Exploring A New Search Landscape, With Microsoft’s Jacquelyn Krones
In my last Just Behave column, I shared the first half of my conversation with Jacquelyn Krones, a senior product manager from Microsoft that spearheaded a large scale user study that explored how we use search. In that column, we laid the foundation for what should be a significant shift in user behaviors. Krones said […]
In my last Just Behave column, I shared the first half of my conversation with Jacquelyn Krones, a senior product manager from Microsoft that spearheaded a large scale user study that explored how we use search. In that column, we laid the foundation for what should be a significant shift in user behaviors. Krones said one of the goals behinds the research was to understand how search fits in the context of our every day activities.
More and more, we find search as a relatively inflexible connector between our intentions and our desired actions. Search has struggled to continue as a one size fits all solution to the diversity of our needs. While the search interface has tried to accommodate this by offering more and more types of results through a single portal, it’s unclear how much longer this approach can continue to meet our needs.
One of the main drivers behind this fragmentation of intent is that often are searches are being launched from different devices. And, as Krones found in the study, our search behaviors can vary dramatically from device to device. The required experience on a smart phone can bear little resemblance to success on a desktop or tablet.
“There’s clearly a different profile of these activities on the different platforms. On desktops and laptops, people do all three of the activities – they conduct missions and excavations and explorations.
On their phones we expected to see lots of missions – usually when you use your mobile phone and you’re conducting a search, whatever you’re doing in terms of searching is less important than what’s going on with you in the real world – you’re trying to get somewhere, you’re having a discussion with somebody and you want to look something up quick or you’re trying to make a decision about where to go for dinner.
But we were surprised to find that people are using their mobile phones for exploration. But once we saw the context, it made sense – people have a low tolerance for boredom. Their phone is actually pretty entertaining, much more entertaining than just looking at the head in front of you while you’re waiting in line. You can go check a sports score, read a story, or look at some viral video and have a more engaged experience.
On tablets, we found that people are pretty much only using them for exploration today. I had expected to see more missions on tablets, and I think that that will happen in the future, but today people perceive their mobile phone as always with them, very personal, always on, and incredibly efficient for getting information when they’re in mission mode.
As a result of these different profiles, you can see that we do different things. Even though we hadn’t put this framework together, we intuitively understood that the kinds of things you want to do on your mobile phone are different than on a PC. We didn’t necessarily anticipate the exploratory behaviour but we certainly did anticipate the missions, which is why right from the home screen of Bing you’ve always been able to get to the kinds of things you’d be doing on your mobile phone very efficiently. If you look at the Bing tablet experience, we built that with the idea and understanding that people would want to explore. It’s a beautiful and delightful experience because it matches what people actually want to do.”
Mission Mode vs. Information Excavation
But how does a search engine know whether you’re in mission mode or gearing up for a major excavation of information? And, if we’re “creating knowledge” through the use of other applications and tools, is a stand-alone search the best match for our needs?
Almost 10 years ago, Microsoft was working on a major revamp of their OS that was code named “Longhorn.” Their ambitions were lofty, including integrating contextually intelligent search functionality. Search would cease to be a separate destination and would be a simple right click away from whatever we were doing, without ever having to leave the app.
In fact, Microsoft was working on something called “Implicit query”, where search suggestions would appear in a sidebar based on what you were doing at the time.
While the idea made sense then, and still makes sense today, Microsoft discovered that they were introducing a feature we weren’t ready for. The functionality got scaled back for the release, which eventually became Vista. Yet, the idea still intrigued the folks in Redmond and they decided to re-explore it in their 2007 study:
“The last time that we did ethnography, in 2007, we were all excited about this idea – why should people have to go to a destination search site? Search should be everywhere, it should be easy to get to and you shouldn’t even really be thinking about it, you shouldn’t have to break your context in order to conduct a search.
The best example today would be on mobile platforms, including Win Phone 7. There’s a search button built into the device and no matter where I am, I hit the search button and it searches, understanding the context that I’m in. If I’m in mail, it searches my mail – if I’m on the web, it brings up Bing and I can do a web search there. I think you’ll see more and more of that.”
Understanding the context of a “Mission” search depends on picking up signals “in the moment” – including location and time. But an “excavation” search presents a much bigger challenge to the search engine. By their nature, these searches involve extended research and aggregation of information.
In the past, I’ve referred to them as “master intents” that we then divide up into individual sub-tasks. We do this because the heavy lifting required resides squarely on our shoulders. Search assists in the role of a “go-fer” but remains ignorant of our overall intent. Krones speculates that this might not be good enough any more:
“The interesting thing is that when you think about excavations, there really is no perfect tool for managing your excavation activity. Excavations can range from pretty small to pretty big. If I’m searching for a pair of shoes to wear to a wedding, that can be an excavation for me because I want to find the best pair to match this dress, or it could be a mission or it could be just an exploration.
