save you a lot in terms of premium space cost as well as minimising lost opportunity.

Another related question is where to place the promotion items. I am sure you have noticed many of these are neat the cashiers. This is premium space. So are the shelves next to the cashiers that cater to impulse buys; you see sweets, chocolates there, not rice or milk. One step further goes to pricing the promotional items, you might decide to take a hit on one item, knowing that most people will also purchase another item where you more than make up for the discount.

Please don’t say “diapers and beer”! A friend of mine who has intimate knowledge of the project where this idea originated gets very angry and declares this a myth that nobody has the incentive to debunk. Anyway, the fact is some products sell together, and placing them strategically is a good way to increase sales.

Take a look at this beauty:

This is a picture from an NTUC Fair Price I visited over the weekend. Let me list the collection of goods in this picture:

o   Red wine

o   Tomato Ketchup

o   Chocolate

o   Hair Colouring

o   Toothpaste (whitening and for sensitive teeth)

o   Facial Tissue

o   Titanium Kitchen Scissors

o   Cooking Oil

o   Children’s Milk

o   Instant Noodles

o   Chicken Essence (a common energy/health drink)

I have no clue what MacGyver would make of all this, to me either this is the mess that lies between your kitchen and your toilet, or a vain survivalist’s dream. But since there are well stocked bomb shelters all around Singapore, I am not sure this NTUC Fair Price is catering to vain survivalists.

Hence, to the “data scientist” who came up with this planogram, please check what language your AI is speaking. Actually it’s probably laughing.

My rant doesn’t end there...

It came time to pay.

In another amazing attempt at increasing efficiency, the supermarket decided to allocate their space to 1 human manned counter, and 5 or 6 self-serve counters.

One of the interesting things about the self-serve area of NTUC Fair Price is that the baskets need to be placed on a weighing platform (a basket just about fits on the platform); hence people who are using trolleys, unless their trolleys are only carrying a basket worth of goods have to go the human way.

The human manned queue was loaded with people pushing well packed trolleys. Basket people were therefore moving to the self-serve. So far so good you’d say. (Not for me, I don’t enjoy corporations pushing work down to me the customer while they increase their profits, and for other reasons (3), but I won’t go there now).

When it becomes fun is when the other interesting thing about the weighing at NTUC Fair Price is that there are 2 scales: one where you put your basket of purchases before you pay, one where you place your purchases (without basket) as you scan them. Your scanning is accepted when the weight that leaves one scale is equal to the weight that is added to the other (I don’t know the tolerance level).

This becomes really fun when you have purchased many items. While you can stack items taking advantage of the sides of a basket, on a flat platform this collapses. Anything that decreases the weight due to falling off the scale for example triggers and alarm and you have to wait for a human to intervene, you can also cancel and restart.

It’s even more fun when you purchase light items, such as a toothbrush. The alarm will ring likely because the weight of the toothbrush is within the sensitivity range, and the system cannot decide whether the item has left one scale and ended up on the other; hence human intervention is required.

What was the outcome of all this application of technology?

Well... The 1 human cashier was very busy dealing with the long queue of trolleys (4). The human helper of the self-serve robots was the busiest bee, buzzing from alarm to alarm. The human customers at the trolley queue had to wait longer since there were so few cashiers able to serve them. 

The customers at the self serve have to wait longer since the alarm kept ringing and ringing. In sum the cost to employees and customers went up.

But eventually, I guess the data will show that the NTUC fair price had brisk sales, a large portion of the customers went through self-serve – they were incompetent hence causing the alarm to ring, but that doesn’t change anything. Someone’s KPI has been hit. This is even more true if the NTUC Fair Price employees have not been compensated for the increase in their workloads.

Furthermore, if people have bought oil, or hair colouring, or red wine, then may be the record will show that the planogram worked.

I can hear the machine laughing while it trains us... (If it was autonomous enough it would probably be rolling on the floor laughing)

Not me; I am not going to that NTUC Fair Price again (5).

(1)    You could argue there was no “data science” involved; but then again I’d ask what the NTUC “data scientists” up to then, or whether NTUC Fair Price has a policy of wasting shelf space? or may be they are just testing our reactions... ? https://www.wired.com/story/virginia-self-driving-car-seat-disguise-van/

(2)    https://www.theatlantic.com/technology/archive/2017/06/artificial-intelligence-develops-its-own-non-human-language/530436/

(3)    http://thegatesofbabylon.blogspot.sg/2015/12/think.html

(4)    Actually it was a single basket queue, so customers were actually expected to put baskets in their trolleys and pay basket by basket at the self-serve, or break the rules that NTUC itself imposed, if the cashier allowed. Fortunately, the human cashier showed common sense and allowed trolleys at her queue; with a single basket, I understood her situation. Human accommodates, what’s new?

(5)    To add insult to injury, the supermarket was very tight, the queues spilling over beyond the minute queuing space, and someone in wheelchair was much inconvenienced.

