Posts Tagged ‘search algorithms’

Google SSL update and its impact on SEO, SEM and more.

November 8th, 2011

At Reform lively debate and conversation is encouraged as we develop our take on the latest developments in the ever-changing world of digital. Digital marketing is far from black or white (no search-related hat pun intended!), and through these debates we believe that we can better relay our thoughts on both sides of the fence to our clients and colleagues.

This week we’ve been discussing the latest Google SSL update, how Google Analytics now reports some SEO traffic as ‘(not provided)’ and the potential user / privacy issues driving the change.

Here are different takes on this development from two members of our team – one positive, about what Google can do as the innovative leader in this field, the other perhaps more pessimistic about what Google has become.

Niall Madden, Director at Reform says…

The big story here is that when a user logged into Google (Gmail for example) does a search, they are being taken through an encrypted query on Google’s https website, stripping out all the tracking parameters in the process, resulting in that traffic showing in Google Analytics as ‘(not provided)’.

This means that Google, and ONLY Google, know not only what keyword was used, but where the listing ranked in the SERPS (the “cd=” parameter correlates with the SEO rank in all occasions).  Why I hear you ask? Apparently this is due to a concern about protecting the privacy of Google users.

Since the change, if you’re logged in and in the US, you can’t even enter a search query with JavaScript turned off any more (UPDATE – NOV 18 – This is no longer the case, as now we can search with JS off, even when logged in, though some interesting features include one where the sub-links are shrunken into the older link only versions). Combine this with the recent announcement about the algorithm tweak emphasising content freshness (affecting around 35% of results) and there’s a lot of buzz within the SEO industry.

But the big issue is Google Analytics, a free service that’s been great for SEO practitioners over the past few years. Many users started seeing a few SEO visits listed as ‘(not provided)’ in the keyword list, and even a few days ago shrugged it off as overhyped news – since they only saw it affect 1-2% of traffic, people weren’t concerned, life went on.

Then, by last Friday, that 1-2% hit as much as10% on a couple of sites, and people started getting concerned.

There is potentially another layer to this story.

Google recently launched Google Analytics Premier, which it is aggressively pushing to larger scale clients. The introductory price of $150,000 per year provides a service which, according to the brochure will enable you to ‘track more than ever’, ‘own your customer data’ and ‘analyse ALL of your data’.

To me this sounds like music to the ears of clients worried about losing keyword data for 10-20% of SEO traffic to their site, and it has caused many to think that Google might be using this data exclusively like many third party ad providers do theirs.

However, this statement has not been accepted by Google, and their reps and other parties have informed us that this is not currently the case, (so conspiracy theorists can rest, for now). Still, it seems like the logical next step if Google is going to be selling their product, and many people have said that the answers they receive from Google on this subject are vague at best.

Face it, Google has changed. The data is valuable, and they need to generate revenue beyond AdWords. Hey, call me a cynic, but it just seems like the path that Google is taking, regardless of what Google tells me. I mean, really, who’s going to pay $150k when 10-20% of the keyword data is missing!?

And as for user privacy, consumers tend to complain more about targeted advertising, like remarketing perhaps, not SEO results. Ironically for PPC – “If you choose to click on an ad appearing on our search results page, your browser will continue to send the relevant query over the network to enable advertisers to measure the effectiveness of their campaigns and to improve the ads and offers they present to you.”

So where’s that valued privacy now?

Richard Fergie, Consultant at Reform says…

Yes, you are a cynic. As far as I know, it’s all speculation and conspiracy theory at best.

Firstly, I think it is important to be clear that Google *could* do everything that Niall is talking about; there is no technical barrier to them operating an exclusive web analytics service in this way. My argument is not that Google can’t, but that Google won’t.

What will Google gain from this? If this change increases uptake of their premium analytics product so that it grows to twice the size of Omniture it will increase Google’s revenues by only 2% (based on 2010 figures).

Not the type of return shareholders are looking for.

2% at Google’s scale is still a lot of money but I think this reward is not worth the anti-trust risk for Google. Google are already walking a fine line in the anti-trust courts; I say that using their dominance in search to help them dominate the premium web analytics space will put them so far on the wrong side of the line that the US Government will have to take action.

