I loved “Read This Before Installing Rails 3.1.3” – near perfect. It even includes a link for what to do if you have XCode 4.2. Here’s the extra bits you might stumble over:

Install Ruby 1.9.3

Instead of the installation of Ruby step in the “Read This” article, do this:

rvm pkg install iconv
rvm pkg install readline
rvm remove 1.9.3
rvm install 1.9.3 –with-iconv-dir=$rvm_path/usr –with-readline-dir=$rvm_path/usr

I found no other problems – there are, though, things that need to be done when upgrading an application from Rails 3 to Rails 3.1.

Updates

After nanofunk suggested that the “–with-conv-dir” and “–with-readline-dir” weren’t needed, I tested with “–gcc=clang”, as he suggested and I had compilation failures. I upgraded to rvm 1.10.0 and tried that – it also fails.

Using rvm 1.10.0 also results in the compilation commands that I give above, also failing. So, as of rvm 1.10.0, I don’t know how to build a working Ruby 1.9.3 on Mac OS X Lion with Xcode 4.1. I may have to upgrade the Xcode version?

WIth rvm 1.10.0, using “–with-gcc=clang” or without, and no additional command line options, I get an error during compilation of the intrinsic readline.c file. So I suspect that I need to do something about the additional library… Just can’t work out what, right now.

When I work it out, I’ll add the details.

In late 2008, I set up AdWordsHelpExperts.com (AWHE) with a group of other Top Contributors to the AdWords Help Forum. It is referenced by Google, in the Google sponsored and managed AdWords Help Forum, as a resource. I’ve now had a letter from Google requesting we abandon the domain name and threatening action against my agency if we don’t do so – and I suspect that it’s connected to a critical comment I made in a public forum.

Here’s what has happened, and I’d be interested to hear what you might think. Is this just strange coincidence or are some staff members at Google both arrogant and vindictive or is there another explanation that I’ve missed out?

History Of The AdWords Help Experts Website

I’d been involved with the AdWords Help Forum since about 2004, and made something like 2,000 or 3,000 posts – and I found that I had a real problem with the forum. It had no way to attach screenshots, and a lot of repetitious problems were much more easily solved with an article that embedded graphics… So I proposed that the Top Contributors at that point should join forces and make a website, so we could create really helpful articles, and direct users to them. Save time writing repeat responses, and show users exactly which link to click, etc.

We didn’t set it up to make a profit. We don’t run adverts on it. It’s cost my business about $1.500 in hosting fees over the years, and is deliberately set up to be very-low-level-promotional and very information rich. We (between us, all the top contributors that post there) do get a business enquiry about every month or so – there’s about a half dozen of us that have been involved; most of the enquiries are for troubleshooting small accounts. So it’s not a major source of leads for any of us. The site is an extension of the community-minded spirit that had us contributing in the AdWords Help Forum in the first place.

Links from Google's own AdWords Help Forum to the AdWords Help Experts Blog

Conflict With Google

Around 2009 I began to be really concerned by the way the user to user AdWords Help Forum was being used by Google. I was one of the volunteer contributors, unrecompensed, getting very few business leads (and not minding that, at all – it was never about getting new clients) – posting on the forum to help others and learn about the strange edge cases of Google, and coming up with interesting business problems to tackle. These volunteers were now being asked to deal with a flood of distressed messages to the forum, as Google stopped providing any service at all to small accounts. Around that point, if you had a problem of certain types (credit card payment failures, claimed unexpected suspensions, indefinitely long editorial reviews and slow low budgets), you were sent on a dreadful trail around more or less useless resources until you realised that at every turn, Google was sending you to a forum for other advertisers to offer advice. We got a lot of upset business owners and doubtless some fraudulent users, too, trying it on… but mostly we seemed to deal with some people who had real gripes about being unable to access any support staff that could help solve a problem with the main way they brought customers to their business.

I’ve made clear my opposition to this “policy”, in the AdWords Help Forum and on this blog. I even organised a “strike” by the volunteers to stop posting in the Help Forum until Google did something about the flood of postings from users that were impossible for another advertiser to usefully help. When you need AdWords account access to understand the problem, another user can’t offer anything more than fatuous platitudes in a user to user forum. Community minded active posters tend to come from a position where they want to help, not just say, “yeah, I feel your pain”.

Nothing, in all that time, caused a ripple of concern from Google staffers about our using AdWordsHelpExperts.com. It’s still referenced, today, in the AdWords Help Forum.

Today (2010-11-13)

Yesterday, a Google staffer accidentally publicly posted a rant on Google Plus. A really interesting and well thought through rant, with lots of interesting details. My attention was drawn to it by an excellent programmer I know, and it came up in various other forums, too.

I commented on a reposting of that rant, specifically about the commentary on Google’s arrogance. I think that shutting off access to support for small business AdWords accounts and sending those users to a public forum was an act of corporate arrogance. I said so in the comment.

