Company Secrets From A Former Director

Hi friends. As many of you have heard, I'm quitting this game and offered to spill everything I learned about companies over the last few years. Most of you who stumble on this thread will have likely already read the Companies wiki page or the Director's Handbook and might have found them... slightly lacking. Well, I'm here to try to fill in some gaps in the community's knowledge on companies.

I learned most of this by running companies for months with secretaries to learn how roles interact with one another and ultimately affect employee income. I've talked to many directors as well and helped some of them out... For those elite directors who are about to lose some competitive advantage, I do feel a little sorry. Just a little. I'm tailoring this discussion towards TV stations, but I know many of these principles apply to others as well. I won't bother going into the basic things that you can find on the wiki... This is intended to be an advanced but not comprehensive guide. I'd like to add the caveat that I am only presenting my understanding of how companies work. I could be wrong on some things here.

Essential Positions: Extremely Important!

First, you need to understand which employee positions are essential. When you hover over the Sales Executive (SE) for a TV station, you see the phrase "Without them, the company will not make a profit". Similarly, the Ticket Agent at a Cruise Line says "Without them, the company will not be able to make money". I don't know if every type of company has roles such as these, but for these two company types, these roles are so so so important. When you read that description, you may think "easy... I'll just get one of these and fill up the rest of the company with higher earning roles". No, no, no. Think like a Torn programmer here. Would a programmer write a script of the form "if there is an SE, then this company earns money"? The answer is no, and the actual equation here is simpler in form but depends on a lot of things.

Through a LOT of trial and error, I learned that the effectiveness of a TV station's SE's determines how much income each other role, such as the Producer, makes. You can think of it like this:

Producer Income = (Many other things)*(The combined effectiveness of all SEs)*(The effectiveness of the Producer)

Similarly for someone like the deckhand at a CL:

Deckhand income = (Many other things)*(The combined effectiveness of all TAs)*(The effectiveness of the Deckhand)

Many, many directors have no clue about this effect. I had a well known director tell me that the main reason for the income difference between companies with high stat employees and the same advertising budget is "random number generation"-- that some companies were simply born with a better chance of making money. That is incorrect. Instead, the SEs' drug usage, stats, or activity are most often the root cause of poor company performance. Make sure employees in these roles have high stats, are active, and aren't addicted to drugs.

You might think... Great, I'll just hire 6 SEs for my TV station, then! Well, no, that doesn't quite work either. That brings us to our next topic.

Number of Employees at Each Role

Assume that you have one employee in one role and his or her contribution to the company income is x. If you hire a second employee into that same role, both will have their income reduced by a certain fraction. From my testing, it might be around 0.85x or 0.9x each, meaning two SEs get you 1.7x-1.8x total. With three SEs, each has their income contribution reduced further. Each might produce around 0.7x, giving 2.1x total. At a certain point, adding more of a role actually decreases the total income contribution from all the combined employees in a given role. For example, I'd expect 5 producers to make you more money than 6 producers. Note that I'm uncertain on the exact fractions here, and it may be possible that these fractional differences change with different roles. With sales executives, I routinely saw marginal benefit to the company as whole when I had 3 instead of 4, but 4 was close enough that I'd say it warrants further testing. This puts our producer equation at:

Producer Income = (Many other things)*(Fraction accounting for # of SEs)*(The combined effectiveness of all SEs)*(Fraction accounting for # of Producers)*(The effectiveness of the Producer)

Role Variation

You have probably seen that some roles are just terrible. I'm looking at you, Attorney. There is clearly a fundamental "role factor" which is part of each role's income calculation:

Producer Income = (Many other things)*(Producer role factor)*(Fraction accounting for # of SEs)*(The combined effectiveness of all SEs)*(Fraction accounting for # of Producers)*(The effectiveness of the Producer)

Pricing and Company Loyalty

Oh boy, this one's a doozy. We can test that at a certain point lower prices get us more customers but lower income. However with lower prices you gain this mysterious thing called "company loyalty". Company loyalty a hidden stat. It's how some 9+ year old TV stations can run optimally charging 120k+ per contract (daily income / daily customers) to get to the top spots, but new TVs see their income tank at 100k. We know from other posts that company loyalty is gained slowly over time, starting with low prices. I suppose if you want a top, completely optimal company some day, start your price below the RRP and slowly increase. With all your company-related educations completed, you should be able to start somewhat above RRP. And how slowly should you increase the price? I'm not sure on this one. Based on the age of these top TVs, I was guessing you could conservatively increase $99 per week for TVs and not run into trouble, but I really have no idea here. This takes months/years to test properly. Better yet, buy an old company that you suspect has high customer loyalty.

Producer Income = (Many other things)*(Contract price)*(Company loyalty factor, which is a function of price)*(Producer role factor)*(Fraction accounting for # of SEs)*(The combined effectiveness of all SEs)*(Fraction accounting for # of Producers)*(The effectiveness of the Producer)

Advertising, Marketers, and TGP

We all should know how advertising works by now. It's not how much advertising you have that matters, it's what rank you are of all others who are advertising. Since that's the case... why do so many people set their advertising budget to exactly $1,000,000? Why not $1,000,001? Think about what others are doing here, or better, test/ask around to see where there are clusters of advertising budgets and go just above it.

TGP will modify your advertising budget in a fixed way (likely some multiplication factor... 1.5x? 2.5x?), and Marketers will modify your advertising budget in a way that depends on their effectiveness for that day. From my brief testing I would guess a standard 5* marketer would get you somewhere around 1.3x-1.8x your marketing budget, with multiple marketers having lesser effects via the same mechanism above.

