• Status: Solved
  • Priority: Medium
  • Security: Public
  • Views: 397
  • Last Modified:

Statistically explain the weight of sales comming from same region

Hi you all geniouses,

I have the attached set of sales. This sales are from a trading company. The company  connects vendors and sellers and arranges fleet services for them (more or less like Amazon.com). Basically I can see that there is an obvious impact on purchases (or sales) given that the vendor is local. For instance, Region 8 Vendors sell 83 thousand dollars to buyers within this region, and the second most important buying region is Region #1 with 57 thousand. Of course Region 1 is way greater in terms of population and market size, nonetheless, Region 8 sellers manage to sell considerably more to buyers in its region.

There is indeed also a big impact on sales given the size of the region itself. For example, region 1 is the greatest buyer and seller to all other regions except from same region transactions.

I want to make some smart conclusions on whether making available all the products of all regions to all other regions will have any impact.

Can you help me out, on how to structure this problem?
0
degaray
Asked:
degaray
  • 6
  • 4
  • 4
1 Solution
 
TommySzalapskiCommented:
There is nothing attached, but the sales amounts won't really tell you if you should offer goods across regions.

Does it cost anything more to ship it farther? Usually no.
Would people in region 1 want products from region 2? Probably yes.

What is the cost of "making available all the products of all regions to all other regions"? It seems like it would be very low.
Without seeing any data, I would assume that this would be a fairly easy decision. Of course, for some services maybe, it wouldn't make sense. I would say leave it up to the vendors but give the option.
0
 
aburrCommented:
Of course it will have an impact. Whether the impact is good or bad depends on factors not listed.
It seems obvious that if you make all products "available" you will sell more but are there costs to increase availability. If the availability is on the internet, probably not much.
Anything that will help a vendor increase his non-local sales will help him.
The only question is  will it disadvantage the local seller in the non-local region. I do not think you present data will permit a reliable answer to that question.
find the pattern now (you appear to have done so)
Make the change and look for a change in the patten.
0
 
degarayAuthor Commented:
@TommySzalapski here it is the file

flow.xlsx
0
The 14th Annual Expert Award Winners

The results are in! Meet the top members of our 2017 Expert Awards. Congratulations to all who qualified!

 
degarayAuthor Commented:
@aburr: suppose there are no costs to make them available (or at least they are not to concern about)
0
 
degarayAuthor Commented:
I know it is a straight forward  decision, I am willing to know at what cost or how much statistically does it impact on sales?
0
 
aburrCommented:
The more availability, the better.
0
 
aburrCommented:
You will only find the impact by doing it and then looking at the before and after data
0
 
degarayAuthor Commented:
@aburr: Well, that implies some investment. Of course nothing is free, I would like to have a smarter answer prior to investing on this. I know I am not going to have a perfect answer, I want to have an idea so that I can forecast sales.
0
 
aburrCommented:
"I want to have an idea so that I can forecast sales."
If you change the conditions, you cannot reliably forecast the results, unless you have a history of the changes or an example of what somebody else did.
0
 
TommySzalapskiCommented:
Looking at your data, almost all of the vendors sell less than 25% of their products within their own region.
In fact, overall, only 23% of the sales on the sheet are from the diagonal. So 77% of total sales are where the vendor and buyer are in different regions.

According to that data, making products available outside the region more than quadruples the sales even with the current system.
0
 
TommySzalapskiCommented:
I agree with aburr though. You need data to do forcasting. If you can find a corporation that made a similar choice, you could extrapolate from there, otherwise it's sheer guesswork. But all the guesses point toward higher availability if it comes at a low cost.
0
 
degarayAuthor Commented:
I have a shy idea, on how but I would like to hear a more statistical (maybe multivariate approach) just to see how much does availability locally and globally affect.

My idea is, that if you consider two regions, say #1 and #2, you would expect that both will likely be similar in terms of sales to regions different from #1 and #2. I mean with a similar proportion of sales to non #1 and #2 regions compared to the whole sales of each region.
0
 
TommySzalapskiCommented:
Well, you could do some kind of linear regression but you still need variables for that. You could include the size of the region as well as anything else that could matter and see if you can get an estimate of the proportion of in-region sales vs. total sales. A more simple statistical approach such as looking at the average should be enough to convince people that it's a good idea.

In fact, sometimes using words people understand like "average" instead of "linear regression model" are more convincing (though less rigorous mathematically).
0
 
degarayAuthor Commented:
The answer does not explain how to do it. It was a coffee chat about what else could be done, but no one answered the question.
0
Question has a verified solution.

Are you are experiencing a similar issue? Get a personalized answer when you ask a related question.

Have a better answer? Share it in a comment.

Join & Write a Comment

Featured Post

Keep up with what's happening at Experts Exchange!

Sign up to receive Decoded, a new monthly digest with product updates, feature release info, continuing education opportunities, and more.

  • 6
  • 4
  • 4
Tackle projects and never again get stuck behind a technical roadblock.
Join Now