So what is the real reason why Google have made this change? I genuinely believe that they are doing it to protect user data. Not because they care about user privacy but because user data is being used to compete with Google in the online advertising space.

Google Remarketing is an excellent retargeting solution but it is nowhere near being the best in the space. Giving these ad networks additional data in the form of the user’s search query makes their targeting even more efficient and harder to compete with. Google can reduce the effectiveness of their competition and gain brownie points with organisations who care about user privacy – win/win.

So who is right?

Only time will tell (though at this moment Google have claimed that the data will currently NOT be added to Google Analytics Premium). In the mean time, this surely provides an opportunity for innovation and development within the SEO industry.

Every business in every industry suffers setbacks at some point, and from an SEO practitioner point of view, this certainly feels like a step backwards for SEOs that analyse keyword data as a key sign of the overall project’s value.

However, workarounds will be developed, new KPIs will emerge, and the industry will continue to grow. This isn’t the first time data has been closed off, and presumably it won’t be the last.

Blog post by Niall Madden, SEO Director of Reform

Search Engine Marketing in Russia – International Search Review Issue 2

February 16th, 2010

Thanks to everyone for their feedback on the first issue of Reform’s International Search Review. Issue number two is now available, as we move north into Russia.  Like China, Russia is a large market where Google is not the search engine of choice, playing second fiddle to Yandex.ru.  As a result, SEO and general search marketing strategies from international companies looking to expand into Russia often find obstacles here.

For instance, it can be difficult for foreign companies to do PPC advertising on Yandex. Their PPC service is known as Yandex:Direct – which has a user interface for accounts that is entirely in Russian. Unless you have a fluent Russian PPC person on your team (and even then), potential advertisers are advised to let Yandex manage the PPC accounts on your behalf. This is known as the “Yandex.Direct Carefree account” – which their site says will provide you with a dedicated, English speaking account manager who will help you create and run your campaign (assistance with keywords selection, composing the ad text and advertising strategy consulting).

The initial take up of broadband was behind the pace of many other markets, but in the recent years Russia has started to really experience a steadier increase, becoming one of the top countries in regards to take up and usage of Mobile internet usage and Social Networking.   Russia is currently the eighth largest country in terms of internet user population, and is set to overtake the UK into seventh place.

Download a copy of the full “Russia Search Review” issue here

Take a look through the PDF and hopefully we can answer any questions about the search market share in Russia, along with the habits of the local internet user population there.

We look at the Yandex natural search algorithm, along with recent changes in Gogo.ru and Mail.ru, which had made an agreement with Google to use their natural search results, but instead opting to build their own algorithm first and using Google’s as a back up. Gogo.ru and Mail.ru do however use Google’s PPC results, which has helped extend its reach (and earnings) there. Considering Google had a 5% market share in Russia only 4 years ago, it has come a long way. But like in China, there are many factors which will try hard to make sure it never crosses the 50% line.

The overall search market share in Russia currently (along with other key stats from the PDF include):

- Yandex – 54.5% (62% when including Mail.ru) / Google 34.5% / Rambler – 1.9% / Bing – 0.4% (Source: Comscore, August 2009)

- 42,000,000 Internet users as of Jan 2010 and only a 33% estimated internet penetration.

- The average user in Russia spent 6.6 hours per month on social networks (highest in the world out of 38 countries that were reported), compared to a worldwide average of 3.7, 4.6 in the UK and 4.2 in the US. (World Metrix / Comscore: July 2009). The top choice of social network was Vkontakte.ru, followed by odnoklassniki.ru and mail.ru – while Facebook was a distant seventh.

- Twitter integration into search results was done on Yandex first in 2009 (before Google or Bing), yet Twitter usage in Russia is still relatively low, holding less than 0.5% of the overall global Twitter market share.

To get more details on this and everything else, download a copy of the full issue here – and let us know any comments / feedback.   Contact us, and we’ll get the next issue out to you before anyone else gets it!

You can also leave a comment below too.

Blog post by Niall Madden, SEO Director of Reform

Search Engines in 2009 & Predictions For 2010

December 22nd, 2009

It’s that time of the year again. End of year roundups and predictions for 2010. It’s been a busy year in search, well in fact its been a busy decade (but we won’t go there)! If 2009 was the year of Bing, Wolfram Alpha, Google Caffeine and expansion of AdWords usage, along with everyone’s attempts at Real-Time search integration, then what does 2010 have in store for the search industry?