   

  

The death of the “Data Scientist” or the return of the programmer

The “Data Scientist” is dead; long live the programmer! The world will soon not need “data scientists”, but instead programmers will rule the world.

This brutal assessment comes from the World Economic Forum (WEF), and I must say that their arguments make a lot of sense.

In this age of fake news, credentials are important; so what is the WEF? Founded in 1971, it is committed to improving the state of the world, is the International Organization for Public-Private Cooperation (1). Its board of trustees includes world leaders such as Jim Yong Kim – president of the world bank, Christine Lagarde – Managing Director of the IMF, but also Jack Ma – founder of Alibaba Group, Mukesh Ambani – Chairman of Reliance Group among others. (2)

Furthermore, the “centre for the fourth industrial revolution” is one of their major undertakings (3), and one of the focus areas is artificial intelligence and machine learning (4).

So this is not fake news, neither is it flimsy covfefe gossamer; sad!

In an article entitled “You’ve heard about it, but do you understand?  Everything you need to know about Machine Learning”(5), the WEF gives a very easy to understand convincing description of what machine learning is, how it can be used, and what is the way forward.

The article covers the history of Machine Learning, broad classes of machine learning (supervised – the machine is told what the different groups are, unsupervised – the machine is left alone to find out the different groups, and reinforcement machine learning – after finding out the groups it is told whether it is right or wrong this feeds back into the learning cycle.

In illustrating the concept of reinforcement ML, the WEF shows a clip where a machine learns to play Atari Breakout – where a horizontally mobile paddle at the bottom of the screen is controlled and the aim is to reflect a ball back to destroy an obstacle made of bricks at the top of the screen. The results are amazing (6).

But even more astounding is the commentary:

“The most important thing to know is that all the agent is given is sensory input (what you see on the screen) and it was ordered to maximise the score on the screen”

“No domain knowledge is involved! This means that the algorithm doesn’t know the concept of a ball or what the controls exactly do”

After 10 minutes the machine is clumsy. After 120 it is playing like and expert. After 240, it has developed its own strategy!

Isn’t this amazing?

So you will ask, what does that have to do with the death of the “data scientist”?

Well, if you, like me, think that Drew Conway was on to something when he came up with the basic “Data Scientist” skill diagram (8):

 

In this diagram, data science requires substantive expertise, or “domain knowledge”.

What the WEF agreeing with is that Machine Learning does not require “domain knowledge”. Furthermore, there is no mention of the “data scientist” in the WEF view point, but rather the article highlights “the programmer”:  ML “needs a programmer to tell it what to do when it is fed with data.” Not a “data scientist”, a “programmer”.

But does that mean the death of the “data scientist”?

Yes it does.

Think about it, if an problem can be solved without “domain knowledge”, then ML is all you need. 

And what does ML need? Not pompous “data scientist”, but the humble “programmer’ who had been cast aside and finally makes his/her triumphant return.

There is a wave washing away the veneer off the “data scientist” and restoring that of the programmer. An example argues that coding is not fun, but it is technically and ethically complex (9). 

What I find most interesting is the acknowledgement of the complexity of coding (I am sure most people would agree that it takes serious skill to develop industrial grade code), but also the introduction of ethics into the mix.

To quote from the article “As well as being highly analytical and creative, software developers need almost superhuman focus to manage the complexity of their tasks. Manic attention to detail is a must; slovenliness is verboten. Attaining this level of concentration requires a state of mind called being ‘in the flow,’ a quasi-symbiotic relationship between human and machine that improves performance and motivation.”

To push the idea further, the latest triumph of machine over man was Alpha-go. Alpha-go was built by the same organisation that developed the machine playing Atari Breakout above. And the same principle was used: “no domain knowledge”, but lots of Monte Carlo Simulation. (basically the machine learns which move is best by, from the current position, randomly trying different plays and evaluating the end result, then assigning a value to that play and picking the one that gives higher change of winning/lower chance of losing, and repeats the process at every turn)

No domain knowledge

No data scientist

One of the greatest triumphs of machine over man was enabled by programmers, not data scientists.(9)

The “data scientist” is dead, long live the programmer!

 

 Further readings/references:

1. https://www.weforum.org/about/world-economic-forum
2. https://www.weforum.org/about/leadership-and-governance
3. https://www.weforum.org/center-for-the-fourth-industrial-revolution/
4. https://www.weforum.org/center-for-the-fourth-industrial-revolution/areas-of-focus
5. https://www.weforum.org/agenda/2017/05/what-is-machine-learning
6. https://www.youtube.com/watch?v=V1eYniJ0Rnk
7. http://drewconway.com/zia/2013/3/26/the-data-science-venn-diagram
8. https://qz.com/987170/coding-is-not-fun-its-technically-and-ethically-complex/
9. https://www.tastehit.com/blog/google-deepmind-alphago-how-it-works/