Gaming the system

By Slow Dad - April 22, 2017

Ever wondered why people do what they do? Imagine being able to monitor and test how people respond to financial incentives. I did just that!
One of the most fascinating projects I’ve worked on was a very “big brother” style behavioural monitoring system for a large retail bank.

The project sponsor had figured out that while most staff generally try to do a good job, they’ll try much harder at the tasks for which they perceive they have been incentivised.

The sponsor hypothesised that if the incentivisation was aligned to the bank’s strategic goals, then in the staff would do more of the things that would help achieve those goals and hopefully do them better.

Like all great plans it sounded simple and plausible.

Everybody has a plan until they get punched in the face

Unlike many executives I have met in my time, the sponsor didn’t just like believing he was
omnipotent
infallible
possessing the looks of George Clooney/Julia Roberts, the brains of Stephen Hawking/Rosalind Franklin, and the sporting ability of Sachin Tendulkar/Ellyse Perry correct, he liked being provably correct. To achieve that he needed to have his hypothesis tested.

George Clooney and Julia Roberts in Ocean's Eleven.
George Clooney and Julia Roberts in Ocean's Eleven.
A random (actually it was located on the ground floor of the bank’s headquarters building, located next to a great coffee shop, and down the road for a place that did amazing chicken curry roti rolls for lunch at student prices) bank branch was chosen. The staff working in that branch were to become the guinea pigs in what proved to be a fascinating behaviour economics experiment.

A quick review of the current remuneration and incentives on offer was performed.

The staff were all permanent employees, earning above average but far from amazing wages. This meant they got paid just to turn up. In fact with paid annual leave, sick leave, maternity leave, bereavement leave, study leave, and the like they pretty much got paid regardless of whether they turned up!

If a staff member managed to sell a credit card product they received $25.

If a staff member managed to sell an owner occupier mortgage they received $125.

The thing was a credit card is a “do you want fries with that?” convenience product. It takes almost no time for a staff member to complete the required paperwork whenever some gormless lackwit inevitably said yes to the opportunity to take on even more unsecured personal debt without sparing any thought about how they would pay it back.

A credit card earned the bank an annual fee of $25, plus around 28% interest on outstanding balances.

A mortgage on the other hand was a never-ending paperwork nightmare for staff to endure. Background checks, identify checks, credit checks, valuation checks and so on. The conveyancing process must be successfully traversed in order for the mortgage to be drawn down, and the staff didn’t receive their $125 incentive until the punter had jumped all of those hurdles.

A mortgage earned the bank an establishment fee of around $600, plus around 6% interest on the outstanding balance.

Unsurprisingly branch staff didn’t invest much of their time trying to sell mortgages. In fact the opposite was true. Staff actively discouraged any random drop-ins enquiring about mortgage products, simply because of their sheer horror at the prospect of running that compliance driven paperwork gauntlet.

Keeping your eye on the prize

The sponsor and I asked ourselves what we would do in the branch staff’s situation. The sponsor said he would sell as many credit cards as possible.

I called bullshit on this. The incentive was so low that it wasn’t worth chasing. If a customer asked about a credit card I’d sell it to them, but I wouldn’t put myself out trawling for business.

Then the sponsor nodded and begrudgingly agreed with me. $25 was chump change to him, it would not make a material difference to his pay packet. It wouldn’t pay for his groceries for a week, or even a full tank of petrol in his car.

The question was how much incentive was enough?

How much is enough?

To answer this I needed more information.

  1. I needed to understand how much each individual product was actually worth to the bank, as sustainable maximum incentive levels had to be below this amount.
  2. I also needed to understand the volume of organic product inquiries for each type of product. This would provide a baseline against which the effectiveness of any incentivisation scheme could be measured.
  3. Next I wanted to know what the conversion rates actually were. What percentage of product inquiries were successfully converted into sold products.
  4. Finally I needed to know the amount of time and effort required to actually sell each type of product.
Once armed with these facts I was in a position to commence the incentivisation mix experiment.