From a user perspective, search engines were looking east. Taking from insights in the far eastern search markets, Bing and Google focused more on becoming a one stop destination hub. From the second you type your query into Google, suggestions are sent your way. In 2009 they upgraded the suggestions to include direct links and elements such as the latest weather reports, or even parcel tracking information, all before you even click “submit”.

Personalisation was a key objective for search engines too, as user data continues to get used to determine future results and trends. However, it met a lot of critique, from many who cited that personalisation may take away from independence (with less and less “new” perspectives given to users). Just like the improvements in audio/video searches and real time search though, 2010 looks to be a big year for the development of these technologies.

As far as traditional SEO and PPC goes, Google impacted these strategies too. Lines began to blur between the two, as Google integrated local results (map results) and expandable PPC ads (integrating PPC and Google Base results together) to include sitelinks and product prices / listings. From an SEO perspective, Google spent the latter part of 2009 emphasising the need for speed – advising that a sites load time may become a big factor for optimisation in 2010.

And Google didn’t stop there, continuing their foray into a vast range of things – some of which it has been working on for a few years, such as voice recognition search, along with working on more efficient translation tools and various apps for the Android operating system. And then there were the many mash-ups of their existing offerings, such as City Tours and Social Search.

It wasn’t all bright lights for Google though, as they continued to seem out of place when it came to other forms of media, including a much criticised attempt to push the Google Chrome browser on TV and print media ads. They also took what many people felt were a step back with a more traditional pricing on “paid placement” local listings, and YouTube ads, opting away from their cost per click rates – and instead going towards the old CPM rates in some cases.

At the same time though, marketers looked closer at the impact of TV on search. Going beyond slogans that say “search for (name of brand) on Google” – marketers looked at having the celebrity presence on their search listings too. Hilary Swank, Scarlett Johansson, Oprah Winfrey and Jessica Alba were celebrities used to endorse PPC ad copy in 2009.

The merger between Bing and Yahoo looms ever present for 2010, while Yahoo Search Submit Pro calls it a day. Yahoo site explorer is rumoured to be next in line, which was the source of many an online SEO tool that analysed links. When this will happen is unclear, but it does remind many of a search marketer of the times when the Yahoo / Overture keyword tool was silently put to sleep.

And of course, it wouldn’t be an end of year / new year posting without some predictions, so here goes. Among our predictions for search in 2010, are:

1. SEO will become the darling of marketing, as recession strapped companies look to get the most of their budgets

2. SEO’s will need to know how to communicate with designers and developers (and vice versa), as factors such as load time, script usage and site coding become key SEO factors

3. Data (advertiser and customer) debates will heat up over the year, as companies like Facebook, Google, Bing, etc tread the line between insight and privacy

4. Local and retail advertisers will see Google Maps and Google Base traffic volumes soar

5. Launched in 2009 – Scoopler.com will become one to watch in the “real-time search” race

6. Search becomes the marketing channel for geniuses (or genii), as from keyword research to semantic psychology – things are about to get really interesting!

7. International SEO and PPC strategies will evolve fast in 2010, as search engines such as Baidu spent late 2009 moving towards a more westernised search methodology in both paid and natural search listings

8. Real time search still needs a fair bit of work and will look vastly different this time next year

9. Mobile search will finally catch up to all the hype its received in the last few years

Drop us a line and let us know what you think is in store for search in 2010.

Blog post by Niall Madden, SEO Director of Reform

Bing & Wolfram Alpha? Research Based Search Engine Results Could Bring That Personal Touch

November 16th, 2009

You may (or may not) have heard last week about Microsoft’s deal with Wolfram Alpha – which has been in talks for the last few months and now allows Bing’s search engine to integrate Wolfram Alpha’s database results (in the US only for now). It should be noted from the start, that Wolfram Alpha is not a search engine itself, but more an information database.

The results will generally be integrated for certain niches, such as health (like nutritional information) and statistics (population, GDP, history, etc), but with a very unbiased and real time approach (with results you can interact with). And perhaps this is not huge news to some people, but it sort of hit me as big news.

It’s better than news of other recent Bing developments, such as how natural results on Bing offer content snippets and related keywords along with dividing the results across the keyword you entered, videos, local listings and several variations – a souped up version of Google’s “universal” project, or even the fact that the first page of results now includes several keyword variables, that results in 20 natural listings on the first page (which is quite a bit longer, but still shorter than the results page of Naver in Korea), because face it, its not new. Same goes for the Bing “XRank” thing they implemented, where users find out how popular the name of someone they entered is. Nothing new there.

So why is this still good news for Bing in my opinion? Because its about time! It’s about time Bing (previously known as MSN, Live, etc) started looking towards a different direction for providing search results. It had been wasting way too much time and probably too much money trying to be like Google in recent years.

FINALLY, Bing may be thinking “long term”, with this acquisition. Maybe it will use this and its Facebook share to give more of a “person to person” feel in results (even the real time Twitter results could help, if they figure out a way to get past the “junk”). Combining Wolfram Alpha with Wikipedia results (which Bing US already integrates under its own site) might help too.

Maybe even add a bit of new colleague Yahoo and their Yahoo Answers offering? Suddenly, instead of a commerial search engine going against commercial search engine, we’ve got one that is taking potentially taking the more “person to person” and “research based Wiki” approach. The one that looked many steps behind for a while now, is starting to look more modern? Strange, but its starting to hint at that.

Many innovative search engines have come and gone over the years, but the main reason many of these failed were that they simply lacked the budget to get seen, let alone have competitive results. Microsoft’s Bing at least has the budget part checked off.

But how about this Bing, forget about profit for a while. Let Google concentrate more and more on its increasingly aggressive marketing model (by this I mean, behind the scenes and how AdWords teams are concentrating on generating client spend. Not conversions, not traffic. Spend.) Take some more time behind the scenes and develop a more “personal” based search offering. Do it right and become a research based medium – and traffic might come back over (more than the single digit percentage share you have right now).

Then, when you’ve established a new type of user experience, turn up the profitability a little – in order to cover increasing resource and overall company management. This sort of approach worked like a charm for someone else, um, what were they called? Oh well, I guess this isn’t 100% new either.

Anyway, here are some reference points (saved as screenshots for non US users).

1) Wikipedia Content being served under Bing.com URL

2) Screenshot of a long page in Bing’s results for “auto insurance” featuring several keyword variables

Do I think Bing’s improved in the past couple of months? Yeah, but I also think it’s got a long way to go. Still, it’s finally looking in a different direction though, and thats a start.

UPDATE: You can integrate WA results on Google too, but only if you install this Firefox plugin – https://addons.mozilla.org/en-US/firefox/addon/12006

Blog post by Niall Madden, SEO Director of Reform

Google and Bing to Add Twitter Tweets in Search Results… Is That a Good Thing?

October 22nd, 2009

News came out yesterday that claimed both Google and Microsoft Bing are going to include a bit more tweets from Twitter in their search results.  The search marketing world was abuzz, with comments and ideas for how they can expand their search marketing capabilities and how Bing and Google will benefit greatly from these “real time” updates being integrated in their search results.

Really?  This is good news?  Perhaps I should tweet it to the contrary.  I say, hold on a second.  How relevant are most people’s tweets to the general public anyway?  Lets have a look.  Bing is already doing a beta of this in the US.  All I can say is that it’s good that it’s separate from the real results. 

Google’s blog at http://googleblog.blogspot.com/2009/10/rt-google-tweets-and-updates-and-search.html tells me that with Twitter integration, I can now find up to date snow conditions at my favourite ski resort.  So I tried to look for the “weather in Chamonix” and “weather in Whistler” on Twitter Bing (http://www.bing.com/twitter/).  Which by the way is the same as http://search.twitter.com for the most part.

One had no results, the other had ONE.  And it wasn’t even of any use.  No real time info, just some person tweeting that the weather was nice…. yesterday – so much for real time.  

Lets try something more commercial then… what’s more commercial than insurance!!  Here the results are split between blatant sales pitches (that came from automated twitter accounts and have links that redirect to affiliates) and people’s every day lives – such as one that reads “I just got car insurance, what up yall!!!”  True story, not mine.