But maybe I have a health issue and I have to make some really important, life-changing decisions. What tool is there that allows you to search, compare analyze, get input on, save, organize, and come to your ultimate solution? What we found is that search has a very high mindshare for excavations in terms of being a fundamental tool.
Following this research, I had another project that I did with an online panel of 20 people where I had them work on an excavation over a period of 3 to 10 days. Every day they reported back what they did, how they did it and how it went. What I found was, to a person, everybody started with a search engine and almost everyone used more than one search engine throughout their process of excavation. So search is very connected to excavation.
But what it doesn’t do today is allow you to do what users are trying to do on their own – they’re cutting and pasting information from different places, trying to compare, analyze, synthesize and come up with their own answer. For informational kinds of excavations, we hear people say they’re not finished until they start seeing the same information over and over again in different ways from different places and they feel like it all makes sense to them.
You can tell that this is an incredibly effortful process on the part of users. It’s important to them, both from a functional perspective and from their own self-concept as a person who creates their own knowledge perspective as they go. Search is a natural place to start thinking about how you make that easier.”
Specialized Search Apps
It’s these extended online sessions where the greatest promise of what could be exists. The “in the moment” problems are increasingly being addressed by specialized apps that are very good at doing one thing. But let’s raise the bar a little bit. Why shouldn’t there be a perfect application for shopping for shoes or all these other little segmented parts of our lives that can all incorporate search where it’s useful and functional?”
If suddenly we have this broad universe of all these little very specialized applications that we use as required, then you have this need for search to find the right app or maybe the right collection of apps given where we are and what we’re trying to do in our lives at that time. Ultimately, this “meta-app” approach could assemble the right pieces required to help us with the bigger “excavations” as they arise.
“One of the reasons why apps have taken off is because instead of having a website that tries to do everything; an app does one thing really well and really easily. Instead of trying to find something, the information is almost instantaneously there, which is a great experience when you’re in mission mode. But you’re also right that what people are essentially doing is creating their own kind of super apps – let’s talk about mobile first, then I’ll talk about PC – by taking all these apps that do one thing really well and stringing them together in terms of a solution that makes sense for them.
So for instance, when we were in San Francisco, almost everybody was a heavy Yelp user. Let’s say the way that I use Yelp is when I’m in a different part of the city that I don’t know very well. I’m with a friend, and I look at what restaurants are around me, then we find one that has a high star rating and a lot of reviews, and we go. And that’s how I use it most of the time.
But let’s say that I’m going to run an errand and it’s a place I don’t know very well, so in addition to picking the restaurant, I also have to figure out how to get there by bus. I think what we would all tend to say, “Yelp needs to integrate bus information,” but what if only 5% of their target user base in any particular city uses the bus information and another 10% walks and another 80% will drive? Rather than make the app like a website that tries to do everything and as a result makes it a little less efficient, people will pair those apps together on their own – they’ll find the restaurant and then they’ll go use their bus app to find the best route.
People are already kind of creating their own customized solutions for the tasks that they do often. And you can see that most clearly on their phones, but people also do this on the Web too. While we would love to say “We’re the perfect shoe store because we can be comprehensive,” that doesn’t really sell well with people. And when you get into something like a health issue where the stakes are so high, that definitely doesn’t sell well with people. We can bring in information that we think is incredibly high quality and trustworthy, but people are still going to want to look at other things.
For things that people do often or that are very important for them, they create their own suite of sites and applications that they trust for this area. When they go out and start searching for a new topic, they know where they’re going to go. They’ll go from site to site to site and do the same query over and over again. Then they use search to make sure they didn’t miss anything – to make sure there’s nothing new and better. But what search isn’t doing for them today is making it easy for them to manage that process they have of using the set of resources they love or trust for a particular topic.”
Microsoft As The Search Underdog
If there was one word that could describe Microsoft’s role in the past, it would be monolithic. The sheer bulk of Microsoft sat at the center of our computing experience. But the world of a million different apps is a far cry from the virtual monopoly that Microsoft has enjoyed in the past. How might Microsoft reinvent itself to be relevant in this new, much more fragmented marketplace?
“Our strength is in creating a platform where an ecosystem can do an awesome job of bringing the content and the actual transactional capabilities to play. So our role would really be more about enabling these things to take place across different providers, different publishers and retailers, not to actually create a bunch of small apps on our own.”
Finally, I asked Jacquelyn to look forward 3 years and predict what behaviors she might see when the next study, in 2013, again looks at how we conduct online searches…
I really hope that what will have happened by three years from now is that these behaviours that we’re seeing around people creating knowledge and creating their own customized solutions – I hope that the industry has caught up with them. I hope we see that people are doing this much more efficiently and that they feel empowered not just to create their own knowledge but that the industry has supported them creating their own customized personal solutions. I hope the industry provides what they need as an individual and provides a wonderful experience.
Opinions expressed in this article are those of the guest author and not necessarily Search Engine Land. Staff authors are listed here.