To trust or to check?

To trust or to check?

Vasiliy Ivanov Nov 21, 2017

I want to share with you a small victory, which was for us a confirmation of the chosen strategy and the importance of the decisions made. 

After 4 years of the company's operations, our revenues began to plummet. At the same time, according to all the KPI, the catastrophe was not foreseen, only small seasonal fluctuations falling on the summer.

Once again, I did not invent anything new, but I was lucky enough to apply the administrative techniques correctly. It is about the key performance indicators of the company and the interpretation of these indicators with different meanings. Have you asked yourself these questions:

  • What are KPI really like?
  • Why do I need KPI?
  • What KPI do I need?



The fact is that key indicators have become a popular term nowadays. But this term is taught by fashionable business coaches, whose results in business we have never seen. You see, there are common key indicators, for example, income per month, the number of new customers per day, or the average lifetime of the client. But such indicators would occur even to a student who has set up his own startup! So why then, like thousands of other businesses, we got caught up in an unpleasant and unexpected situation?

I can draw an analogy of this situation from the experience of my ascents of the mountains - when you stand on top above the clouds, you see only beautiful horizons and other peaks towering above the clouds. And what happens at this moment below, under the clouds? Is it rainy or sunny there? You do not know the answer to this question until you come down.


In the company, such a descent from the top can be considered the leader’s temporary abandonment of strategic tasks, the visit of all departments, and daily communication with employees. But what to do, when you do not have time to talk even with key employees during the week?

Recently I wrote about the red flags, which allowed Churchill to correctly understand the situation during the war. I recall that he realized that his subordinate generals would not always report the real state of affairs at the front, so he created an intelligence department that was under his control only.


So what should be really considered as KPI?

At my level, I watched the growth in sales, the increase in new customers, and other complex indicators. Every morning I looked at them, felt glad that in the current product everything was fine, and headed for developing new promising directions. But one day, four years after the company's life and a month or two after the dismissal of key executives, the statics broke off dramatically. I even thought that I made a colossal mistake and did not see their true potential. We began to learn the figures, from which the complex indicators are composed. We studied the history of changes in those indicators and found out very unpleasant details. In a manner of speaking, I felt raped by people I trusted. More narrowly focused indicators told us that we were diving down for a year and it was not visible from above because of the speed that the product gained.

It turned out that we were in a situation where nobody in the company managed and controlled the engineers, although I believed that we had a worthy leader. In turn, the sellers confirmed the global statistics to be the least involved unit. The departments that worked closely with these two subdivisions, often could not get clear information about the timing, quality, and other figures. The application in the Mac App Store, in general, dropped to the 2.5 rating, sales began to plummet. As a result, having constructed a chronology of increase and decrease in customers, we got the following schedule:


Was that my mistake? Certainly! I unconsciously replaced the key performance indicators with the trust to the top managers. And under the trust, I hid the reluctance to understand the inner workings of one or another department and the work of its leader.


How we dealt with that

Analyzing the above figures and examining in depth the user complaints, we conducted additional customer surveys. Further, I "announced a state of danger" for those activities that did not fulfill the required functions, and took the power over certain departments in my own hands for a while. I established the release mode of product updates with weekly frequency and strict requirements for the content of these updates. After the release of each update, I measured changes in the product evaluation by users.

In addition, I personally tested a number of technical specialists and was forced to admit the sad fact - after several years, technical experts had not grown in their knowledge. They received a lot of information, but not knowledge. Many of them did not even have a full understanding of the product architecture they were working on. Consequently, we immediately began to develop the training of specialists. As a technical leader, I collected literature, allowing to raise the level of specialists in a few months and prepare them for the next cycle of training. For example, programmers were missing books on software design, UML, design patterns, and C++ (because all our products are cross-platform and we need a single product core for these platforms).

 It took us only four weeks (iterations) to return the product to the 4 rating and restore the previous sales level.


Next, we have to go through the correct steps to consolidate lessons learned from this sad experience. We plan to keep the current pace of development in the area where the order is restored and improve the directions that are still not working optimally.

That is why I consider it important for each leader to independently think of the key performance indicators of his company.