And I pointed to another area where I think Google has failed users through failing to think through what users do – the appalling identity mess surrounding Google Plus, Google Apps and AdWords. I need three identities across two browsers to use Google Apps, AdWords and my original Analytics identity, because of the way that Google has messed up Apps users identities and the simplistic username and password authentication from 2004-2010 that gave me serious concerns about AdWords access (I use private email addresses to login to AdWords, not addresses used in public forums, to make account guessing harder, so I need at least two accounts, in normal cases, but used to be able to have both open in one browser).

I’d like to suggest that that’s arrogance, too – ignoring a paying user base, and making it more difficult for them, without a real personal explanation or apology? Isn’t that the action of an arrogant business? You might disagree – and I’d love to hear why you think paying users should have their services made more difficult to use, so that free users can have a new service. I may have a limited imagination or too narrow a vocabulary, so go ahead, and inform me.

There’s an old article about that Google Apps/Plus (One) Identity Mess, on this blog, too. It’s months old – July. Dead history on the web.

However… this one critical comment that I make is on a highly read page, distributed to a huge audience and attracting serious attention. Not like this small, rarely updated blog.

Within about 8 hours of that comment, I get an email from Google telling me to take down AdWordsHelpExperts.com:

Is this Google taking vindictive action against a small business, or just an accident of timing?

I’ll gladly admit that I don’t participate in the forum any more, or even the AWHE web site. I found 2009 to be a bruising, time consuming and deeply unpleasant experience for a volunteer. I get a sick feeling in my stomach just thinking about going there and what it was like for month after month.

Will I take the site down? Yes. I have better things to do in my life than argue about this. My already weakened respect for Google has crunched down another notch, though.

My reply to the Compliance Team.

Questions, questions

Is this another example of Google shooting itself in the foot with customer and top contributor relations?

Has Google now made it clear that external criticism is not permissible, or is this official and officious email just coincidence?

Wouldn’t an approach through the continuing Top Contributors to ask them to rename the domain, have been less confrontational?

Does Google really take vindictive action against external critics?

Am I excessively paranoid?

Suggestions (or donations) of a domain name, too, please!

Updates

2010-10-19

The compliance team have been very helpful, pointing out where they feel we’re too commercial on AWHE. So it looks as though we can keep the domain name – the other TCs are happy with reducing the promotional content, which was always pretty light and never the primary motivation for the site.

We have not yet discovered why the complaint was brought at the time it was, nor why the initial approach to a group of people who had been actively and voluntarily working with Google, was so aggressive.

So there’s still some questions – why does Google always appear to assume that people it interacts with are fraudulent, deceptive or malicious, and must be threatened for compliance or abruptly removed from advertising? Isn’t there a way of approaching clients and partners that is better customer service – assuming for example that it may be oversight or ignorance rather than something requiring threats of action? I’m pretty sure that none of the customer service experts I know, would recommend starting every new dialogue with threats.

But what was the number one listing on Tuesday and Wednesday has vanished by Friday:

What’s really remarkable is how thoroughly this site has vanished:

It’s usual for a site to have the home page as the first page in the sitelinks list. For the home page to vanish… well, looks like Google hates this site. And I wouldn’t bet on any rank being passed.

Unsurprisingly, none of the Google penalty checking websites that I could find suggest that there’s a problem with this site. After all, it has results in the listings. Just very low ranked and none of the top pages (like the home page) show up. Sure sign of a problem of some sort!

So, what happened? Well, the site was probably reported to Google, possibly even before I saw the solicitation. And Google acted, quite quickly.

If you have a Google Webmaster Tools account, you can use Google’s Undeclared Paid Backlinks report form to report sites that are using and offering undeclared paid backlinks. I don’t know what you do if you aren’t a registered webmaster – perhaps find one and ask them?

So, what happened to the companies mentioned in the story? I can’t find them, now, either. Certainly not on page one of results (I’ve used the non-personal search on our tool to view Google organic and paid search as if in another country)

Google really isn’t happy with undeclared paid links. At all.

I've still got a problem with how this was done - no email notification that it was fixed, and I didn't see anything in the Apps management console either; you find out by following some Google blog or other, or keep trying it. Google ought to improve communications with paying customers.

And as a result of Google's approach of non-communication and the forced creation of transition accounts, I now have a pointless Google Account which I must merge back into operations - more work for me and my staff, and no apparent help or guidance from Google for my having to fix a problem they created. It's just not good customer service. I suppose I'd like a way to merge Google Accounts - so I can get back the access that I had seven months ago.]

This has to be the strangest thing that I’ve come across on Google Plus and Google Plus 1. Google has created a paying service to manage user identity, and then excludes those long period paying customers from taking part in Google Plus One and Google Plus. Yup, if you have Google Apps, you can’t use social networking, and you can’t mark the resources you find useful. Everyone else who *doesn’t* pay Google for services, can use Google Plus and Plus One. And Google’s response to Google Apps users? Silence. Total silence. Thanks, Google guys. Thanks a bunch.

[Update: 2011-08-12 - Google Apps Help Forum has a Google Staffer response, but unfortunately he promises to keep users updated on a thread that already has 8 pages of comments. Find the updates if you have time... What's wrong with a pinned posting, maintained by the Apps Help Forum Advisors and locked against user content - a reasonable way to publish the Google position in a forum, with such a strong set of questions?]

