0

I am looking for a solution that gives our BI team AND our Magento system better insights into user behavior and user needs. Actually we use GA for our webshop and something else for our ERP system/online stores.

In terms of big data it would be very helpful to combine such data or even better, have one solution.

But I did not find any good solutions on market where even Magento could 'ask' the system for user specific data to present him best products that fit to the users needs (online and offline stores).

Actually I thought (from a high point of view) every shop has SKUs and users that buy SKUs with some extra parameters. So I expected that I can find tools or extensions that support something like "that might be interesting for you too" on a point of view from the customer and not "other people bought this too". Or a better example would be "you bought this product a year ago, do you need another one, don't you?".

I know there are large shops, that have such data and they use it - and I guess they all invented it by themselves, while I think 90% of such application should be the same on each shop system - so why is there no extension that can do such things?

Or are there such extensions but I can not find them?

E.g. Spryker has such BI system closely implemented to their shopping system and use Mondrian as tool to view the BI data - but I guess you have to develop the shopping part by yourself too.

2 Answers 2

1

Typical BI tools are not very well suited for such tasks as running Recommendation Algorithms. If you are going to build such logic above Mondrian, you would need to write user defined function for it and it may not work as efficient as you may expect.

Better way would be to apply such algorithms on lover level. You may apply recommendation algorithm using in-database or "in-tools" technologies for data mining as R, Python, Skark, Weka.

Lot of BI vendors support execution of R scripts. Implementing recommendation algorithm in R language should not be a problem since it has lot of ready to use packages for different data mining problems.

R script might be connected to Power BI, Qlik, Oracle, HP Vertica. There is also R connector for Magento, BIM.

The workflow with BIM extension would be:

  1. You export data needed for the algorithm into csv files.
  2. You write an R script that will consume data from csv and do prepare data for fast analytical processing.
  3. You write an R script that will apply recommendation algorithm to make suggestions for particular customer.

R scripts will be called from PHP code of Magento and results might be returned in JSON format. Examples of R scripts you may find inside the extension.

0

Because, from personal experience, this is far beyond DIY and often requires working with a software development house experienced with the 2-3 different environments you want to integrate or else, a company with an in-house multi-discipline team and a sizeable warchest to fund it.

The extensions, if available are often proprietary to the developer. One integration package that can help translate data between different systems is In-Synch by ROI Consulting which is used to arrange data exchange with for example, Sage 100.

1
  • Yes, we have multi-discipline teams and everything to build it by ourselves. But I am still curious why there is no software that covers a part of this BI part. Especially functionalities like get buy recommendations from the users buy history.
    – Jurik
    Apr 13, 2016 at 7:21

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.