Extending the DDM modules in Liferay with custom field types isn’t that easy. This blog post is the result of our attempts...
How to turn Search into Find #interestingreadJelle Van der Linden
I don’t think I have to tell you that looking for content and documents on the world wide web is different from looking for something within a company. Each time we talk about Search with our clients, it comes down to the same main goal: “we want to be able to search in our intranet as we search in Google”. Actually they mean: “we want to be able to find in our company how we find in Google” 😉
Mostly our answer is not what they want to hear.
The two main reasons why Google Search is so efficient ?
- Google has invested a lot in optimizing their search.
- Others have invested a lot in making their website, content and documents BE FOUND by Google.
Finding a solution with decent search features is not impossible these days. The mid and high range tools all invest in Search as the user experience on the web has also improved. They know that users can’t work without a decent Search tool.
But the second bullet is something that a lot of companies take for granted. They think that a new and optimized search will do the magic trick. And although it may already create some difference, it will not be comparable to the Google Search Experience.
Combining Google Experience and #search
Did you ever use a #hashtag in a search on Twitter, Facebook, Instagram or other social platforms? Did you notice any difference between a Google and a # search?
I’ll continue by using an example with “Company Culture”.
I searched for “Company Culture” on Google. And these images were the top hits:
I also looked for #companyculture on Instagram…
Do you notice the difference?
Google Search will give you calculated results: on which site can I find images with “company culture” in its name or metadata?
Instagram will give you personalized and human results: which images were TAGGED by a user as important for the subject “Company Culture”?
Okay, now I can understand that within a company you would probably look for the first set of images. But what if you could also look for more than just the metadata information on your content and documents?
Wouldn’t you like to scour your company data for #difficultclient #bestslidetemplate #importantforsales #needtoknow #interestingread, … ?
Improving Search in your company
It all starts with knowing what there is to find and how your users search:
- Identifying content types and user groups will get you a long way.
- Define synonyms and word clusters.
- Talk to your users to find out how they would like to look, eg. “I want all the maintenance contracts of Client X between 2001 and 2005”.
- Try to locate where your users store their content and documents. Okay, it’s kind of an unrealistic quest as users will never tell you that they have 200 documents stored on their local machines or on their private computer. But if you get some to cooperate, you can get a good view on data storage and content flows.
Enterprise Content analysts will be able to help you sort those things out. They will map the results in a structured report with a personalized approach and the next steps.
Once you know how to tune your Search engine with the above information, you should also consider the following.
Centralize and integrate
How wonderful would it be if you could just find the right content and documents with one search. This would be a dream come true. In real life you often have to scour in several applications: CRM, ERP, Intranet, email, DMS, S:Drive, your own laptop, … And if all these apps give you no result, there’s the paper contracts in your legal department if you’re lucky.
So here’s my advice to you: centralize and integrate. Try to evolve to one central content repository. Platforms, such as Alfresco, are build on open standards and can act as the central content store for all your other applications. So regardless of the source application you are searching from: if you have the rights to find the document, you will find it.
If for some reason the above scenario is impossible due to compliance or security reasons, go for a federated search. This is a retrieval technology that allows the simultaneous search of multiple searchable resources. A user makes a single query request which is distributed to the search engines, databases or other query engines. The federated search then aggregates the results in one single search result.
Dress your content in information
As mentioned in a previous blog, metadata can create a world of difference in guiding your content and documents through workflows and provide you with data to report on these workflows.
But metadata will also help in making your content findable. The more information your search engine can find about the document, the better it will be able to classify it and show it in the search results for the right key words.
So investing in a metadataplan is priceless.
Starting with pushing your documents through OCR and a classification engine is an investment worth considering. Your search results will improve rapidly, which will take a user less time searching AND you relieve the user of having to input the metadata manually. And to be frank… the classification engine will not have a bad day on which he doesn’t feel like wasting time inputting metadata
What if you could profit from all the information and content on the world wide web, without having to look for it. Have you ever heard of Semantic Search? Semantic Search tools will improve your results by understanding the searcher’s intent and the contextual meaning of the search query.
For example: try to look for “Paris” on Google.
You will get:
- a number of hits about the capital of France
- advertizing for hotels in Paris, France
- a few hits on a town in Ontario, Canada
- and various sites which can give you all the dirty details on Paris Hilton
But what was I looking for …
Semantic Search will drill down the search results based on a deduction from all search queries, my search queries, clicks in search results, … These results are combined to estimate the searcher’s intent.
Allow human identification
Apart from metadata the #search, as mentioned above, can also improve your search experience. It will create a whole new search and find experience, ideal for knowledge transfer and experience management.
Although lots of companies are doubtful when it comes to allowing personalization and giving users the rights to create their own tags (the more formal #use within corporations ), your employees could really benefit from it.
I have to admit, this is not going to work in every company or even in every team within the same company. It requires a change in mindset and in #companyculture.
Does it only work in companies with younger employees, the so-called digital natives? Probably not, but a prerequisite is the wish to share. And this is something that Facebook, Twitter, Instagram and other social platforms encourage.
But if your company culture encourages Agile teams, knowledge transfer, peer-to-peer coaching, personal growth, teamspirit, after work drinks, … Hashtagging content will be supported by the teams.
Won’t there be a rapid growth of the number of #’s within your company? Probably…
In the beginning everyone wants to be creative and think of their own funny #’s. But once certain tags will start to get popular and the searches come up with good results, people will start tagging their documents according to the same principles.
Same as with metadata, #hashtags can be tracked and used in reporting.
While metadata will give managers insight in how many documents are in a certain status or step in a workflow, hashtags can give insight in interests and hot topics.
Wouldn’t you like to know how many hits you get on #difficultclient, #needsimprovement, #wip, … ?
Combining the best practices
The era where Search was a click and install is way past us. If you want to provide your company with the right search tools and the right steps in getting to the dream scenario #onesearchfind, get some help!
Content specialists will be able to detect improvements and quick wins and advise you on the best approach in infrastructure change, but also in change management.
If you give us insight in your search struggle, we can turn Search into Find!
Interested in knowing more? We’re just a phone call away.
Blog written by Jelle Van der Linden
Digital Flow Consultant COIN (ACA IT-Solutions)
- Wikipedia: Semantic Search
- Blog: Semantic Search
- Wikipedia: Federated Search
- Blog: Metadata Management
- Blog: Taxonomy and Tagging
- Product: Alfresco