Human guinea pigs for fun and profit

Over the next six weeks various incentivisation mixes were developed and applied. Results and the ensuring staff behaviours were carefully monitored in an attempt to identify the optimal combination of carrot and stick required to achieve the bank’s strategic goals.

So far, so good.


However once I started evaluating the results I noticed something interesting about human nature.

People are largely driven by self-interest.

They are also for the most part lazy and impatient.

These factors combined meant employees seized on the immediate rewards offered by the easy to sell commodity products such as credit cards. However they still shied away from those products requiring more work, like mortgages and insurance products.

Nothing earth shattering there.

Selfishness and self interest win out

Then I noticed that some of the behaviours I was observing didn’t make economic sense. Staff were undertaking activities that would not result in the optimal financial outcome for themselves.

When I dug into this I discovered that the staff had interpreted some of the incentive mixes differently to how I had intended them to be applied, and were attempting to game the system.

On the one hand I was relieved that I could now understand their behaviours, and self interest was still driving their prioritisation decisions about which activities they would pursue and which they would avoid.

This presented the project team with a challenge, as not only did I need to evolve an optimal incentivisation mix to elicit a desired set of behaviours, I also needed to factor in the branch staff attempting to game the systems I was developing. This resulted in my attempting to second guess the likely responses of the staff, creating an elaborate game of cat and mouse that was in equal parts frustrating and fascinating.

Lessons learned

At the conclusion of the evaluation period I prepared some recommendations for the sponsor (who had been actively involving in pulling the incentivisation leavers throughout). They weren’t what he wanted to hear initially, but once he sat down and thought about what I had to say he conceded not every problem can be solved through incentivisation alone.

The report concluded that the compliance paperwork burden of many specialist products was too great for it to be sufficiently incentivised in a bank branch. The staff just didn’t have the time (nor the interest) required to do this job well.

Therefore I recommended all specialist products such as mortgages, insurance, and financial planning be referred to a dedicated team of professionals who specialised in just that one product. This proved to be controversial at first, but several years on all the other leading retail banks appear to have adopted a similar approach… because it works!

I suggested that the relative amounts of the incentives be driven by that product’s anticipated profitability of that product. They should also be aligned with the bank’s business planning and marketing activities. There was no point promoting one product while incentivising another, the conflict of interest for the branch staff was obvious.

There were a bunch more data driven recommendations, but those were the relevant ones for today’s post.

So what?

What does this have to do with personal finance?

Much like the bank branch staff, we are all easily led and manipulated.

Firms constantly dangle carrots… loyalty programmes, frequent flyer miles, and the like.

It was recently tax season, and there were a ridiculous number of posts in the blogosphere talking up the paying of (often large) tax bills using reward earning credit cards. However hardly any of those posts actually ran the numbers to compare the relative costs of the 2-3% credit card surcharge versus the value of the reward points earned.

Using my own reward card (albeit a much less generous UK based one) each £1 spent earns a reward with a monetary value of £0.005. To pay a £10,000 tax bill using that card would therefore earn a £50 reward. Earning that reward would incur a surcharge of £300.
Doesn’t seem like such an amazing deal to me.

What about an airline reward card?

Paying that tax bill would earn 10,000 reward points... again at a cost of £300.

My reward points trip to Cuba earlier this year “cost” 45,000 reward points per return ticket.

Were that reward “earned” via surcharge incurring purchases such as tax bills it would require me to “spend” £45,000 plus a surcharge of £1,350 to pay for a ticket worth £550 (excluding the taxes which I had to pay in cash anyway).

That doesn’t seem like such a good deal either.

These reward schemes are certainly not the most generous or the most smartly used, but they are certainly representative of a common problem confronting all of us.

The moral of the story?

There is no such thing, despite what bloggers may tell you, as a free lunch. Trust, but verify, advice you receive... but always run the numbers to validate whether you’re receiving sage advice or being sold some complete and utter bullshit.

Much like the mindless acceptance of the "4% safe withdrawal rate", unquestioning acceptance of the merits of "travel hacking" should also be challenged. There are definitely times when doing it can result in a win, but there are many others where the blind pursuit of loyalty points will result in a sub-optimal outcome.

  • Share:

You Might Also Like

0 comments