One generic tool for getting the right results on top is the “term frequency–inverse document frequency“, or tf-idf for short. It’s a combination of how often a term is mentioned in a document compared to how often it’s mentioned in all of your documents in the index. So, rare terms within the whole index used often in one document makes it a good search result when searching for that term. But most likely, not good enough. You need to figure out what’s the characteristics of your content, and what are the most characteristic use cases and user stories for your users. Only then can you achieve great relevancy, …  I mean get the right result on top of your search result.

Model for relevance tuning

We’ll use our Recipe app as an example…

So, for our food recipe app, we have some obvious content characteristics:

• The more ingredients in-season for one recipe is good. We’re doing an OR-search on all ingredients in-season so this comes out-of-the-box … almost.
• Quite a lot of recipes doesn’t stand the test of time. We know that most of the recipes at oppskrift.klikk.no from 2008 or newer are quite good and have nice photos.
• We’re not sure if we need this, but we know whom of the writers to trust. This may be an overkill when we already have a boosting on newer recipes.

And we know a lot about our users as well:

• Most grown up people in Norway have a job, thus limited time to prepare a meal. This means that recipes that takes shorter preparations should be boosted from Monday through Thursday. The verdict on Friday is still not decided.
• During the weekend people have more time to make dinner. The recipes that takes a short time to prepare most probably cut some corners, and are not that good compared to recipes that takes a little longer time. So for the weekends, we should do a demotion of really quick recipes, at least for dinners.

This is the info we’re going to use to sort our search result. But we have more knowledge about our users that we can use to auto-set filters for certain times of the day:

• Most work days, people don’t plan a breakfast meal or lunch. The whole day we can auto-set the main “course” filter.
• During the weekend, people may also plan a lunch. We’ve decided to auto-set the “light meal”-filter during weekends up until lunch time. After that the “main course” filter is auto-set. We’ll log if the first thing our users do is to set another filter.
• On Friday and Saturday a lot of Norwegians drink beer, wine or liquor. After some hours of drinking, they get hungry. Maybe we should have an “afterparty, quick and greasy and tasty-meal”-filter auto-set for late Fridays and Saturdays?

So, what’s the filters we’ve decided on:

• Light meals
• Starters
• Main courses
• Deserts
• … and maybe the Afterparty-thingy

What do you think?

Sounds nice? This is work in progress, so check back every now and then for new blog posts.

Tapping into the mental model

Each variable combination will do an OR-search containing a lot of ingredients for that particular place and time of year. Our relevancy model so far:

1. Recipes with the highest amount of ingredients hits
2. Newest recipes
3. Recipes written by Christopher Sjuve

Number one on the list is given, but why number two and three? A lot of the older recipes doesn’t stand the test of time, and we know we trust Christopher Sjuve’s recipes.

The places we’ve chosen:

1. The sea
2. The farm
3. The garden
4. The forest
5. The mountain

It’s an odd bunch of places, but so far we think it will work. Logs and usability testing will tell us later if we’re hitting the target or not. The farm doesn’t fit that well with the others, since you don’t normally enter a farm and steal a cow or some potatoes. But it will be what’s closest in content to your average supermarket, and will be the default choice. Almost all of the places will have some overlapping ingredients. Each search is a combination of a place and a month. 5 places x 12 months means a sprite of 60 images where you swipe horizontal to select a place. Month will be selected for you, but to open up for exploration we think it will be valuable to  have a vertical swipe to select month.

Here’s the first wireframes on the UX concept.

First version of the query matrix. So far not organized by places, but types of ingredients.

The search is already up and running, but lack every sign of graphical user interface.

Thanks to Qbox.io for letting us use a Hosted Elasticsearch instance for this project!

Sounds nice? This is work in progress, so check back every now and then for new blog posts.

Using the phi-spiral as a visual relevancy hierarchy

So, how could the Phi spiral help us?

A search engine list out what it thinks is the most important first. But the list has several issues:

• You could, but should you?
  Just because your template really want you to render a logic list as a visual  list, doesn’t mean you have to do it like that.
• Not representing the information well
  A news article looks like a news article, no matter which version you see: The front page teaser, a short version or the full blown thingy. But a search result almost always looks like just that: A dull list of items.
• Too many items
  Results at the bottom of the list tends to get higher click rates than just above the bottom. I guess this have to do with how people scan web content in F-shaped patterns and that a list of 10 items is too much information for the user to digest.

So, what if we started to use space and position to show relevancy? The Phi spiral, building on the Fibonacci sequence would make a nice search result.

Phi spiral as a search result. We’ll get all sorts of other issues, but I think it’s a good start to getting somewhere better.

You would maybe not be able to show more than 5 result items, but we could put the search box in the middle of the page and then get 6 items.

Looking less like a search result and more like a content filled page.

So, what do you think? Others are using treemaps: Newsmap.jp. Not a very usable implementation, but a nice idea.

A few weeks ago, my colleague Harald and I carried out user interviews at a customer for their “internal Google” solution. Before starting to developing concepts and interaction design, we wanted to properly figure out their users’ actual needs.

One of our findings was that the users generally know the type of content they are looking for. When caseworker Paul searched for “Vietnam Workshops” he knew that it was an application he was looking for – and not a colleague, a PowerPoint file or a news article on the intranet.

Armed with this knowledge, we got some ideas about how to improve their current search experience.

Previously, we thought that this was a good results page when searching for “Vietnam Workshops”:

But maybe this is the way to give Paul a better answer?

By asking Paul to help us decide on the “search mode” before we begin to guess his answer, we will most likely be able to give him a more accurate result.

However, when we user tested this a few weeks later we received clear feedback that the tiled design left the user feeling a little “cheated”. They had searched, but still needed to make a choice before finding the answer. This is also a well known search pattern – even explained by Monty Python.

Turns out Winnie the Pooh has the answer: “Yes please, I’ll have both” is the solution! The search results present a clear call to action to select a category, but with some results already presented on the first page:

A good argument for displaying results in the various search modes right away is to show Paul that we “know more than what he thinks.” By quickly scanning the results page, he may learn something new about his colleagues, see who’s associated with “Vietnam Workshops”, or maybe he’ll find something interesting in an intranet article; hopefully making his current task simpler and quicker to carry out.

But making it clearer that he has the opportunity to choose a search mode, can help Paul to help us give him a more accurate answer.

1. Enlarge your search box! Why are you hiding it ? Screenshot: TMZ.com  My favorite search box: 2. Make your search button big and clickable. Screenshot: CNN.com (Edited. CNN uses the correct version) 4. Use colors common to your users! Screenshot: google.com

5 quick fixes for non-designers who want to enhance their search results page:

1. Enlarge your search box. Almost every second site I see, internal or external, “hides” their search box – maybe because designers think they’re ugly. If you want your people to search: make it easy. I like to add a little gradient on the top of the search box, and to make the corners round. That makes it look inviting to write in.
2. Make your search button big and clickable. Make sure the search-button is as tall as your search box, but make it come out of the page, with a clear color or an embossment. The goal is not to make it look cheesy… but to make it look clickable.
3. Use already designed material as an inspiration. Make sure that your search results, and the rest of the website for that matter, matches your company brand identity or design hand book. Use the correct colors and fonts – they are probably thought through by a designer already.
4. Use colors familiar to your users. - Blue is for links – Green is for URLs etc. telling visitors where the result came from – Red is for alerts and error messages If your brand identity doesn’t have hex-codes for these colors … ask your designer or design partner for then. Kindly.
5. Enlarge your body text size. Many designers tend to user small text in their sketches … maybe to get more room for nice pictures? Your users are probably interested in the result when they are searching, so why don’t we make the text readable. A nice size for a readable body text is 14 og 15px, with a line-height of 17/18 px. Not 12 px Arial…

And a general tip: Don’t try to reinvent the wheel! When working with search design, see how Google, Amazon and Bing format their URLs, links, file type-icons, etc. and make your version of it. Surely you can try another way, but then you should also user test it properly before you launch it.

What are your suggestions to improve search design?

flickr.com/rustychainsaw

One of the pleasures of leading Comperio is that I get a chance to meet a lot of interesting people with exciting ideas. Their undertakings span from how to improve productivity through better information access in an intranet, how to boost revenues and profits in an eCommerce solution through supportive shopping guides and a customer-centric promotion and product placement strategy, to how automate surveillance and investigations on e-Crime for government agencies– to name a few.

What do these ideas have in common? The answer is at least two things; that they all leverage “Enterprise Search” to support new business models and they all show a high level of leadership in pursuing their ambitions. As you will see on our re-launched website the customers are greatly successful with Enterprise Search and the rewards are high for the company they work for as well as for their own careers.

Maybe it’s time for you to pursue your business idea? If you do we would really like to hear from you. In fact, we invite everyone to share ideas around how Enterprise Search can improve businesses to engage in our new blog.

We have exciting days ahead – join us in shaping the future of Enterprise Search!