Search Evolution: Data Convergence & The End of Keywords?
While search technology naturally evolves, we must also recognize that human’s usage, expectations & desires for search are evolving as well. As marketers learn to leverage the information, tools & channels available now, we need to be cognizant and welcoming of the evolving search trends on tomorrow’s horizon. Who should be playing catchup, search technology or the human brain?
Search Engine Strategies San Francisco invited three bleeding-edge thought leaders to examine the question, Search: Where to Next? The attentive SESSF audience were treated to a mix of marketer’s, toolmaker’s, user’s & search engine’s perspective on the shape of search to come. Moderating this discussion was Graham Mudd, Vice President of Search & Media at comScore.
Speaking first was Marc Poirier, Co-Founder & CMO of Acquisio
There’s a huge discrepancy between how much time we spend consuming media online and how much money we spend. 34% of media time is spent online, but online only represents 12% of overall media spend. There’s a disconnect. Media spend is predicted to double in the next few years and this is not new money, it will be coming from traditional channels.
There are three new technologies that Marc thinks will be most important in the next few years:
1. Attribution. The ability to track, not just inside search campaigns, but outside. How channels interact & tracking attribution across channels. Marc says that we have to stop looking at last click attribution and start looking at the bigger picture of various events in the buying cycle.
2. The ability to purchase display like we purchase search. Today, when you want to do a display campaign, there’s faxes and signatures to send, negotiations to make, it’s very complicated. That’s changing as well. Now we have ad exchanges, hubs to procure inventory from different publishers, deciding how you can pay for each impression.
3. Real time Bidding. RTB + better reporting will help us better capitalize on the money that’s there. We have the ability to say how much we’ll pay for each impression and measure success display. We will be paying for display like we pay for keywords. Marc thinks this will be widely available very soon.
Marc thinks search marketing will become part of something bigger. Search is just one of several channels. There is a lot of data across channels & we are going to evolve into performance marketers rather than search marketers. To do this, we are going to have to manage a lot of data. The amount of data you get in display campaign is staggering.
We are going to need different skills in this industry. People with data skills, like statisticians & database admins are going to be increasingly attractive to hire in this industry. There will be a shift.
There’s going to be a change in how we manage internally. This is really a fantastic opportunity to get deeper into other channels, prove performance and ultimately, get more money.
Next up was Brian Kaminski, Chief Operating Officer of iProspect
Brian asks us to think back a few years, to the dominance of the Sony Walkman as a portable media player. It had an interesting ride, not just in market share, but in share of mind. Since then, the technology has obviously changed. Our “portable media world” is now dominated by digital devices like the iPod.
What happened to Sony? As the marketplace continued to evolve, Sony missed a big opportunity to evolve with it. Ferris Bueller best summed it up “Life moves by pretty quickly, if you don’t look around once in the a while, you might miss something.”
Seizing the Opportunity
What will search look like in the future? This may be a bit bold, but Brian thinks that in 5 years people will not use search engines. Think about it, imagine the rapid evolution that’s taken place from the primitive green-screen computers to the devices we have today. The evolution will be similar for search.
Search is an interruption. Brian doesn’t think people will use search engines because there will be new things to move people to a and b. That’s not to say that this technology won’t have or be built upon search functionality.
How do we impact things tomorrow? Brian has 5 clues about the future of search:
1. Today, the advantage goes to those who are genuinely using the data, not just staring at analytics. Attribution modeling, testing, understanding how everything is connected. Those people stand out from the crowd. Today data is an advantage, but tomorrow it will be a basic requirement.
Marketers need to embrace the change and the move towards data. Invest in technology system that facilitate the use of data moving forward. Invest in people with seriously heavy data skills. It seems odd for marketers now, but will become the norm moving forward.
2. Most search results are one too many. This is a far cry from the future where search results will be much more realtime and much more connected. It’s starting now with Twitter, but will morph into a wider group of realtime results. Brian isn’t totally sure how this will come into play, but he feels that the social graph and relationships will play a major factor in your search results. Users are much more comfortable with data validated by humans they trust.
Embrace real-time content. Twitter is a little scary because it doesn’t necessarily have the filter that corporate would like.
3. Convergence of new technologies and device coupled with changes in user behavior. User’s will desire content and devices to be more portable. in the future, search will be dominated by non-pc activity and non-pc searches. Mobile search opens up the opportunity for location based targeting and context.
Embrace this content and become comfortable in providing to users in non-pc users. It takes money and adjustment, but will be worthwhile
4. Search is really a net, a capture mechanism. We use search as the net to capture people from all other marketing. More media will be planned specifically to drive search. Search will be at the center of the marketing agenda as it becomes a bigger part of the budget.
Embrace testing of new campaigns. How does email & display affect lift in branded queries or impressions?
5. Today, search is very keyword dominated, the essence of all search. Search will become less about keywords, it will be about leveraging the camera in your mobile device to take a picture and automatically gather data. It will be about barcodes comparing prices for products nearby. These are just a few examples.
These 5 predictions mean that things are going to change, possibly very dramatically over the next 5 years.
Don’t sit and wait for the change, anticipate and jump on the opportunity. Set aside more budget for testing and innovations. Invest in data system and data skills. Continue to think more outside the PC, take other devices into consideration. Don’t be afraid to jump on new trends more quickly.
Rounding out this session was Shashi Seth, Senior Vice President, Search Products, Yahoo!
Shashi asks, how many of us find ourselves overwhelmed with the search results from a query? How many of us use social networks as a proxy for search, asking our friends for information. How many of us spend time on your mobile devices looking for innovation?
(Matt note: I could not accurately capture audience hand counts because I was facing forward & blogging)
Users have changed significantly over the past 12 years. Everyone in the industry faces this problem. User’s needs have become more complex, the queries they’re doing are significantly different and the amount of time users have is shorter (as is their patience).
People are spending more time on different times of content . More time on images, videos, their mobile devices and social networks. These are all areas that the search industry haven’t invested in as heavily.
The last 12 years have been about the back end of search – crawling, indexing and algorithmically determining relevance. Those were important, Shashi understands the investment. But the front end of search – new sources of information, new ways to get there, getting answers quicker on all the devices you care about, that’s where the industry is headed.
Shashi is often asked, “What exactly is Yahoo search doing, in partnering with Microsoft?” Algorithmic and sponsored results from Microsoft are only part of the equation.
This technology is Yahoo’s canvas for innovation.
They are taking their focus off the backend of search, letting Microsoft handle that. Yahoo is basically taking their scientists and programmers who’ve spent the last 12 years on the backend and moving them towards the front end of search.
Selected Recent Innovations
Deep Web – There is tons of information available with many different way to ask a question. If Shashi is interested in a Spanish restaurant in San Francisco, he can type in Yahoo and get good results. But what if he want to search for a particular menu item, like Paella? This science actually existed in yahoo – taking deep information from structured and unstructured sources where people can ask complex questions and pivot in many different ways.
How often have you wondered about a very specific question like “Who were the last 10 winners of the Nobel peace prize?” or “Which countries where they born in, how many came from Princeton?” Search technology can’t do this yet, but very soon users will be able to get these answers from search engines, rather than a sea of blue links.
Yahoo does a lot of realtime data, recent example with the Olympics and the Worldcup. But Yahoo understands there are many types of realtime, like “when is this train arriving, how much was that stock traded for an hour ago? ” Yahoo is investing heavily into these answers as well.
Shashi says that discover and serendipity will be increasingly important for Yahoo’s users. Data relevant to users’ context and geography.
Panel Discussion Time:
Mudd: I’d like to hear how you think structured data will change the future of the SERPs & how you can use this trend from a marketing perspective.
Shashi: We launched a product a few years ago called Search Monkey, recognizing the immense importance of structured data. That products allows publishers and developers to highlight & markup the information ofntheir pages, so Yahoo can pull that information & incorporate it into search results. We’re recognizing the fact that people want answers, not a sea of blue links. Users are going to be delighted and thankful for that kind of experience.
Kaminski: As a marketer, it’s really about speed and customization. It gives you the opportunity to get your data out there in a customized way. The ability to further customize the data is an opportunity that shouldn’t be missed. It’s better leveraged for customization & for the speed of realtime.
Mudd: One other way that data can be used is for personalization. For an engine, this provides the opportunity to show more relevant results to the users. For the marketer, it presents a real challenge: Every search result looks different. This is something we all know has been happening for a couple of years. Shashi, how do you think about personalization, where do you see it going in the next couple of years? How do marketers react to this trend?
Shashi: Why is this important? When Yahoo looked at their numbers with 65+ million users worldwide, we asked how many of these users looked to Yahoo to engage for search. We found that only 45% of yahoo users are actually using Yahoo search. Diving deeper, many users are actually spending a fair amount of time on yahoo properties, yet not using yahoo search. When we are actually giving them contextually relevant, personalized and timely information, we find their usage patterns go up. We’re betting on that, when you can personalize to your user base, you’re likely to have higher engagement.
Poirier: For Shashi, do you look at personalized search results and see ways to monetize these better?
Shashi: When the results are personalized, the user engagement goes up and happiness goes up.
Poirier: As a marketer, one of your biggest opportunities is to have as many 1 on 1 conversations with your customer base as possible. Typically, this is very challenging to do because one set of results is pushed and it almost collides with the userbase. I view this as a huge opportunity to better initiate and sustain 1 on 1 conversations with customers. We all agree there’s a threshold where 1 on1 conversations become annoying and where to cutoff as well.
Mudd: The notion of social as a corpus of information that you’ll see increasingly in the SERPS. with more social results incorporated, the more risk you’ll see in spamming those results.
Kaminski: It’s a tricky line that varies wildly from client to client, from one product or topic area to another. Understand the number of times you want to show the user an ad, at what point are you just plastering the ad in front of them? I’ve seen it be fairly all over the map, but the thought that I had is that many clients go “we don’t understand the data or the magic formula, so we’re almost afraid to start.” I sort of argue with them at times - if you’re hitting 3 out of 10 you’re still going to be in the hall of fame.
Shashi: Social is definitely very important in the next generation of search. For example, if you’re looking for a Nikon D50 camera, wouldn’t it be great if we could cull several of your Facebook friend’s reviews of that particular camera. That obviously brings challenges to bear, like spamming. The interesting juxtaposition for that is that the opportunity there is significant. Can we facilitate this exchange, these connections on Facebook and make sense of layering this over your search results?
Mudd: From the marketers perspective, when you’re thinking about social, it may be that the way to think about it is, do things you think are truly useful for the user. The Nikon example is a great one. Nikon could really be encouraging users to write reviews of it’s product on Facebook, or whatever channel.
Thus concludes a fantastic session & a valuable peek into what the future might be for search. Examine the implications some of these predictions might have on your own marketing endeavors & end users. You may find you’re aligning with the evolving needs of search users, or maybe you’ve only just learned that “wheel is good, but fire hot.”