Googlebombing may be a word some newbies have never heard before. Once it was a political trick and a prank to arrange mass link networks to connect a message with a keyword. The most famous one was the Whitehouse's biography of George W Bush ranking for the phrase [failure].
Friday, February 25, 2011
I found it rather hard to put together a presentation for Search Engine Strategies this year.
Thursday, February 24, 2011
A busy room! Mobile is still the hot! This morning we're moderated by Erica Schmidt of iProspect and our speakers are Jen Brady, the CEO of Fred & Associates and Steve Page, CEO of Mobile Commerce.
Let's not forget to look at what customers want. Other pain points: agencies are moving too slowly says Jen. This frustrates customers. Jen also suggests that marketers think mobile is smoke and mirrors. I suspect this view my by influenced by the fact that Jen's from the States as I've not personally encountered that view here in the UK or elsewhere in Europe.
Don't just do a mobile campaign to do it. Make sure it's relevant and make sure it covers the right platforms. Don't forget the BlackBerry.
Jen's found her trip to London interesting because she's not been able to find local restaurants by using her mobile.
Another pain point is the multi-platform demands; iOS, RIM and Android. There are also a lack of standards for the likes of analytics and quality guidance.
How to respond?
The world no longer has borders but consumers have different demands. Geography is still important. Mobile users in Japan are different from those in Scotland.
Publishers are now aggregators of demand. Publishers have been successful at this but marketers are yet to follow.
It's easy to set up mobile when setting up your PPC campaign. It's a missed opportunity if you forget, Jen says.
Tips on going mobile:
- Focus on desktop SEO
- Use paid search for discovery
Steve points out that on different handsets you get different ads but the same natural search results. In fact, in his examples the natural search mobile results match the desktop results. He suggests you might as well concentrate on pure desktop SEO.
Steve pulls up one airline - let's call them Rainbow Duck - that has a site which isn't optimised for mobile. So once it's been found there's very little chance of of anyone being able to use it.
People don't scroll much on mobiles. You need to be at the very top of mobile results to be in the game.
About 5 million searches a day go via Steve's company. They work with some of the mobile networks.
What are the biggest search terms on mobile as tracked by Mobile Commerce. In 2010:
1: Facebook (15% of all searches)
13: You Tube
14: Face Book
Bebo is that high based on its performance at the start of the year. It drops like a stone afterwards.
Steve ponders why so many people search for Google. I ponder why people try and type in Facebook.co.uk.
Sports are popular. When England are playing people just search for [England]. They don't search for [England football] or [England cricket].
Steve's tried to compare this to Desktop searches. What becomes clear is that some search terms are more likely to appear on mobile. Turns out that only a small percentage are looking for location results. Social networking has about 15% all all searchers. About 45% of all mobile searches are "single user search", ie unique long tail searches.
Local only get about 1% or 1.5% of all searches. Questions arise over searches like [manchester city] - is that a sports search or a local one?
(* Live blogger note: Just seen a tweet that says "Google rep confirms 1 in 5 search queries have local intent. 1 in 3 mobile queries are also local. #SES" via @NickWilsdon. Funny.)
The trend? Single user searches are on the increase.
Which ads get clicked on the most? It's important to keep in mind the low end devices have a very different search experience than the smartphones. 49% of people click on the first ad. 22% click on the second ad in mobile.
In the States only 4% of searches come from a high end device. In the UK 15% come from high end devices.
The CTR of an average mobile ad in the UK is about 3%.
When launching an app it's crucial to appear in the "Featured" section in the AppStore. This is controlled by a team in iTunes. This is where your launch spike of downloads will come. Those downloads will put your app into the top of the category sections.
A rather sneaky way to remind your customers that your app is there - is to release updates. On the iOS this encourages people to re-download/update the app.
The app PositionApp for the iOS will tell you how well your app is doing.
Do you keep your rankings high? Companies like o2Media can help. A SMS campaign can be sent from o2 to help promote your iOS app. The very best way is to feature in one of Apple's own promotions. In other words, when Apple are doing an "app for Christmas" style promotion then it helps hugely if they recommend you!
Free or paid? Free certainly encourages those downloads. Some of Steve's clients have moved budget from development (trying to earn money back from paid apps) to marketing (brand boosting via free downloads).
The final collection of takeaways; Apps are always developed late, buy at least one handset on each OS you want to target (ideally one on each operator!), learn how the OSs work by using other apps in anger and understand what you're trying to achieve. Lastly; think about how the app works when it suddenly looses signal.
My last day of live blogging at Search Engine Strategies 2011 here in London. I'm speaking later today. First thing in the morning and I'm in the main room for "Video Search Optimisation". It's a busy panel with Sean Hargrave as the moderator, Paul Carff of Google (he would be a good person to interview!), Will Critchlow of Distilled and Jonathan Allen of SearchEngineWatch as the speakers.
I must admit that I feel this session has something to address. What on Earth do we mean by "Video Search Optimisation"? I think the term has become overloaded and confused. That messes with client expectations and that's a bad thing. I wonder if we'll touch on this today or whether we'll dive into tips.
Google's up first. I expect we'll be talking about video sitemaps. Paul's had some volcano issues.
YouTube streams 2 billion videos every day. 70% of all US online video viewing happens on YouTube.
Video is fasting growing medium in history.
53% of people took action as result of watching a video online.
77% of Americans online watched a video last month.
Streaming on YouTube from Sept 2009 to Sept 2010 is up 36%. Average time spent per view per month is up 46%.
Google wants to create a large video index. Webmasters can help by using the Google Video Sitemap. Paul suggests that SEO comes into play when writing the title and description for videos in the XML file. Recommended but not necessary are
publication date and
restriction relationship which is a geo-restriction.
It can take a couple of days for Google to process all the videos in a sitemap. This is a big improvement on before.
Paul announces http://code.google.com/p/video-sitemap-drupal. It's version 1.0.0 and Google are eager for feedback. There will be Joomla one soon.
Are we done with that? Hope. Paul reminds us there's more than search. We cut to a video about a tongue cleaning product. The message of the video - YouTube is the great leveller. It lets SMEs compete with giant international companies when it comes to some forms of advertising.
Back to Paul and the highlight that video is better for discovery than search. Related videos are a good space and you can now buy promoted videos to appear here too. With "True View" you can ensure that you only pay when people watch enough of your video.
Creative content is always the big winner. Think about what will interest your desired audience.
