PLM, Search and Findability

Searching for information is a tricky thing. Search may sounds as a simple operation, but in fact, it is translated to a complex computational, information and organizational task. Search isn’t a new problem. Lots of work was done in this domain for the last 20 years. Google clearly changed a consumer perception about internet search. Google "lady Gaga" and in less than a second you will have all relevant internet information about Lady Gaga.

However, searching inside of corporate data is different. What works for "lady Gaga" type of search, doesn’t work for MPR-345000 type of search for product information, part numbers, documents in SharePoints, emails, intranet sites and other enterprise data sources. It is complicated and daunting task. Sometimes, it even goes beyond of what is most important – it is about where to search. During my previous work at Inforbix, I found customers are usually very confused about searching information. They think about search as an easy and complicated thing at the same time and usually blame IT for not bringing a right solution to solve the problem of search.

I’ve been looking on materials coming from Enterprise Search Europe 2013 conference earlier this week. looking on Enterprise Search and Findability Survey 2013 published by FINDWISE and presented by Kristian Norling few weeks ago. The following two slides caught my special attention. It speaks about what are the obstacles to find the right information. Take a look on the following statistics:

According to the research, there are top 5 obstacles – we do wrong tagging, we don’t know where to search, tools are not perfect as well as tagging made by tools. The last one is even more interesting – we don’t know what to look for.

What is my conclusion? Information search is a complicated, but fascinating topic. For the last few years CAD and PLM vendors started to put more emphasis in order to improve their way to search and actually find the information. However, solving the problem of a single tool is still not enough. Even if company has PLM/PDM system in place, the information usually located in disparate sources. As we learn from the research, the problem is going much beyond searching for a specific set of keywords – data sources are not defined and how to search for right data is not clear as well. Just my thoughts…

Best, Oleg

6 Responses to PLM, Search and Findability

  1. David Bell says:

    I think context will play a huge role in “search” for the next generation of PLM tooling. For example, if you have many complaints about a single defect, you are the lead product manager responsible then the context of this is pretty clear – what solutions have worked in the past, who are the experts to call. You can do this with keyword searches, but context is critical.

    How you capture context and use this smartly is another story!

    We are working intensively on such a use case at the moment here at uPraxis.

  2. David, you are spot on. Context and semantic relationships will play a huge role to make search easier and precise. Thanks for commenting! Oleg

  3. Mikel Martin says:

    I think the thing to remember is that in most conversations about search is referred to as a technology. We should remember the real problem. The user needs to locate some data. Search is usually only very useful if that user has enough information at hand to know what objects will provide this information. This is something that Google realized a long time ago. Which is why they are not just providing a search tool, they are using what other people find useful to help you locate something that might be interesting to you as well. Even more interesting is when you do a search and find something unexpected but useful. What troubles me is that I am not sure most enterprise systems are good at this, but often think users just need search to locate data. Not only (in my opinion) is the idea of search foreign to the human mind, in many cases it returns to much data to be useful.

  4. Mikel, you mentioned something very important. This is not about ‘search’, but about ‘find’. Google identified algorithm that works good for internet search with the assumption that people are looking for similar stuff. Like pages with good references to “Lady Gaga” will be good for everybody looking for Lady Gaga. However, in enterprise, it is different. All searches are more individuals. That’s why it is hard. To explore semantic relationships between information is right way to go. Just my thoughts… Best, Oleg

  5. ramesh says:

    Thanks oleg for sharing…in the recent PLM connection conference they have introduced a very effective categorized structure data search for TC called AWC(Active Web Client). The demo shows a non-plm user can zero-down to the search result upto a great extent of what he is looking for. And also for CAD users they have come up with Shape based search in TC which is also and interesting topic for designer to restrain them to do a duplicate design work which is already created.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s


Get every new post delivered to your Inbox.

Join 290 other followers

%d bloggers like this: