I would like to know some common (or not) applications of OR in a E-commerce (meaning "electronic services").

As a hint I would say: Use Bin Packing models to organize the ads/banners in your website.

asked 13 Apr '12, 17:58

Florents%20Tselai's gravatar image

Florents Tselai
600516
accept rate: 7%


Marketing Optimization

This isn't limited to E-commerce, but certainly includes it. When you have lots of campaigns, it may not be feasible or desirable to target all of your customers with every campaign. Marketing Optimization (MO) is choosing the optimal assignment of campaigns to customers.

You usually have a set of "contact rules" that specify how often customers may be targeted with various types of communication (eg "don't email more than once per week, except important notifications") - these are constraints. You may also have channel constraints (can't send more than 10000 letters or make more than 50 calls per day). You also have a propensity for each customer to respond to each campaign, and an expected return if they do respond - these are at a per-customer level and come from prior modelling.

Naive marketing departments tend to either spam everyone for every campaign, or simply send everything to the best 10% of customers. Even with contact rules, the situation can happen that a customer is not sent a campaign that they would be great for (high propensity to respond and expected profit) because they had recently been sent a previous campaign that they may not have been as suited for. MO lets you consider your whole planning horizon and optimally assign campaigns to customers.

It can dramatically reduce costs and increase response rates.

Note: all this can still apply if you are a purely web-site-based business. Emails, and website ads are your channels, and these are constrained. Choosing which ads to display to each customer could be done this way, although I don't know if it is or if they use simple keyword-type matching. I'd hope ad companies do some sort of propensity modelling.

link

answered 13 Apr '12, 19:54

DC%20Woods's gravatar image

DC Woods ♦
4.1k22546
accept rate: 5%

edited 13 Apr '12, 20:03

  1. Analyzing customers' behavior and data using data mining techniques, which are highly rooted in optimization and statistics, is a great example. A specific example would be the recommendation systems used in many e-commerce website where they bundle and offer products based on a single specific customer's past behavior along with information of multiple other customers with similar interests.

  2. Near to the @DC's great example, there is devising pricing strategies for different services belonging to a specific product (e.g. Amazon's multiple delivery dates with highly different prices is a good example).

  3. Distribution network design and optimization for e-commerce businesses offering product delivery is a good example rooted in supply chain management. In addition, inventory control is very important for such businesses.

link

answered 13 Apr '12, 21:53

Ehsan's gravatar image

Ehsan ♦
4.8k31122
accept rate: 16%

Similar to Ehsan's second point, there is dynamic bundling of goods. For instance, if I order "Drifters - Greatest Hits" from an e-tailer, they may offer to bundle "The Coasters' Greatest Hits" at a discount. This is tailored just to me, based on some combination of what other people have purchased, my other orders, or their managing to hack my Pandora play list. :-)

There's also quite a bit of math/OR behind online auctions, although I can't really say just what.

link

answered 14 Apr '12, 10:03

Paul%20Rubin's gravatar image

Paul Rubin ♦♦
14.6k412
accept rate: 19%

I second the online ad auctions point. One of many interesting papers on the topic is,

http://people.ischool.berkeley.edu/~hal/Papers/2006/position.pdf

The anual Ad Auctions Trading Agent Competition (TAC:AA) has also generated a number of interesting results in that area.

(17 Apr '12, 01:02) Carleton

Optimal pricing strategies come to mind. Objective would be to maximize revenue with your customers. Constraints would be ad budgets, customer demographics or segmentation, demand, costs of business. Here is an example paper from INFORMS "Optimal Pricing Strategy for New Products" (Krishnan, Bass, Jain)

link

answered 20 Apr '12, 08:45

larrydag%201's gravatar image

larrydag 1 ♦
3.2k51326
accept rate: 9%

Your answer
toggle preview

Follow this question

By Email:

Once you sign in you will be able to subscribe for any updates here

By RSS:

Answers

Answers and Comments

Markdown Basics

  • *italic* or _italic_
  • **bold** or __bold__
  • link:[text](http://url.com/ "Title")
  • image?![alt text](/path/img.jpg "Title")
  • numbered list: 1. Foo 2. Bar
  • to add a line break simply add two spaces to where you would like the new line to be.
  • basic HTML tags are also supported

Tags:

×29
×3

Asked: 13 Apr '12, 17:58

Seen: 3,940 times

Last updated: 20 Apr '12, 08:45

OR-Exchange! Your site for questions, answers, and announcements about operations research.