Old Spice Stats - 120,795,481 views, +107% last month's sales, 151, 136 subscribers, 180 videos and 1 basic proposal.
More than 50% of all videos have been rated or commented on.
With the launch of Google TV and others the role of TV is becoming more central for online video. YouTube have launched Lean Back to make the large screen experience as good as possible.
We're starting with on-site tips.
One strong reason to do video is to get into the SERPs for key terms. Videos in the SERPs can link back to your site.
In addition to sitemaps there are video microformats; the Yahoo Search Monkey one and the Facebook one.
Another tip is to publish a transcript of your video. It's a good use of text and will help you to rank. Speechpad is a cheap and effective way to achieve this.
Will recommends tools in addition to YouTube. One such tool is Vzaar. Wistia is another and it makes analytics much easier than working with YouTube.
Another tip is to provide a textarea box with your own embed code in it. In this code include a link back to the page that hosts the video. One thing to watch/exploit is that the subscribe to channel link is automatic without confirmation.
WIll promotes Unruly Media as a good way to get your videos out there. It'll be a good day for them in Search Engine Strategies as I'll be mentioning them too.
Jonathan has stared as Bilbo Baggins. Well. At a school play. We'll be looking at mid and high budget video tactics.
Whether this was planned or not - we're now quoting Guillermo del Toro who sees video as a story engine.
What's the video agenda?
- Production values
- Marketing strategy
- Channel / monetisation
- Measuring success
In production values let's look at UGC, cheap or semi-professional - thus leaving "Broadcast" or TV quality out of scope. A word of caution on cheap videos; they look cheap. Jonathan thinks it's therefore better to do UGC better.
No budget at all? Xtranormal and Moviestorm are two sites that will let you create video from simple text scripts.
If you have photos then Animoto will put them together, to a soundtrack, and create a video. A tip from Jonathan is to use this to create intros
Create a video by submitting your sitemaps to Treepodia.
Tips for getting for extra traffic.
- Upload all the videos you get. Google will sort out the copyright issues for you. It'll either be automatically be blocked or you'll get the benefit for it.
- Report on events; get those voxpops and shoot crazy stuff.
- Keyword rich titles on YouTube, target known search queries, copy other people's tags and tag people on Facebook.
Eww. We've a video of Jonathan eating a live octopus. Let's hope he doesn't play it. As it turns out "Eating live octopus" is second in Google's search suggestion on [eating l]. In fact, "eating live [anything]" is pretty popular!
Link videos together as a playlist. This works as well as it also makes sense to cut longer videos down into smaller chunks.
A way to make a cheap video look less cheap - try and frame the the subject, try and get eye level at about 2/3rds towards the top of the screen, pick a nice background and add some "pick up shots" to spice things up.
Moving on to professional content... branded content is increasingly popular. If your brand doesn't lend itself well to video you can still come up with interesting twists. For example, Diesel Fragrance Factory have a "Flute Boxing" video. Diesel's had their video chunks turned into a TV episode.
A final tip is to pay video influencers to do branded product placements. Let their own production values come through and don't try and control these placements.
Drats! Out of time.
Wednesday, February 23, 2011
My last live blogging session of the day from Search Engine Strategies. Search Engine Watch's Director Jonathan Allen is moderating and we've got Howard Ogden from Mobilistar and Mark Lister the MD of Imano as presentors.
Layar has had a $14m investment from Intel. The major research companies are predicting big things for AR. THey're calling it disruptive and the new mass medium.
AR is a revolution in vision, a new interface for reality and a window to the data age - a lot envaglism for Augmented Reality from Howard. He's keen to suggest that AR isn't a piece of technology.
He disagrees with Gartner in that it'll take five or so years for this to become mass. It'll take less time.
Where did it begin? The start was 1D barcodes, to QR codes and then to 2D marker based AR. We're beyond that now.
He holds out Word Lens as an example of AR available today that's simply mindblowing.
Three major AR platforms to watch; Junaio, Wikitude and Layar. Howard suggests that Junaio is more technology advanced than Layar. Wikitude has Wikitude Drive coming out.
There's Google Goggles - possibly the most famous AR tool. Howard says it'll be very big. He managed to speak to Larry Page about AR. Howard predicts that Goggles will be key to Google's location based advertising AR options. After all, they already have Street View that maps the world. They could augment building faces with data. They've the Android platform in place to let them do this en mass. He predicts that Goggles will be built into the browser - no need to download the App.
Howard agrees with Google. Mobile first is the right policy; especially HTML5 web apps.
Junipter predicts that revenue associated with AR will grow from $6m in 2008 to $350m by 2014.
Where are the top regions for AR. He did a keyword search regional distribution.
1: South Korea - 26.8%
2: Singapore - 14.5%
Top cities; Seoul, Singapore, Munich, Amsterdam...
.. but there's a catch with his research technique. We're finding the homes of large mobile maketers and the large AR platform providers.
In the "Internet of Things" we'll see four key areas;
- Social Search
- Augmented Data
- Ambient discovery
Phone mobiles are almost people like with their sensor abilities now.
We're heading to a future that will have contact lenses that augments your vision with data. They have the technology now.
The singularity is coming. Time predicts 2045 is the year we become immortal through technology.
Both Google and Apple have bought technology that specialises in facial recognitional.
Google use the phrase "the bar of daily engagement". Howard admits that AR isn't there yet. What will that take?
- Useful AR
Mobile app development is the real growth behind the digital agency Imano, says the MD. Recent changes in client demand is a move away from simple brand and towards utility apps.
Smartphone + Tablets > PC shipments, says Morgan Stanley Research. By 2013 the mobile hardware greatly outstrips the desktop stuff. A lot of the growth in this area comes from BRICS. Smartphone growth in those geographies are picking up.
The benefits of mobile?
- Bypass fixed telephony (very useful for the developing world)
- Smartphones bypass PC/Mac
- Quick answers
- Personalised content based on location
- Social context based on location and status
Hyperlocal marketing is coming. For example, a bagel shop might be able to advertise to people within 100 metres. A far better idea than printing off tens of thousands of flyers to hand out at Tube Stations.
Mark illustrates with some client examples. One theme is the merger of social into AR.
What's coming for AR?
- 3D Facial Recognition
- Near Field Communication
- Biometric apps/heart rates
- Character recognition apps
- Second screen recognition apps
- Mobile growth is unstoppable
- Growth of apps continues - for now
- AR is a powerful tool for mobile use
- Not just fun and gaming. Use for security, health, payments, leisure and automotive.
My Search Engine Strategies live blogging continues, typos, warts and all with coverage of "Crossing the Digital Divide: The Leap from Search to Display". Ciaran Norris from Mindshare will moderate while Sacha Berlik, Founder and CEO of Mexad and James Yancey, MD of SearchIgnite Europe, present.
James makes a prediction; next year this room will be overflowing. I tend to agree. Behind him the slide says it's not a battle between search and display; it's about understanding the cross-channel.
From among SearchIgnite's clients in Europe, James works with five or six who have set up biddable media departments. However, he goes on to explain it should not be about search and display departments fighting for this budget - it should be about the cross-channel budget.
Why do DPS sound familiar to paid search people? It's an auction, it uses tech and automation, it's objective and accountable and is a direct response medium.
In search there is no cookie ID, represented by a keyword and cannot be bought in bulk impressions. In display there is cookie data, demographic survey data is available and you can buy in bulk (CPM) on the impression level.
The growth in display is not simply down to the auction - it's about the data you can have. Keyword "can". If you don't have the data then you don't have it. Data lets you understand the value beyond a simple keyword.
James suggests that real-time bidding is sub-set of auction based display. You're buying single impressions rather than collections of impressions. You're targeting someone based on their cookie ID. If your data is certain someone is about to transact - then it may be worth paying more to target that person - hence real-time bidding. It's that ability to target that person at the right time.
A DSP is a technology interface that sets across many auctions. By themselves, James says, that's not impressive. It would be like having a search tool that let you work across Google and Yahoo at the same time. The key thing, the interesting part, is having the data.
DSPs can be built in basements with friends. It's the data source that's important.
What are some of the challenges?
- Premium inventory
- Privacy legislation / behavioural targeting
- Attribution modelling
Sacha comes from a Display background, does not have the Search background so from his point of view the Display world has changed dramatically. He founded MexAd in 2008.
Display bid management is the entry point for search marketers, says Sacha, it's a good chance for them to take a big piece of the cake from the traditional media agencies.
The traditional media agencies ignored search for too long and stuck to their old display buying ways. This has left them vulnerable but it's a lesson for search marketers to avoid.
What do we need to know? There are no pre-committed budgets in display any more. This is a big difference from the pre 2008 days. No need to buy a month of impressions, hope for the best and see what happens. Today you can optimise in real time.
What's different? In search there is only Google. In display there are at least six; and they're different down to the algorithms. In a few months there will be at least eight platforms you will need to know.
Sacha predicts that click optimisation with just session cookies will no longer be effective and don't expect high post-click conversions.
Some exchanges; AppNexus, DoubleClick and RightMedia. We're going to go deeper.
RightMedia was the first - been here since 2005. Pro: solid technology, massive reach worldwide and great optimisation algorithm. COn: no real-time bidding via external DSP, lack of media quality (so many dodgy sites), very slow in development and it's blind by design.
AppNexus: don't like to be called an ad exchange, call themselves DSPP. Pro: great technology, very fast, RTB on many major suppliers, transparent, audited inventory and server side cookie storage. Cons: Many limits of creative and third party tagging.
DoubleClick Ad Exchange. It's Google. Pro: giant reach worldwide, full transparency, accepts DFP partners, audited inventory and great targeting options. Con: poor UI (AdWords!), no proper pixel tracking, limits on ad spec and without your own technology you're lost. You need to use a large ad server to get into this and you need your own RTB to compensate for their lack of pixel.
There's also OpenX. It's blind by design and Sacha doesn't think it's mature.
On to yield optimisers aka SSPs. We have AdMeld, PubMatic and Rubicon. MexAd predicts that these SPPS will re-brand themselves towards exchanges.
AdMeld. Pro: integration of 3rd party data sources, real-time bidding and focus on technology that makes buyers and publishers happy. Con: High floor prices.
PubMatic. Pro: High quality in European markets. COn: RTB only on US inventory. Non-English inventory is high CPM.
Pubicon. Pro: nice publishers, worldwide, good reach and RTB. Con: A lot of non-premium inventory. Okay, not as dangerous as RighMedia (no p0rn) but it's not wise to rely on their internal classifications.
Two blog reading tips; ExchangeWire and Adexchanger.
(*Live blogger's request; dear Adexchanger please move to full text RSS)
Sacha reminds us that the data driven marketplace is very young. Technology is immature and there is a fight for market share and survival. It's a tough fight, for example, Google blocked AppNexus and advertisers were unable to buy Google Inventory. It took eight weeks for AppNexus to get back into the Google display buying system.
Technology doesn't compensate for lack of knowledge and this is a complicated marketplace.
Try and gain access to your RTB sources and don't just lean on your DSP.
There's been a change to the line up for "Is Search Remarketing/Retargeting Right For You" and Jon Baron, the TagMan speaker, can't make it. This is a shame because I really wanted to ask about TagMan's achillies heal - that they throw away data in order to de-duplicate - and how that must negatively impact Retargeting.
However, the good news is we've got some solid speakers for this Matt McGowan moderated session. There's James Yancey, MD of SearchIgnite Europe, Magnus Nilsson, Search Marketing Director at Banner Corporation and Guy Levine the CEO of Return on Digital.
We're going to censor client names for this blog. I'm going to call them all "Rainbow Duck".
Guy begins by pointing out retargeting is only possible on the content network. It's a simple way to connect with users based on their past interactions.
A client - Rainbow Duck - tracks when visitors come to the site but fails to fill in the "Contact Us" form. That triggers the retargeting. Their banner ad starts to follow that visitor around the web.
Guy moves on to show how to set this up in Google. Navigate to the right place and create your list. These lists are based on the actions people take on your site. For example, you can target visitors who've abandoned the shopping cart. You can also time delay this, so set a cookie to last a year and then retarget people for the next season with the next ad campaign.
On a note of privacy; your list needs to be at least 500 long before Google lets you use it. Guy recommends starting to build your list now so you can build your lists.
For one client - Rainbow Duck - the first retargeting campaign wasn't capped. This sparked a bit of a storm as people wanted to know why the ads appeared all around the web. I actually saw someone blogging about it. Guy points out the capping option if you want to avoid this.
Despite the blog drama about the frequency of the ads - Guy's figures show how the retargeting was a fantastic success. Leads went from a low of >150 to a high of >400. Costs wen from a high a >£350 to a low of >£150. On reflection Guy does admit it would have been better to have used more segmentation (ie, more lists) and a wider use of creatives.
The campaign saw an increase in ROI. Why? More targeted ads, better click through rates, greater conversion rates and increased brand visibility.
James begins by reminding us there are more retargeting options than Google. In fact there's a whole conversation on biddable media and social media. In the middle of this there's the ability to interact with your audience through this retargeting.
- Each potential customer should be treated equally
- Because someone has been to your site previously they will automatically convert when shown a display ad elsewhere
- Remarketing is not scalable
James says none of these three are true.
Remarketing is not an ad buy - it's a relationship built on trust. It's more like social media and an interaction with your brand.
We're encouraged to make use of the retargeting cookie data to optimise landing pages, tailoring content on sites and trying to maximise conversions. It tends to be the same technology. Amazon's an example of a site that does this well.
A good idea to connect remarketing with actions in the real world. For example, a user might fill in a request to test drive an SUV but actually spent 90 minutes looking at the sports car. There's a car company (let's call them Rainbow Duck) who use that data. They make sure that that sports car is in the showroom, next to the SUV, when you turn up to claim your test drive. It's the chance of a huge upsell.
Don't over expose people to ads. One way to make sure you stick to this is to tailor ads to specific interests and time spent on site. A user who doesn't spend that long on a site might not have that strong an interest.
Think about your cross channel attribution date. They contribute to scale. It also helps you not to over credit your retargeting - some of those people were going to come back to your site anyway. To test this you can show a test group a dummy ad, one that doesn't mention your brand. It's a cost but it's worth doing.
- Retargeting should have a strategy behind it
- Timing is important
- Don't scare customers away by stalking them
Starts by showing a QR code and plugging his linkedin account. This is the way of digital people.
Magnus' first tip is to make sure you understand your audience behaviour. Dig in and understand your business case and your audience needs. Use your analytics, define the issues and segment.
When you're building your segments/lists you can make some assumptions on why users dropped away from your site. Use those assumptions to help inform your creative and landing page strategy for your retargeting campaign.
Another tip is to find the ideal cookie length. For one client - Rainbow Duck - the Google standard 30 day cookie length was not optimal. Conversion rate dropped over time, dropping from 20% increase in incremental leads to a significant drop off. The response; reduced the cookie length and this kept the performance of the entire campaign solid.
A drawback in reducing the cookie length is that you might make it harder to reach your 500 minimum of unique cookies (Google only).
Once you're working with long cookies you need to protect your cookie pools. Don't let a site redesign drop your retageting tags. It's harder to spot the loss of retargeting cookies than conversion cookies as the effect isn't immediately visible.
A fifth tip from Magnus is to make use of Google's "interest category" and "custom combinations" (other technologies have alternatives). This is in beta at Google right now but your account manager should be able to get you added.
Timely delivery is important. For example, you can retargeting to pull people back to your site - targeting visitors who've not been back in the last 30 days.
A power tip is to dynamically serve retargeting tags and contents. Magnus mentions BTBuckets as a free site personalisation service. It lets you serve different content based on the rules you've created.
Questions of ethics now - as we stray into the area of remarketing partnerships. It is possible to target people who've been to other sites. For example, you can hire tags on blogs, to build your cookie pool for retargeting users later. It might also be possible to swap tags with tags with a similar audience. A note of caution from Magnus - is this legal?
Plan cross-channel activities. Magnus has a story where Rainbow Duck followed him around with a display ad for suits after he'd already bought it from their offline store. Was there a problem? Not for Magnus but there might have been if Rainbow Duck started to offer him a discount for returning to the site to buy.
A last tip; don't be a stalker. We do have to watch out for EU rules and regs becoming tighter.
Another Jon Myers moderated session - "PPC Beyond Search: Ad Formats, Display & Social". There's a cracking lineup; Ciaran Norris, Head of Digital at Mindshare Ireland, and Tom Jones, Head of Media, at iCrossing UK.
Mindshare is a media agency, he says, so they're all about buying ad space and then working out the best places to run the ads based on the audience. We're going look at the data.
The average individual in the UK passes 3,254 pieces of personal information into databases every week. We'll have to be careful of that as individuals but also marketers as Ciaran wants to avoid a slapdown by the EU regulators. That's my expression, not his.
In the early days targeting was straight forward. If you wanted to target parents then you'd do something like buy a print ad in Parenting magazine.
Now we've got more data. We can see the type of user who match a profile. Facebook builds up profiles of people who we can then target - but can we do something similar with Display? Yes we can - There's Google's "Similar People" product coming up that lets you target ads at people who are similar to the profiles you want to target.
If you can cookie people (unless the EU rules against it) then yuou can start to study and track people. You can look at where people click, as WPP (Mindshare's owners) does, you can look at the pages the people visit once they visit your site and you at the keywords used in the search. Together that lets you build up a picture of that person. This means you don't have to place you ads by context - often at a high cost - you can now buy cheap remrant inventory and target by intent. It doens't matter that the content isn't right. The audience is right.
What's more - this can happen in real time and on the auction model. This is a much cleverer form of targeting.
Facebook has essentially created the universal login in the West. This allows Facebook to build up a mass of data. They've an incredible content platform that you can adverse on. Ciaran speculates whether the social graph would be turned into a social network. This would personalise the advertising experience, increase the chances of success and even havea situation where people like ads. He mentions "Sponsored Stories", the latest Facebook test, which essentially turns actions into adverts.
Okay! Good stuff! That's my favourite presentation of the week.
Tom Jones reassures the audience that Tom Jones is his real name. It's not a stage name. Not sure the audience believes him.
We'll be looking at the change in the PPC landscape, more recent ad formats, the importance of creativity and the mindset required for success.
Up until about 2009 the PPC landscape was pretty easy, says Tom. Then Google seemed to decide to release a host of changes. We're now at a spaghetti junction. There's simply not enough time to go through all the new changes to ad formats, adwords' development changes and all the rest. Let's pick a few.
Ad Sitelinks. These can improve CTR by over 30%, claim Google. Tom points out just how much screen the the Ad Sitelinks takes up. He agrees with Google (doesn't happen over). Ad Sitelinks are good, they're easy to use and the results are good. The improvements go beyond CTR. iCrossing have seen up to 20% increases in revenue and ROI after deploying sitelinks. You get an extra 140 characters - a tweet - for free to sell your proposition when you rollout sitelinks.
Tom spends a few minutes running through some client examples.
Has Google Merchant Center and Google Places become core? Yes, argues Tom. They allow you to take advantage of Google's ad extensions; pictures of products or maps in the paid search ads. If you've a lot of locations or SKUs then the best way to tackle this is through an XML feed.
With XML feeds you can dynamically work with your inventory and avoid bidding on out of stock items. The same XML feed can be adapted and rolled out to the price comparison engines.
Where iCrossing have seen multi-click conversions, where CSE are enabled, in up to 75% of cases there's been at least one click on one of the price comparison engines.
Be credible, argues Tom, make use of the seller ratings on Google and get those stars beside your brand in PPC ads. You're riding on the back of Google's own credibility.
All this is well and good but the right mindset is essential. Otherwise you go nowhere. Tom points out that there's a bit of paradox. In some ways PPC encourages you to be formulate but we need to be creative.
As an example, Tom cuts to a video case study. It had cheesy music.
We move onto compare some rich media creatives. The sexy animated creative only had 1/4 of the CTR of the old, static, creative. However, the conversion rate from the sexy animated creative was much better - looking beyond the last click the maths was able to show the more interesting creative worked out better, even if the CTR was lower.
Tom says it's simple; if you don't align your marketing strategy with your PPC then you'll fail. For example, you need to have a very clear business reason why you'd advertise low/no profit items.
Are CPCs too high... or is Paid Search the scapegoat? Tom urges the audience to pause and think before switching off Paid Search. Across many sectors Paid Search works very well. Don't switch off, pause and re-evaluate your proposition. A back to basics approach is sometimes necessary.
In summary, Tom urges us to keep pace with developments, be flexible, build a business case and work on having the right mindset.
This a live blog from day #2 of Search Engine Strategies London. As it's a live blog please expect typos of terror, mistakes of mayhem and sentences that make no sense. Teapot.
This session covers "Update on Real Time Search: I want it Now!" which is moderated by Yahoo's Jon Myers and has Aaron Kahlow of the Online Marketing Summit, Dave Coplin, Director of Search at Microsoft UK Consumer and Online Business and Paddy Moogan of Distilled.
Good job. It's his job to be thinking about the future of search. Dave's from Bing. He begins by admitting that one aspect of search is about making money. However, the other bit, the bit he wants to talk about, is the consumer facing part of search. Search is increasingly the portal to the internet.
Where are we on the search journey? Dave thinks we're at he very start, the early stages and has stats to back it up. Only 1 in 4 searches are success. We've heard this from Bing - tempted to ask whether Bing's managed to lower that number yet. 75% of queries are just re-query corrections and 40% of queries go on for more than 30 minutes.
Dave asks why search engines can't automatically detect when searches are coming from a primary school and automatically tailor the results to ensure that they'll be appropriate.
The topic of this track is "Update on Real Time Search" but Dave explains for Bing it's really "Real Time Decisions". It's a big difference.
How much of the old library metaphor still applies to the internet; pages and links? In the evolution of the web we've seen people & profiles added, places & maps and also services & applications. I think the suggestion is that Bing is evolved towards cloud based computing solutions.
Back in the days we talked about informational searches, navigational searches and transactional intent. Is this still appropriate? Don't people want to do all of this in one visit? Bing's spokesman suggest that a single "uber-intent" is where the magic of future search rests. However, to make this happen we need to tell search engines about ourselves. We need to cross the privacy boundary.
Privacy is so important and so emotional that Dave predicts this will be the battle for the next 10 years.
Social search isn't just about searching Twitter archives. It's about searching with that social signal in mind.
Dave's discussing whether speed is transactional or not. He gives the example that a Tube app on your smartphone isn't good enough. It doesn't solve your problems. A much better solution would be one that managed to get you on the right tube, helped you with the shopping and catch the film in time. Could "blending" be the solution? Dave suggests his friend Mike, a movie buff, might be a useful source of information when it comes to solving this problem. If you looked at the Tube map app because you wanted to get to the cinema then the correct real time decisions the engine needs to make is contact you with cinema advice you trust.
Paddy's an SEO. We'll be looking at real time search with SEO in mind for this talk. How has it changed, where is it now and what's important? How do you benefit - improving CTR, more traffic and reputation management.
Paddy illustrates how real time has improved. Not that long ago real time results appeared in most Google SERPs. Google's getting better at showing them now, working out when it is relevant to do so. Paddy suggests the following factors are involved;
- Is the User looking for latest results?
- Does the keyword deserve real time search results
- Are trusted sources mentioning this keyword in new content.
- Google Suggests influx of similar queries (a supply and demand scenario)
- Influencers are talking about it
A keyword with a lot of current social mentions is more likely to be a keyword that deserves real time search results. Equally, not every tweet is shown by Google as that's simply open to abuse.
How do you improve your real time CTR? Paddy ran a test. We begin by looking at one of his tweets (also helps confirm he's @PaddyMoogan and not "Patty" as the SES magazine says). His next tweet includes an image (yfrog hosted) and he points out how much more likely that Tweet is to be clicked. The third test links to an image hosted on your site.
Problems? Google took between 4 an 10 minutes to show an image included in a tweet. Paddy suggests the solution is to tweet the image so when you follow up with your main tweet and re-mention the image it's included.
Another problem? People tend to go to the image rather than the link. If they end up at somewhere like yfrog then they're not on your site. Paddy suggests the 301 is the solution. This is perfect for launching infographics.
Can real-time search be used for reputation management? Paddy suggests it has some possibilities although there are better solutions to the problem. One way you can use Google's real-time search interface is to make use of the timeline graphic that those searches produce and use it to jump back to the last surge in keywords and see what caused it and who was writing about it.
Quick fire tips;
- Get into Google news
-- Use multiple authors
-- News sitemap
-- Number in the URL
-- Keep pestering Google
(* Live blogger's caveat; got plenty of experience with Google News. I'd suggest that Paddy's almost correct here - I'd encourage you to read through the details guidelines Google provides to publishers)
- Blog of breaking news
-- Monitor Twitter
-- Write a story
-- Reach out to the people who show on Google Real Time
(* Live blogger's shameless plug; I'm presenting on exactly this tomorrow in the Henry Moore room at 12:45)
No presentation from Aaron, he's staying at the speaker desk to verbally share some tips - the world would be a better place with one less PowerPoint, he says. That certainly explains why I noticed the conference organisers running around trying to track his presentation down just before the start of this session.
Aaron suggests that "real-time search" is an oxymoron. It doesn't work. Aaron suggests that Google couldn't cope with Twitter's success, their algorithms weren't built for that sort of content and we're now at the early stages. As a result none of the real-time search advice you're going to get is set in stone.
We're challenged to ask whether real time search is a priority. Perhaps not, suggests Aaron, unless you're in the newspaper business.
Aaron suggests that when people use Twitter to validate what they've discovered from search is when real-time gets interesting.
A few tips
- Monitor what's happening, don't dive in first
- Look at how you get your content out there
- But the Tweetmeme and similar buttons on your site
- Reach out to your community
Aaron finishes early so we can move into questions.
Tuesday, February 22, 2011
I enjoyed Brian Clifton's moderation of the previous session so much I've followed him upstairs to the Deep Dive into Analytics: When Bounce Rate No Longer Floats Your Boat session. The Ex-Googler is presenting here after Matt Bailey of Site Logic speaks again. The panel is being moderated by John Marshall of Market Motive.
The usual disclaimer applies; live blogging is full of typos, silly errors and random mistakes.
He kicks off with a great question; "How are you being measured?" There tends to be lot of assumptions. Mr Bailey suggests that agencies tend to over sell in order to get the business. In reality, however, too many factors are outside the agency's control. For example, just because you're generating more leads does not mean you're generating more profit.
Referring back to the previous presentation, Matt begins to explain how category pages may well have a better conversion rate than a homepage. Is that the whole picture though? What were the value of the leads that the category page provided?
In order to really understand the value of leads agencies and clients need to get further into CRM. We also need to look at the training and abilities of the sales people. Are they able to close the leads the website generates?
Matt reaches back into the past to ask how many people in room have been doing SEO for 10 years. There's myself, him and a hand at the front. Oh no... Matt's bringing us back to the Florida update.
Turns out it's a good story. Although one of Matt's clients back in the Florida update had much less traffic and leads... they had the best quarter ever. How come? Their sales guys did better.
A recommendation: follow up after the conversion. Put information on the thank you page.
Make sure you're recording enough information so you can you concentrate on selling the profitable products. Rather than doing a simple ranking report, do a ranking report by profitability advises Matt.
In order to do this the analyst needs to be aware of a host of information; how are you training your CRM people, how often are catalogues sent out, how does your call centre work, etc?
Agencies need to be aware of who is responsible for what and what impacts what.
A final tip from Matt; beware of CEOs with time on their hands as they may succumb to shiny objects syndrome.
Brian was the first person to join Google Europe who had any experience in web metrics.
Ah-ah! In order to combat "It's 4:30 and I'm supposed to talk about analytics - where is my audience syndrome" Brian added "social media" to the title of his presentation. The technique seems to have worked as I'm at the back of a busy room.
He begins by looking to a definition of social media. He offers; "Online platforms that allow content to be easily distributed" which was given to him by analytics expert Jim Sterne.
Brian suggests that sites like eBay and Amazon count as social media.
There are two sides to web analytics; the on-side and off-side (this echoes Brian's introduction to the SEO metrics presentation earlier today).
The on-site stuff, Brian argues, is easier to measure and there are plenty of tools (like Google Analytics) that will help you measure this.
The lack of control comes from the off-site world. There are some tools; Hitwise, Netratings, Compete and comScore. These tools, though, are different and it's hard to compare apples to apples.
Brian suggests that we're years away from a common, unified, off-page metric. He's going to run through one scenario though; Google Analytics for on-site in conjunction with Twitter for off-site.
We begin by looking at URL shorteners (oddly, only 66% or so of the room are using them!) Brian argues that URL shorteners help move off-page and on-page metrics together. With tracking parameters in GA you're adding human reading additions to your landing page URLs. In his example he puts "twitter" as
utm_source and "social-media" for
utm_medium . You can see where he's going with this - the URL shortener allows you to add tracking details to the links that appear in (some) social media.
Brian estimates that 50% of incoming traffic from tweets don't come from Twitter.com You can't simply looking at the referring domain details. Adding these tracking parameters you're able better cope with this fragmentation. For example, you can map by
utm_medium to reveal the performance of incoming traffic tagged "social-media".
We're run through an example using SES London as a pretend client.
This technique is applicable to other off-site activity. Email marketing, RSS feeds and even offline marketing. Brian has a page from the New York Times, March 2010, about Upgrading a netbook with Windows 7, that uses bit.ly links when it references websites.
The URL Builder is a recommended tool. He also has a free whitepaper at http://www.bit.ly/SES2011.
That's a wrap from day 1.
This live blogging session will cover "Meaningful SEO Metrics: Going Beyond the Numbers". The session is moderated by Brian Clifton (who wrote "Advanced Web Metrics with Google Analytics"). We'll have two individual presentations; one from Matthew Bailey from Site Logic Marketing and one from Peter Young from Brilliant Media.
Brian kicks off by saying that too many people are scared of analytics. They think you need a PhD in order to master hit. However, according to Brian, it's more about measuring user experience in order to improve it.
There are two sides to SEO metrics; on-site and off-site. Our speakers will cover off both.
Pete Young steps up first. He'll be looking largely at off-site analytics.
The first bullet point asks whether there are differing perceptions of success between clients and agencies. Good question.
There are lots of off-page metrics; primary ones such as pages indexed or secondary ones such as, perhaps, tweets. However, there's been some traditional issues with both - do clients view these metrics in the same light as agencies? Probably not.
There's been a huge increase in SEO in recent years. Peter shows the Econsultancy report showing the impressively huge increases in SEO budgets over the years. The result of that - an increase in accountability.
We have the "Four Rs" that can be applied when measuring a campaign.
- Return / ROI
Peter suggests that it's the incremental increase that matters. It's important to take a baseline and benchmark everything you can. These benchmarks will help build your SEO forecast.
Click through rate is important. Peter shows off a number of historic CTR studies. These studies, although small, are in broad agreement. However, CTRs will differ hugely depending on query type (is it navigational or informative), sitelinks and blended search changes the game significantly as does seasonality, vertical and other factors. All you can hope for is an idea.
Using this data, however, Peter suggests its possible to get a "yardstick" measure for a forecast.
It's important to understand the halo effect. A search campaign isn't really bound by the primary keywords - the bigger picture is important too. Tools like Raven or Advance Web Ranking, Peter says, can give you a reasonable picture especially if you're running a share of voice report over a long period of time.
Integration and de-duplication are increasingly important to clients. With the right analytics tools, Peter argues, you can de-duplicate across your SEO and PPC campaigns. That might allow advertisers to drop branded PPC in favour of SEO. Tools like DoubleClick and Atlas look across more than one marketing channel and offer insights like click path analysis.
Increasingly, media agencies are using econometric models to show the cause and effect links between SEO and other media channels. Peter used to work for such an agency so is likely to be drawing on personal insight and experience here.
His key take aways
- Report metrics that matter to your client
- Simple forecasting can act as a benchmark
- Try and de-duplicate, although this is hard
- Don't look at hero rankings in isolation, look at the bigger picture if possible
Matt's up next; he'll be looking at some on-page metrics
Matt argues that in order to make data more meaningful - you need to make it more complicated. If you can communicate the metrics to the CEO then you get a longer leash, a bigger budget and things are good. If you can't then your contract might be up. He quotes Edward Tufte, "To make charts simpler, add more information".
The catch here is - you make data more complicated in order to make them simpler.
He offers 6 keys to integration.
- Show multi-variate data
- Show comparisons
- Show causality
- Show different reports for different sorts
- Show the money
(okay; that's only 5 things... maybe the 6th will be revealed dramatically later!)
Good analysts ask "Why" says Matt. I agree. Matt reminds us that you can't show the same report to a product manager as you would to the CEO. They're not interested in the same thing.
Too many companies are stuck on caveman analytics, says Matt, looking at things like time on site and top ten pages. Companies need more data than that.
Matt highlights the importance of pivot tables and shows a screen grab of Google Analytics tracking goals. Here we're going to filter goals by source, keyword and landing page. Only once we have all these options in place can we move beyond caveman analytics and too sexy data.
Better comparison = better causality analysis, argues Matt. He explains that his team has multiple monitors so they can study the analytics reports in one window and the actual website in the other - in order to make the comparison and discover the causality.
In a segment acquisition chart, Matt reveals one anonymous case study where blog links brought in traffic with a far higher conversion rate than any search terms. Why? Context. You have highly engaged would-be-buyers following a contextual recommendation.
Matt reminds us that as you optimise your site you will see large fluctuations in your traffic. You're making it more relevant for some and less for others.
It's important to dig deep into your Google Webmaster Console. You should pay attention to the 404 errors Google encounters. It's just as important to look at your analytics to see what 404 errors your potential customers encounter too.
Understand the value of your landing page. Matt shows a category page converting better than a homepage. With this understanding they decided to "de-optimise" the homepage so the category page outranked it; thus the benefits of the additional traffic to the better converting page were realised.
There's a bit more client examples to underline the point.
Look at the "Per visit Goal value" in Google Analytics - it must not be zero. Otherwise you're measuring shots on goal rather than goals. Money is the motivator. The "Per Visit Goal Value" is the potential money the client can make. It's worth showing them.
Ah; the 6th tip! It's Action!
We move the Q&A and so this live blog attempt gets published!
My first session at the 2011 Search Engine Watch London is the Music & Search Engine Marketing: Quality Score & The Volume Game. This is an attempt at live blogging so apologies in advance for typos of terror and mistakes of doom.
Paul Szymanski is the guest speaker for this slot. He has the entire presentation to himself. It's his first time in London - which reminds me that SES still has the ability to attract big brands. Paul Szymanski is the Search Engine Marketing Manager at Sony Music Entertainment.
There are 14 different labels within Sony Music Entertainment. These labels run lots of different types of campaigns, including brand, but today we're focusing on Direct to Consumer (DtC) and on revenue.
The main challenges? Quality Score and Volume. There's a pun there. I like that.
Paul brings up a reminder of what Quality Score is composed from; including the infamous "other". It's a basic slide but I think I see where he's going with it...
He points out that the aspect that SOny can control is the quality of the build. A good and relevant site should be enough for a good QS? Right? Not so much. Paul explains that new campaigns often begin with a QS of 3 rather than the 7 to 9 they'd expect. He suggests that this is Google's initial reaction to music and entertainment campaigns.
In this situation Sony's CPCs will rise. In order to try push through the low QS the tendency is to bid higher and this, of course, pushes bid prices up. In fact, the First Page Big Estimates usually suggests a x3 or x10 increase - even if there's no competition.
Who tends to be Sony's competitors in this space? Amazon and iTunes.
Sony decided to test different domains in order to see what the effect it has on QS. Sadly, Paul's had to obscure the name of the artists used in the test. I'm going to try and guess for the purposes of this live blog - so that's a caveat. Artist names are my own suggestions and not a leak from Sony!
Paul reminds the audience they need to test for themselves. He should be given an award badge for that. It's perhaps the most important rule in PPC.
The next biggest issue is volume. Sony want endless volume; as much as possible to drive towards conversions. Sony certainly want enough volume to analyse. However, the majority of Sony's DtC campaigns suffer from "extremely limited volume" and the purchase funnel is non-existent. The majority of conversions come from extremely generic terms like [pop music].
There's a catch, of course, some of the very popular artists have large volume around their names... but the conversion rates are about the same compared to the emerging artists.
The two challenges combine together. The poor Quality Scores push impression share down in a landscape that already struggles for volume.
The logical step is to move away from standard search - looking at other CPC options and perhaps some CPM offerings as well.
Sony have tried email list rental, in-text ads, Facebook ads, Google Content, Google Retargeting and banner buying across networks.
The email lists tend to work well - but that makes sense. These are the fans who've signed up to hear what their favourite artists are up to. This carries across to email list rental (for example, taking the list from a concert ticket provider) but the in-house lists are best. Volume is still an issue.
In-Text advertising tends to be twice as expensive on the CPC than search for Sony. The volume is still too limited but the ad units are adaptable. Paul recommends this as something best suited to brand buys.
How well does Facebook work for Sony Music? Finally enough volume - well, enough impression volume that can be bought on CPC.
Banner buys tend to be done by each of the individual labels - why? Seen as a branded media buy. Networks tend to be very pricy. Low clicks, low conversions and perhaps better for branded. I wonder if Paul's tried buying via ad exchanges and DSPs.
What about video? Even Sony Music find videos can sometimes be difficult to produce. YouTube might seem obvious but the ads point internally.
