Controlling what people do (part 2)
How to use money and other levers to steer people towards your goals
OK but I still don’t know how to control people
So lets say that you are appointed as Head of Legal in a large corporation and your objective is to improve their risk management practices since the company has been involved in several “small problems” recently. Projects are launched without enough scrutiny, and although compliance requirements are published in a fancy PDF on the company’s intranet… people don’t pay attention to them. So what can we do?
There are two things we need to solve: i) where do we “intercept” what people are doing and ii) what hard incentive we put so they follow
On the interception
This is harder sometimes than it sounds. Common sense says that decent companies should have (as least!) a list of projects they are running, after all, most of them consume some sort of resource like money or people’s time, but… thats not always the case. So option 1 is that they do and that list will point you to the people you need to “convince” to do a certain thing.
If the list doesn’t even exists, the best option is to “probe” a certain moment where all projects meet. For example, lets say that in order to get funds teams have to navigate some sort of approval process, fill some docs, etc. Therefore, the guys running this process, a portfolio team or Finance, are your best allies to embed your own requirements. For example, lets say that in your new role as Head of Legal you want all projects to staff a lawyer from your team to review what they are doing, therefore, the portfolio team approving a project should also ask for that lawyers signature during their review. This usually is not that difficult because the portfolio team have no incentive to put the company at risk or to approve projects that are not compliant.
Other places can be in some sort of project committee where people check on the progress of things or during go-live where there is usually some sort of user acceptance test or closure process, for example, in IT projects have to pass some checks in order to release their product to the public. It’s important to try to establish this checks as early as possible (“ex-ante”) since it will be easier for people to comply without rework. In technology there is a whole trend talking about this called “shift left”
It’s good to acknowledge that if you decide to put a roadblock on people you better be ready to act when they come for you. For example, if you want to approve all these process designs you better have a pool of well trained, sized and motivated engineers to help projects do things “in the right way”, other wise they will rebel and raise to the CEO that you are a stopper.
On the incentive
There are only 2 ways to incentivise people… money and automation.
Money can be used in multiple ways, for example, our previous case of asking the portfolio team to not approve projects without the sign-off of a lawyer is a VERY EFFECTIVE way of cutting funds to bad corporate citizens. Some companies are also releasing funds every quarter, so they can be sure they are compliant also during execution.
Another way is to play with people’s financial incentives aka their variable salary. So you can link a certain objective to that and unless they make a miserable amount of money they will tend to follow.
A corollary of money is reputation, for example, a long time ago I worked in a factory that published the time people spent working (calculated with the turnstile) in the company’s intranet, so literally everyone knew how many hours you sink at work. Very evil but worked because people think that a good reputation leads to money and promotions. So for example, creating a public dashboard about anything will create this peer pressure effect where people dont want to appear at the bottom showing they are not doing something right. Lets say you have a metric to measure team’s morale, who would want to have his name in the bottom of that list?! (some people but thats another story)
Companies also going through an operating model transformation need to think about money as the main incentive. Lets say that your big corporation is shifting from an local model where every business unit do things their way towards a homogenous global way. You have two alternatives, the nice democratic way where you beg local managers to please listen to how you want to do things or you take control of their budget (and people, even better). It’s common sense, like when your dad didn’t want you to do something as a kid by not giving you any money, and yet many managers because of politics, fear of accountability or inexperience choose to play the democratic way thinking maybe a consultant will give them a nice trick on how to solve this without having to go through those nasty budget discussions.
“Although painful, its easier to just kill the elephant in the room”
The second incentive is no incentive at all, is automation. If there is no way to skip a control then you need to live with it. There is a cool case in Google where all processes, for example when you click on Gmail to send an email, need to run under a certain time, otherwise stopped automatically. This is great because is forcing engineers to ensure their code is fast.
A more human example are employee expenses. Instead of paying for the corporate credit card of an employee and then waiting for their expense report, what most companies do is link the card with the employee’s personal account so they are compelled to do the expense report. This is an evil model that works perfectly, no begging for people to do reports, they just have to do a lot of paperwork to even out and get their personal money back and any mistake will be on them… brilliant.
A cool side effect
Intercepting the flow and incentivising people also allows for a major change in a company’s culture, you can start managing by exception. For example, if you create 10 requirements projects need to follow, any project that proves they are compliant can move ahead automatically. You can even create the time to market incentive to follow the standard instead of reinventing some greenfield way of doing it. On the contrary, anyone that doesn’t follow the standard needs to ask for an exception and prove its needed. Lets say you, as CIO of a company, declare you want all your applications to be coded in X programming language from now on… 95% of projects will have no valid reason not to follow this rule (otherwise no money for them!), but 5% will actually bring good reasons why they need to skip this rule, so therefore only 5% of project will require additional analysis and consume valuable time form your team.
Surely, the most difficult part of controlling any organization is Human Resources. The normal individual differences between people are a natural factor to cause this difficulty. Things that are great for some are simply "unacceptable" for others and so on. Even incentives can cause havoc sometimes, when those in charge remain "surprised" and cannot understand what's going on. People are complicated and personal choices cover a vast spectrum that can be very difficult to understand by the manager implementing control measures.