Controlling in Internet agencies: Why recording hours is nonsense!

For a long time, I thought that the only way to run an Internet agency efficiently from a cost perspective was to have the most comprehensive controlling possible. In an agency, this must necessarily be based on the time spent by employees on projects. Gradually, however, I am realizing that what should actually be the basis for decisions on efficiency increases and improvements is downright counterproductive. Read why in the following article.

(Reading time 4 minutes)

Selling time when business value is required

In the digital economy, we have always billed our services according to time. Depending on the role of the employee (developer, integrator, project manager, consultant, etc.), we set a price per unit of time and then charge for the time spent accordingly. The employees have to work a certain amount of time, receive a salary for this defined time and thus account for the majority of the agency’s P&L costs. This seems logical and stringent in itself.

The customer, on the other hand, wants to receive a deliverable. He wants to receive a trade or a result that he can use for his business. It should be of the best possible quality and delivered as quickly as possible. It should provide him with a benefit that gives him an advantage in his own business. The deliverable should therefore have a business value. Ideally, the fee that the customer is willing to pay for the business value of the deliverable covers the price that the agency charges for the time spent on it. Everyone is satisfied.

Now, the ideal case is like the wolf in Switzerland: it rarely happens. The rule is that either the client or the agency has to lower their expectations. In most cases, agencies’ target calculations do not look like reality. The reality is usually significantly worse.

Pseudo-controlling

In many agencies, there is therefore a certain amount of movement on this point at some point. People are making each other aware of the issue, calculating what margins they would have and what they could do with the money if only they could bill all the hours booked on the project. I have rarely seen solid service controlling. This is hardly surprising, as the people who work in agency management rarely have the necessary knowledge of cost accounting. The fact that the industry is still young with its peculiarities (from the perspective of the old economy) adds to this. There is simply a lack of established concepts, as we know them from industrial production, for example. And so more or less creative and sophisticated concepts are used to at least provide a certain indication of the profitability of the projects.

I find it fascinating that the results are usually interpreted differently from project to project. In the case of projects that are fundamentally successful but in the red, all kinds of soft factors are then “monetized” (at least mentally). Projects that have fallen out of favor usually remain so, even if they are profitable. Personally, I don’t even find this unsympathetic, even if it is gross nonsense from an objective point of view, but it shows that the economic component is not always above all else.

Time recording

The mother of all cost accounting is the recording of hours. It is highly unpopular with the vast majority of employees. For one thing, it demands a great deal of discipline from the employee. On the other hand, it is only suitable for very few employees in terms of working methods. In agencies, a lot depends on impulses, on fast and slow phases. The constant tracking of time attempts to create facts where there are no clear ones. This is because the time spent is usually only related in a very abstract way to the deliverable and therefore to the business value for the customer. And this is precisely the core of the problem. The customer usually does not understand why feature A means that x hours have to be spent. Not because they can judge it. No. But because feature A has a certain business value for them. If the costs are not in proportion to the business value, the customer does not understand the effort. (Not to be confused with the “know a lot about nothing” type of customer, who simply finds everything very simple and questions every effort).

It becomes even more difficult when the time spent fluctuates. But that is exactly what software development is, only approximately estimable. Anyone who claims otherwise has so far only integrated templates. However, the customer assumes that every step is completely calculable and assessable. He is not aware of the partially explorative nature of this work. As if the (conscientious) developer wasn’t already suffering enough, now the agency comes along and crams him into the corset of time recording. The meticulously precise reports and the discussions based on them when times are bad are also great. These are all things that really good developers avoid like the plague. In any case, that’s my experience.

Wrong selection

And this recording of hours promotes false selection. If times are tracked precisely and every absence is reminded or “discussed”, the developer will start to overestimate. The agency management will then go to pitches two or three times with these estimates and realize that they are completely overpriced. At the next pitch, they will leave the developer’s estimate in the room and come up with a lower price. However, the controlling is then still based on this price and the developer is criticized again. These are the typical paths that result. An ideal breeding ground for frustration for everyone.

In addition, if working time is tracked in such detail, this usually results in a lot of overtime. The employees who need a lot of time and therefore stay in the office for a long time have a lot of overtime, those who have logged little time are seen in a skewed light. However, it is in the nature of a good developer to deliver better deliverables in less time. If you use the scale “more time logged equals more effort”, you consistently reward the wrong ones. And the good developers will leave sooner or later.

Minimum time unit: One day

What can you do about it? I think a lot. I would start by only logging working days and these days are not logged by the developer themselves, but by the team leader. That deliberately requires a few things:

1. Raise customer awareness
   The customer, especially with medium-sized and small agencies, will initially not understand that time recording is only accurate to one day. After all, it costs every hour. But if you explain to them that software development is not an exact science and that time tracking has a lot of negative effects, I have found that this is understood very well.
2. A day is a day
   It is in the nature of our daily lives that there are such and such days. Some days everything works out and others almost nothing. You also have to accept this in the project business. By simply logging all days as a single unit, you balance out the (mental) expectation that everything must always run smoothly. Thomas Layh writes about this in his blog:

   We are not machines, we only use them for work.

   The much fuzzier time tracking in days instead of hours takes this fact into account.

1. There is no overtime
   If no hours are tracked, there is no overtime. This does not mean that the employee does not have to work longer if the project requires it. As I said, a day is a day. Sometimes, you just have to work through it. On other days, you can make up for it. For agency management, mountains of overtime are anathema. And rightly so. By tracking in days, the employee is responsible for making up time as quickly as possible. Agency management is required to guarantee this freedom. The benefit is great for both employees and the agency. The employee has less pressure and stress and the agency does not have to deal with time sheets, overtime discussions, etc. And it’s less hassle for everyone.

 

Paradigm shift in the overarching sense

In my article on Labor Day, I wrote that we need to find a model in the economy that no longer links the value of work primarily to the time spent working, as the efficiency gains from technological advances have been reducing the necessary working time for decades. Our system will not be able to cope with this in the long term. Less and less people’s time is needed to generate added value.

However, people themselves are still needed. We would therefore do well to link the remuneration for a person’s work to business value as soon as possible. I already know that this is no easy task. It is no coincidence that more and more services are being offered as products or flat-rate services. The customer is simply much better able to assess the relationship between business value and the fee to be paid.

Modularization of services

In agencies with highly repetitive work, this can already be implemented quite easily by modularizing the services and also trying to create code that is as configurable and therefore universally applicable as possible in development and production. This also pushes time recording into the background, because such an approach greatly reduces the effort required for features anyway. The price for the functionality is still based on the business value for the customer.

And this value does not decrease just because less effort was required to provide the feature. And that brings us back to what we actually want in agencies: More healthy efficiency, more satisfied employees and customers and more profit at the end of the year.