Sold out?

This is really NOT how eCommerce should work: You are ordering on a wednesday evening, you are hoping that your goods will be delivered on friday or latest by saturday, and then you are receiving an email on tuesday (!!) that they are sold out.

This is what happened at the german shoe retailer GOERTZ, as you can see on their email here:

There are two topics within this case:

1. You should not treat your customer as an idiot. You should not excuse yourself by telling your customer that there are rare cases where products get "suddenly" sold out, if you are sending your article-is-sold-out message nearly a week after you have received the order.
   
2. You should integrate your backend IT better to your shop infrastructure. You always compete with Amazon regarding service and response time and Amazon is damned good at that.

I think retailers have to learn that in eCommerce your online shop is what your physical store is in stationary retail. If you are not investing into your stores to make the customer experience as good as possible you will loose. 

Btw. that is what lately happend to Schlecker here in germany. They missed to keep up with their competitors, lost more and more customers and in the end failed.

And how does this whole case relate to master data management? 

Actually "stock on hand" typically is not considered to be master data. This shows that introducing master data management also implies to look into your IT infrastructure how you are doing integration for your IT systems.

And the requirement for an online shop is typically a "near real time integration". This means changes from all sources for the online shop have to be propagated to the shop nearly in real time. Batch jobs which run once a day are not acceptable!

And your master data system should be one source for your online shop.

What are the Positives and Negatives of GDSN?

Lately I received the question whether I could quickly summarize the Positives and Negatives of GDSN.

Guys, I can tell you - that is not an easy task! So I tried my very best and this is what I came up with:

Positives:

1. GDSN is the only global infrastructure to exchange product information
2. It is used by many global suppliers and retailers
3. It has a lot flexibility build in with the extension concept and extensions like the AVP-Extension
4. There is a channel for communicating back from retailers to suppliers (the Catalog Item Confirmation message CIC)
5. -Full coverage for all B2B processes in the verticals of FMCG and Grocery

Negatives:

1. Implementation is complex for suppliers and retailers
2. Although GDSN defines a message standard, this is only mandatory to be used between data pools. The communication between data source/recipient and the data pool is not standardized at all. This means every MDM/PIM/tool provider has to build an extra interface for each and every data pool he wants to connect his tool to. Therefore there are right now not that many tools which have build in GDSN capability.
3. Typically GDSN does not cover all the requirements a retailer has regarding data synchronization. Therefore retailers are often asking for additional product information from suppliers on different ways. Very obvious is that GDSN today does not yet cover B2C data. And the question is whether the current GSMP process really can master the challenges of B2C data (category specific attributes).
4. GDSN Pricesync is very complex and there is only one community which has adopted it (australia) up until now.
5. There is a huge reengineering of the message format under way (Modular Item / GDSN 3.0) to give GDSN even more flexibility. There is not yet any date for putting that into production (rumours are saying, maybe 2015!?). This will put a lot of burden on all data pools and probably also on customers using GDSN today.

What do you think? Is that a fair summary? Do you have any extra Positives or Negatives?

Looking forward to your comments!

Cap Gemini: MDM in 2012 still top IT topic

Do you remember my blog post from nearly a year ago? There I commented on Cap Gemini saying that MDM is one of the Top 5 IT topics for 2011.

Recently Cap Gemini released a report on the top IT topics for 2012. And there MDM and Data Quality Management are both top topics for 2012.

What does that mean? Actually it demonstrates that master data and its quality is key for sustainable IT. Therefore Master Data Management and Data Quality Management will continue to be key topics for IT and business as long as IT exists.

And my comment from a year ago that MDM is not only an IT topic but has to be driven by the business is today even more valid than a year ago. So let's continue to work on these topics!

Amazon is struggeling too with product information quality ...

... I am always very impressed by the product information and esp. by the quality of the product information Amazon does provide in their shop system.

But even Amazon seems to struggle from time to time with the data quality of their product information.

Look at the following screenshot which I have take today while I was shopping for  a landline phone.

For those of you who are not familiar with german here is how it should be written:

"Plug&Play AnrufbASINtworters:" -> "Plug&Play Anrufbeantworter:"

"Bedienung des AnrufbASINtworters:" -> "Bedienung des Anrufbeantworters:"

"AkkuladASINzeige:" -> "Akkuladeanzeige:"

"SignalstärkASINzeige:" -> "Signalstärkeanzeige:"

Got it?

Somebody has done a kind of global "Search&Replace" and has replaced "ean" by "ASIN". Actually EAN is the former name of the GTIN and ASIN is the name of the Amazon article number. Within Amazon you can convert the GTIN to the ASIN and vice versa.

So what does this example tell us:

1. Amazon does obviously not do any checks on the provided features. They obviously do not classify feature names as real attributes and they are obviously not trying to standardize those. You have also not product comparison in their shop.
   
2. Philips either has a manual process to provide their data to Amazon and in this process somebody manually did the global search&replace or they have an automated interface and a very bad testing process. Because then this has to go wrong in their interface implementation.

How could this situation be avoided?

1. Amazon could predefine all feature names and force all suppliers to deliver their data accordingly. I think this is not a feasible approach because Amazon would have to standardize all attributes on all product categories they have plus they would have to force all their suppliers to comply to that standard.
   Do you think Amazon could do this? I do not think so. Therefore I think Amazon can only rely on their suppliers that it is in their best interest to provide the best data quality they can.
2. Philips could use a PIM with defined attributes for the features of their products and implement an automated upload process to Amazon.
   I think this could be easily done and would solve the problem for Philips and Amazon.

Btw. although the data was corrupt I have ordered exactly that phone because I hope that Philips is way better in designing and producing phones than they are in providing product information to Amazon :-)