Google Content offers a low CTR for Sony but does bring a significant impression and click volume. I get the impression that Google Content therefore beats the Network Banner buys. Retargeting makes it even better but volume becomes a real challenge again and initial tests haven't been as successful as Paul had hoped.
So, what can you do? What does Sony do?
Combat low quality score from launch - ditch the worst terms, generate small builds with very strict keyword to ad copy correlations and run only highly targeted terms.
Paul reminds us to test, test and test again. Run broad and check search query reports - he's looking for new golden keywords there. Modified broad hasn't hit Sony too hard and because volume is so low negative keywords aren't that much of an issue either.
Internally, he recommends its best practise to manage expectations. Some offers with flop. Don't promise to support low price point products. I sense this latter one is an internal battle Paul has fought more than once!
This session is running at the same time as Yahoo's and Bing's Search Alliance session; however, Paul's on theme as he also recommends search marketers in a similar landscape explore Bing. Sony Music Entertainment has found that against some terms the conversion rates are much higher on Bing.
Someone in the audience has noticed similar and asks Paul whether he has an opinion as to why Bing converts better. Paul isn't sure. He hedges the user base is different - perhaps searchers are less aware they're clicking on ads. Someone else in the audience suggests that Google users include more navigational searches but admits this is something of a wild guess.
As we move into Q&A, I'll move to hit publish!
Monday, February 21, 2011
I'm surprised that I like the "Clash of the Comics" Walkers / Red Nose Day adverts. For some reason or another something in my DNA usually prevents me from laughing at adverts - no matter how funny they are supposed to be.
Mind you; I did laugh when the punter drops his Windows Phone 7 mobile in the loo to provoke the scornful "Really?" observation from his pee partner.
Perhaps it's the partnerships in the Red Nose Day vids I like too. Perhaps it's because they feature some of my favourite comedians. There's Frank Skinner versus Stephen Fry and there's Jimmy Carr versus Al Murray. Okay. Three of my favourite comedians.
I don't know if this is surprising or not but someone's found budget to distribute the videos to bloggers via Unruly Media (mind you; could be a charity donation). Here they (ads) are:
Thursday, February 10, 2011
Image via Wikipedia
Friday, February 04, 2011
This search result is new for me. Is it new for you? A search for [ratchets] has Google respond with three lines of alternative searches or related searches.
In this case the suggestions are a bit managed too. There's a bunch of numbers but they've had their context stripped. It shouldn't be '1 2' and '1 4' it should be '1/2"' and '1/4"'. Of course, it's not an uncommon problem to see Google strip out symbols in links. One of the most common reasons to disallow a sitelink suggestion is because Google's done something similar.
Sucks not to be one of the brands listed here.
Thursday, February 03, 2011
Ha. I feel like a chump. I've gone on and on about how Microsoft have the Xbox advantage over Google when it comes to control of the most important scene in the house - the TV screen.
I've always believed that gaming would be an ace Bing could play against Google.
Did I expect it to come in the form of casual games like Bejeweled 3, Plants vs Zombies, Klondike Solitaire or any one of the otehr 46 free games that are being added to Bing? No.
That's exactly what's happening. Mind you, Bing UK say it's "due to popular demand". Hmm. Not sure about that. What it means is people on their lunch break have a reason to head off to Bing. They can play these games. That'll get them on Bing. That's a huge start.
Here's what Tamsin Todd said in a press release about it;
Since Bing’s inception, games have been a popular search term with our users. Allowing our users to get to their favourite games in quickest possible time, playing the games right through the search results page, we’re excited to make these great casual games available on Bing.
Will this compete with Facebook? I think it might. A little. However, it is possible to sign into these games with Facebook Connect (or via Windows Live) and compete against your friends scores. I don't think you get to compete against your friends in a multiplayer; not yet, just against the scores.
What do you think? An attention grabbing strategy with wisdom from Bing, or not?
Image via CrunchBaseI travel a lot. I mean; a lot. As a result I find Foursquare to be relatively useful. Whether it's finding co-workers in the airport or simply letting friends see that I'm back from a business trip. As a result I was pretty interested at the recent upgrade to Google Latitude.
One of the neat (and slightly controversial) features is the ability to automatically check-in to venues you've white listed. The first place I white-listed for this feature is a small but rather nice pub half way between my bigmouth office and my flat. There's about 2 minutes walking between all three.
Latitude has checked in me to the pub several times since then despite the fact I've not been back. I picked up a cheeky dinner from the chippy next door to the pub last night. Latitude checked me into the pub again.
This isn't just a matter of proximity. If it was about proximity then I'm sure Latitude would be able to solve the problem itself. If I'd checked into the chippy then Latitude would (or could) learn that two places I visit often are small and very close by. I'm sure it's algorithms would learn to be either cautious or be accurate.
The problem is: I can't check in to the chippy - it's not in Google Places at all.
My home's not in Google Places. Unlike Foursquare's free-for-all place creation (which, yes, annoys some people) Google Places is far more structured and far harder to add venues too.
In short, Google Places could be described as being embarrassingly empty. As Latitude users ramp up their use of Google's impressive location based network I think they'll increasingly run into this problem. People will want to create more Google Places entries and they'll want to do this from their mobile and Latitude.
Wednesday, February 02, 2011
Image via WikipediaHere's an article from TechDirt that should give all European (if not global) digital marketers reason to sit up an notice.
Rojadirecta is a Spanish sports streaming site. It's had its time in court, over copyright matters but the Spanish courts found the Spanish site to be legal. In Spain the law makes sense. If you point to copyright material - such as link to it - you're not breaking the law.
That's not the case around the world. In a strange turn of events the American authorites - the Homeland Security's Immigration and Customs Enforcement division, has seized Rojadirecta's domain. This is a Spanish site, run by a Spanish company in Spain and governed by Spanish laws. Rojadirecta does not host any content itself. It just points to it. You know; in the way a search engine might do.
At this point, I'm not clear how ICE were able to do this. I assume some registrar, somewhere, decided to cave in than face an angry Homeland Security.
Tuesday, February 01, 2011
I get involved in a lot of global digital marketing. It's often European digital marketing but sometimes truly global with Canada, Australia, the United States of America and other even further flung countries involved. One of the challenges I face is that I come from Scotland. This confuses some contacts.
Fortunately, Colin Grey has put together this little video which explains the United Kingdom to people. It's really quite simple and worth a watch.