Visual Entertainment and Technologies Forum

Recently i did some testing to see the effect of price on the sales of vehicles in Gearcity.

The main question i was looking at:
Do you see a increase in sales volume, perhaps a significant one when you lower prices?
The reason i wanted to know this was i believe it was suggested at some point on the forums you could increase your price without having a significant effect on sales.

Preface:
I don't have an enormous amount of time to do this at the moment, so ive only got testing from two different price points and i haven't repeated it yet. I also had a lot of bad luck with bugs and had to do the final part of the testing on 1.11 instead of 1.10.
So you should consider the information very rough estimations, but i think they are still interesting for discussing the topic.

Ill start by summarizing what i think ive found out, if you want to see the procedure, limits and raw data, skip to the testing section.

Basic conditions: (again look at the bottom of the thread for more details.)

• I used a sedan for every test, selling in 1901 from Feb - December in London only.
  
• For reference, the Per capita at this point is around 380-390, population is 6.32mil
  
• (Note: Competitor activity is reported, but is probably not reliable at the moment due to bugs)
  

Results/Summary

• Sedan Selling at $1000 in London - 2.6x per capita
  Average sales were ~ 376 vehicles per month. Lowest was 220 and highest was 506.
  Combined Competitor activity was never beyond 150 units total per month, but on average around 40-70.
  
  
• Sedan Selling at $600 in London - 1.55x per capita
  Average sales were ~ 995 vehicles per month. Lowest was 661 and highest was 1503
  Combined Competitor activity reached 200 towards the end, but was on average 100-130
  

Thoughts/conclusions based on this:

You should consider this a (very) rough estimation, but it looks like the drop in prices from 2.5x($1000) to 1.5x($600) per capita in 1901 gave roughly 3x the sales.

So answering the question asked right at the start of the test, in controlled conditions it seems we get significant increases in sales volume when you decrease price. We can also assume that the reverse, that in controlled conditions you get decreased sales with a higher price, is also true.

Ive suspected for a while that because of the way income distribution works in Gearcity, if you have a lower price you will be able to sell vehicles to a much larger percentage of the population. (i.e. people with lower income who could not afford the vehicle before can afford it at a lower price)
That said, i assume a lower price will also attract more people from the demographics higher priced vehicles sell to as well. (i.e. more rich people might also buy if you lower the price)

Regardless of the mechanics involved, reducing the price has quite the effect on sales by the look of it.
Selling at low price and high volume is something to think about when you design your next vehicle.



Testing Procedure:

To limit the number of other factors that can influence sales, ive tried to sell the same vehicle, under the same conditions in separate saves, in 1901. Without doing this there is so many things effecting sales its hard to tell anything with any amount of certainty.

Just for clarification, i had to edit my money in the first game turn in order to not go bankrupt while testing sales. This should have no effect on the actual sales numbers.

In detail:

Here are screenshots of each of the vehicles in 1901. You can see that while their names are different, they appear identical.
http://i.imgur.com/0qxkNux.png
http://i.imgur.com/DrNde4Q.png

Sales Data:
Ill post the raw data in code tags below this (its large).


Ive got two folders full of saves for each month, for further information on the models and further comparisons/proofs/evidence if needed.
I can also provide any other charts or information if people would like them.

Yay, that's great, thanks for sharing!

As a matter of facts, I would require you to tell us more about your margin profits. That is, did you have better results (well, let's say more money) using the highest price or the lowest?

I could have made the maths, but still, I wonder what strategy is the best to win easy money, let's say for those who do not know how to edit starting money...

Anyway, again, thanks for testing and for sharing, much appreciated!

(01-18-2014, 09:37 AM)Laurent Tommy Flanagann Wrote: [ -> ]Yay, that's great, thanks for sharing!

As a matter of facts, I would require you to tell us more about your margin profits. That is, did you have better results (well, let's say more money) using the highest price or the lowest?

I could have made the maths, but still, I wonder what strategy is the best to win easy money, let's say for those who do not know how to edit starting money...

Anyway, again, thanks for testing and for sharing, much appreciated!

Thanks, i don't mind at all.
I was interested to see the effects of this kind of test myself.

Edit: I wrote some reasoning for the test, but i decided to put it in the first post instead.

On the profits thing
Sorry, but I don't think this test can answer your question about which strategy is more profitable.

Unfortunately, this test was really only narrowly designed to show differences in the volume of sales with different prices in controlled circumstances.
I payed no attention to how much the design of the test would cost, i was only trying to limit as many variables as possible(which is quite hard). I list many of the most costly of these at the end of the post.

Unsurprisingly, these controlled conditions with no concern of profit/cost meant i lost hundreds of thousands over the year of my test. (which is the reason i had to edit in the first place)
I don't think that invalidates the test because it was only ever designed to show the differences in sales volume with a change in price. It was not in any way trying to control/test profits, afaik they don't effect the volume of sales at different prices.

So i think profits at different prices probably needs another entire test to itself with its own controlled variables.