Don’t get me wrong, Twitter has a lot of use for people that already know you or your company, your brand, etc, and want to know more or be kept posted – such as our Reform Digital account that lets followers know the second this blog is posted (quite ironic I know).  But if you’re a complete stranger that happens to query something related to this page’s content, I’d rather you find this posting in the search results and not the much briefer condensed tweet – which will of course put this lovely page an extra click away. 

While I know many brands have taken advantage of Twitter, to launch unique offers to users who “follow” them, in some ways like the email lists of old – putting this sort of thing on the regular search results sort of defeats that purpose.  Brands lose out from the quality of direct connections that are more likely to convert, and users lose out by seeing a wave of search results that get valued by the engines, but don’t really have much use in most instances. 

From the user end, will it open more Twitter accounts?  Maybe.  Critics such as Steve Rubel of Edelman Digital believe that Twitter already has the users who want to use it.  For the most part, he feels “everyone who wants to publicly tweet is already doing so”.  And I agree.  But, I do think there will be a wave of new accounts and more traffic for Twitter with this search engine integration. 

Brands, advertising, just like the profiles that went astray when mainstream media looked at MySpace and Facebook as revenue generating opportunities.  Its one thing to make sure your brand reserves their presence, and I’m all for proactive campaigns – but there’s a difference between a Tweet about a product offer from a person I know and a search result featuring a tweet from a stranger that tries to befriend me.  Can I really trust the latter? 

One positive that Bing’s Twitter integration claims to have is that it will only expose the tweets from the last 7 days.  This is a good start, as the last thing I want is an archive of tweets muddling up the index in Bing or Google for that matter.  On the other side, all the tweets I see in the results are pretty irrelevant.  Only time will tell I guess.   Twitter can be easy to spam also, which is why I liked the fact that it wasn’t fully integrated in search.  And it’s very easy to post an article and just get loads of accounts to link up to it.

Some articles cite that you can get real time weather, news opinion from regular people and sports information now that Twitter will get integrated on search.  Hmmm, I could get that stuff online already.  This just makes sure I get to see everyone who bangs on about it and thinks I care.  Lets see how Google fares. Good luck.

Blog post by Niall Madden, SEO Director of Reform

Update: Here’s what the US version looks like (for those that can’t see it).

Bing Twitter Homepage in US

Time for transparency of Search practices?

September 24th, 2009

At ad:tech in London this week there was the usual good representation from the Search agency community: DBD Media, Oban Multilingual, Jellyfish and Efficient Frontier, to name a few of the regular crowd; and Just Search, Optimize and High Position to name some newer kids on the block.

Each agency has its own USPs – whether this is an algorithm, a technology or planning tool. Some claim to be technology providers; others stick with the agency theme. Some have slicker marketing then others – glossy handouts and snazzy logos for tools; others have fast talking sales folk, who can counter any argument on the planet with “yes, but we’ve got a secret sauce”.

Poor, poor clients. Where do they start in evaluating what really sits behind the ‘smoke and mirrors’ sales rhetoric? How do clients know that they will get quality of service, efficiency of management, a best of breed piece of kit, and – perhaps most importantly – transparency of ROI?

Here’s the thing. What all of these Search agencies and technology providers are selling is a service. It’s a combination of people, process and technology. Technology is really important because it is the enabler, the heavy lifter, the lawn mower. But someone’s got to mow the grass.

Here’s the other thing. There is no perfect algorithm for delivering a service. Efficient Frontier call it an “intelligent learning algorithm”; we call it great minds, nimble fingers, Excel macros, spreadsheets and 50 people in India.

Oh, and clients, do ask about who’s doing the do. Is it in-house, or is it outsourced to a country where labour is cheaper? Many Search agencies nowadays are outsourcing. Because there is a lot of heavy lifting to do. If so, do they understand the nuances of your target language and how people search in that language. Are your business and marketing objectives being lost in the Chinese whispers translation from client to search agency to overseas outsourcing agency.

Search is already complex enough without suppliers masking the truth about delivery and adding extra layers for sales and marketing purposes. When will the mist clear so that clients know exactly what they’re buying? Isn’t it time that Search agencies started being transparent with clients about their practices, so that clients can make informed buying decisions?

Is it too much for clients to expect a straight answer to the questions “how?” and “who?”