SharePoint Search frustrations: what is costing you the most and how to fix it
Let's discuss how we can improve the search function in SharePoint, saving time, money and reducing frustration.
There's some startling statistics on just how frustrated office workers are with the time it takes to find information in their internal systems.
What the survey data shows about employees search frustrations
Some surveys show that office workers spend on average between 20 and 30% of their total work time searching for information.
A surprising 62% of employees say that they spend too much time searching.
Obviously the number is even higher for some classes of employees, like knowledge workers, who need to use SharePoint search even more. Those are highly paid employees who feel that they're wasting their time searching for information that they are having difficulty finding.
A shocking 43% of employees surveyed would consider quitting over not getting efficient access to information through their company’s systems and for younger employees that number is well over 50%.
An overwhelming 79% of employees have a negative opinion of their company's enterprise search systems.
Traditional enhancements to SharePoint search (metadata and templates)
So what are some of the solutions?
Out-of-the-box there are two or three different types of search systems within SharePoint but none of them get really high reviews. Part of the problem is that people use Google outside of work and they have certain expectations based on what Google can give them. Those expectations are not met by SharePoint search even after enhancements. What you get is a Google-like search query, but not Google-like results.
1. Metadata
If you talk to experts in search they're going to focus on certain aspects of search. Most of the time they're going to focus on metadata. Information Management people just loooove metadata. Knowledge Management and Records Management people do too, as do consultants. The idea is that if you have enough metadata then search should be easy. It all goes back to the old days of librarians and indexes where all sorts of metadata was designed for searching and you had card catalogs and library computers and you could find things if you only used the correct metadata. Then you could craft the correct search query to find the information that you're looking for. After the results come back, SharePoint also has metadata-based refiners to further narrow down the query.
Mind you, metadata is a wonderful thing. It has lots of uses, like applying business rules and processes, helping workflow, enabling collaboration and interoperability. Governance and compliance are crawling with metadata. And it’s also useful for search.
The limitations of metadata:
However, metadata search requires skills that not everyone has, and it puts the onus on people who are expected to provide the correct metadata for future searches. When you're generating or receiving content, like documents, emails, and so forth, there's only so much metadata that people will volunteer, and this is key: the people who are generating content generally have different objectives from the people who in the future are going to search for that content. Even if it’s the same person, past you wasn’t considering future you. The metadata provided will be what is useful to them at the time, it will not be what is useful to people doing search in the future.
2. Scope of search and refining search
Many search consultants will spend a lot of time doing things like scopes of search and refining search either before the search results or after they come back, and adding things to search and having more complex search by a different criteria. Those are all good things if what you need is to narrow down what you're looking for, but does that actually save you time? It might help, but it's not the entire solution.
3. Result format enhancement
A lot of other search enhancements will also focus on what the results look like. Again, formatting the search results requires a lot of metadata. Should search display templates be formatted as nice little cards with things in predictable places? For instance, if you're searching and you're returning people as a type of content, certainly having some cards with the person's picture and their name and their title and contact information all in nice predictable places, that will be helpful to know who that person is and whether that's the person that you're looking for. However, not all types of content lend themselves to this sort of formatting to help you narrow down at a glance exactly what it is that you want. Traditional search systems stop short.
The limitation of traditional search solutions
1. Only covers fetching results. What about sifting through results?
One of the big problems with most SharePoint search solutions is that they stop at the point where the search results are displayed on the page. It is viewed as an IT problem, so when the computer has found and ranked and displayed all the “hits”, the programmers and consultants say “OK we have solved the search problem” and put on their jackets to leave. But of course if you are an information worker who's doing search, you know that generating the results page is only the first few seconds of the search and then you are faced with having to look through each of the documents (hits) one by one to determine whether the information you're looking for is actually in one of them. Searching individually within each of the documents is the most time consuming part of the search.
2. Most answers are in the text
One of the big problems is that what you are looking for is probably in the text, not in the metadata, and the display of hits won't have a lot of information other than the metadata. It will have the document title, the document type, and it may have a small snippet, but SharePoint’s snippets are not very good. Generally what you have to do is open up the document and when it displays you have to search again, using the document viewer software’s search function. But the search function within the viewer is not at all the same as the one that identified the document in the first place.
Search in SharePoint is actually quite smart. It does things like finding the singular plural of a term, or different tenses and genders in different languages and it knows capitalization. It can detect the language and find variants that are correct for that language. It can be set up to recognize synonyms, like “Vice President” when you’re searching for “VP” If you have a multi word query it can tell which of the words are important so it will end up finding good documents even if they don’t contain the exact words you were looking for, and ranking them using sophisticated algorithms.
But, none of those algorithms are available when you open a PDF and you’re in the PDF viewer, or when you open a Word document and you’re in Word, or when what it found was a page and the page is opened in your browser. All that you have when you search that second time is the exact match of what you type, because that search function is not that sophisticated.
So most of your time is being wasted after the results have been displayed when you're having to read through all of them and get increasingly more frustrated that you're still not finding the information you want.
Solution?
PointFire Search Summarizer tackles the time wasting
This is where the PointFire Search Summarizer comes in - to focus on that time after the search results have been returned and to help determine whether one is a good match i.e., whether it's the document that contains the information you're looking for.
It summarizes what each of those documents says about your search query. It finds different variants of the search terms and the concepts that you meant even though they're not necessarily exactly what you typed and it finds them within each document, generating a summary in a couple of sentences. It’s not a general summary like the ones generated by Copilot, it is a summary of what each document says specifically about your search query. The technical term is query-based abstractive summarization. You can figure out at a glance whether that document is promising.
We talk more about this in this blog post
It also does what's called query-based extractive summarization: a list of sentences taken from the text, that are most likely to contain the information you're looking for, so even without opening the document you can have a quick look at whether the document contains information that you need, in the correct context.
There are lots of reasons why a SharePoint search might have returned that document - it could be that all the words in your query are somewhere in there but scattered throughout the document, not a single phrase. Whereas the summarizer tells you about the entire concept that you're looking for, including close matches. Alternatively, our summary often says directly that nothing in the document is relevant.
The Search Summarizer gives you:
1) the summary
2) the extract - highlighting the most important parts to make it easy to scan and completely eliminating the time-consuming open-and-search-again cycle. This is where it differs from other search solutions.
Another good thing is that PointFire Search Summarizer is an extension to the free PnP Modern Search software, which also allows you to do all the wonderful metadata search and display templating that I was talking about. Customization using PnP Search lets you narrow down the number of hits that are returned, and PointFire Search Summarizer lets you avoid opening all of those hits, saving you time/money and frustration.
How to try the Search Summaries for SharePoint
The PointFire Search Summarizer is currently ins closed preview. You can:
- Schedule a demo here
- Request early access to the preview here
As always, if you have any questions, you can send us an email at sales@icefire.ca