Producer Income = (Many other things)*(Advertising rank factor)*(Contract price)*(Company loyalty factor, which is a function of price)*(Producer role factor)*(Fraction accounting for # of SEs)*(The combined effectiveness of all SEs)*(Fraction accounting for # of Producers)*(The effectiveness of the Producer)

Employee Effectiveness

First, many directors already understand that employee effectiveness can go above 5* (above 100%). It has been widely circulated that employee effectiveness maxes out at 150%, which is supposedly achieved by having an employee with double the stat requirement set for their position. The equation goes like this: up to the listed stat requirement, an employee's stats linearly contribute to the employee reaching 100% effectiveness, and beyond, the incremental gains are half as valuable but still linear until the employee reaches a maximum 150% effectiveness. To test this, I ran 9 programmers for one day with two sales executives and two secretaries. All programmers were active within the last 12 hours, and none of them had any kind of drug addiction. Look at the programmers' income contributions below. We would expect to see a linear trend when plotted against [I (up to 66k) + 0.5*I (66k to 132k) + E (up to 33k) + 0.5*E (33k to 66k)] if conventional wisdom held true. A linear function does not fit the bill here, and the cluster of values at the high end throw shade on the "stat cap" theory as well.

[image: i.imgur.com]

Long story short, a logarithmic function with half weighting on the minor stat fits the data much better.

[image: i.imgur.com]

However, we still see some oddness around the stat cap. Trying again without a stat cap makes things look even better (below).

[image: i.imgur.com]

From this data I have concluded that an employee's income contribution follows employee working stats according to a logarithmic function, and that there is no working stat cap. There are obviously many implications to this, but rather than spelling them out directly, I'd like to let directors do the math on their own.

I have a very very strong suspicion that number of days worked in a company is another minor factor at play here. Basically, I think that the number of days an employee has worked in a company determines if they fall above or below the trendline. See below, the same plot with # of days each employee had been in my company.

[image: i.imgur.com]
Conventional wisdom states that an employee reaches maximum effectiveness after 3-10 days, however from a game programming perspective it would make much more sense if there were something like a log(# of days employed + x) factor built into the equation. More data would be required to prove this, of course.

Inactivity

The number 1 answer to "Why did my company income drop today?" is employee inactivity. This is one of the few things we have a pretty clear understanding of– we learned from Patch List #104 that Chedburn "Changed inactive employee effectiveness to decrease to 0% over 10 days after the first day of inactivity".

Staff Rooms and Environment

Currently, staff rooms are not worth fully upgrading (at least for TVs). This is because it is very, very difficult (if not impossible) to get your environment up to 100% without a cleaner. And a cleaner, once you have one, is very good. A cleaner is better than multiple staff room upgrades-- a cleaner is a very easy way (and possibly the only way) to get to 100% environment when your company is full of employees. Even having your staff room upgraded further than your company size will not make it easy to hit 100% environment without a cleaner. Just a note, I remember reading a note from Chedburn somewhere that indicated staff rooms might be more worth it in the future, so all of this might be out of date soon.

Now, a question that is not asked very frequently: why should I care about environment? What does environment ACTUALLY do to my income? Unfortunately, I don't have an answer here. This line from the the director's handbook is somewhat maddening: "A low environment can lower income by up to a third." Does that mean 0 environment mean the company makes 66% of its max income for that day? And that 100% environment = 100% max income for the day? Based on everything I've read, environment is a separate factor that affects company income generation, rather than just an indicator of how things are going (as popularity is). I don't have a good way of testing the effect of environment from my current data sets, but perhaps this text might inspire someone else to test this.

Inventory Management

I generally avoided companies where I needed to actively manage inventory, but I did briefly take an adult novelties company to the top earning spot before selling it. The secret to ANs is to sell half of your inventory capacity every day and to never run out of your highest revenue items. Stocking out of an item means you lost an opportunity to sell more of an item on that day. Ideally the daily demand exactly meets your supply, which is half of your inventory capacity. And since you can tailor the demand of each item by adjusting the price, you have complete control of the demand/inventory equation.

It takes 24 hours for ordered items to be received, meaning you have to predict how many items that you will sell the next two days in order to know how much of everything you want to buy right now. Spreadsheets are your friend... eventually you will settle on an optimum price which sets the demand where you want, which will allow you to order a relatively consistent amount of each item each day.

Final Remarks

This concludes my thoughts on how I think companies work. Again, it may be possible that I have a few things wrong here. Hopefully some of you find this helpful!

Running a company would more likely be an end-game play for me i.e. my opinion means nada but still thumbs up on one of the more interesting (and hopefully value-adding) sharing on managing a company.

It's not bad, he really doest get advertising and company loyalty but the rest is fairly solid.

Finally an interesting read about companies!

So any chance you will spill the beans?

Can confirm the Sales Exec stuff, if you have a book reader the rest of your employees income jumps high.

Not yet

Wow thank you for this.

I read this in John suchet's voice, (that's a compliment) keep writing :)

Thanks! I have never been into companies, in fact i have never owned one. But this is a thread worth reading for anyone who wants to learn more about it!

Thankyou, great to have some more insights from somebody with extended experience.

R+

what a loss that this guy quit :-(

Pretty decent explanation for new players to understand before they go making a company =P

Bump great post