[Update: 2011-07-21 - How Can I Get My Staff Connected? - You'll have to send an invite to a non-Google Apps Google Account, from a non-Google Apps Google Plus enabled account, and run everything involving Google Plus in a separate account. If you send an invite to a Google Apps accunt, it is completely useless, because Google appears to check that the Google Account that you are using, matches the Google Account of the invitee. Invites to a Google Apps user are completely useless. Not that Google tells you that, either as sender or receiver, until you actually click on the link and get the error message about Profiles, below.

Note that Google is apparently both intending to allow brands to have space in Google Plus, and is also supposedly adding access Google Apps users, at some unannounced point - so you may face another problem of merging identities or re-establishing an identity when Google does get around to allowing you access. Nothing like making it easy, eh?]

What the farquahr was passing through the mind of the Google Product Managers that chose to do this? Perhaps:

“I know, we’ve had paying users on this service since about 2005, we’re sure of their identity, because they are paying us to make sure we know who they are, so, I dunno, let’s just forget them because, well, they’re obviously idiots. They’ll take any amount of abuse. They’ve been used to working out how our poorly documented systems can be used, so being unable to reach another service we offer can’t possibly bother them. And if they complain? Meh. Paying customers on AdWords go for years without answers to basic problems, except ‘ask another user’. Fuggedaboudid.”

[Update: 2011-07-20 - Dave Girouard, a VP at Google says (but not in the Google Apps Help forums run by Google), that adding Profiles is a priority; even using the guys' name in searches, I can't find this post in the last month of articles on Google's various Blogspot.com blogs - probably because Google are such lame users of SEO. Running a separate blog article with a probably non-useful title and perhaps missing important cue words to allow search to operate properly, rather then replying in the Help Forum to questions from Google Apps Administrators and users, on a forum set up by the service organisation, is pretty peculiar. It's like me getting a question from a client, writing a blog article and neglecting to let them know it is relevant - irritating for everyone, and less than helpful. Blogs are for public announcements, not client communications. Learn. The. Medium. And learn how to write so that useful stuff can be found on search, or fix your search engine to work with the opaque way that you blog. One of them is ineffective - either search doesn't connect relevant material, or the writing needs fixing.]

Do Google Customer Service Staff Keep A Chart Of “Clients Crapped On This Month” and Compete To See Who Wins?

Doesn’t excluding Google Apps users from two services means that a whole bunch of federal and state government employees, several significantly sized corporations and a whole bunch of iddy-biddy liddle companies have just been excluded? Oh, and didn’t I see some national governments in Google Apps “list of informationally endangered organisations” client list?

I used to think that the hundred thousand or so small businesses denied any access to AdWords Customer Service was an abomination. But whoever thought up the idea of silently and without warning denying access to the great new experiment in social communication, to all Google Apps customers, has to win this years’ award for “Most Google Paying Customers Excluded From Reasonably Expected Service Levels For No Stated Reason”. We’re talking ten thousand users *at a time* for some Google Apps clients. By my estimate, it totally dwarfs the scale of ignoring small business AdWords advertisers by about an order of magnitude. And to omit the same user group on *two* services, in one month – absolute genius. Not sure if they get double points for that, or an exponential powerup.

Why This Might Not Be a Good Idea For Google

For most businesses on the planet, paying customers come first, not last. If they did so at Google, then when Google have a new and exciting service, Google should make sure that the people who pay the Google payroll and keep the lights on in the Google datacenters, get an early crack. Rather than the current policy, which is apparently to keep paying customers both excluded and totally in the dark. AFAICS, that’s really not very clever customer service – is it? Am I really that out of step with how organisations should be treating clients?

You know that number one thing in the Google corporate mission statement? You know, the one about “Focus On The User“? It comes ahead of the one that people usually talk about, that you can do business without being evil? Point of fact for the Google staffers… the people who pay Google to do things for them, are Users, too. Just because they pay Google, doesn’t mean that they should deserve less respect.

I’m really annoyed by Google. For now, we use Google Apps. Or at least, we will probably do so for about 30 more days *WHILE I WORK OUT HOW TO LEAVE GOOGLE APPS*. I’d prefer to not-have access to a service that I’m not-paying for, than be denied a free service because I am paying an organisation for their services.

Cloudy Implications

Why would being disallowed to use Google Plus and Google Plus One make me reconsider our usage of Google Apps? There’s some really cute things you can do with apps, from scraping sites to sharing stats. But… Google is supposed to be unifying the Identities of Google Accounts and Google Apps. This is the first crucial test of whether new services will be available. And the answer is “FAIL”.

({sarcasm on} Good product naming system, BTW – makes it totally clear in the users’ minds what they are doing and how separate those services are. I can’t see anybody ever being confused about them. {/sarcasm})

Which would I rather have? And what would my users in the business rather have? The ability to mark useful resources and engage in controlled social networking, or the opportunity to use the not-as-good-as-Word-or-Pages word processor, or the not-as-good-as-Powerpoint-or-Keynote presentation tool, or… Well, they already hate using Google Docs, unless I force the use on genuinely shared data. So most would vote to kill it – they use the mail system, and otherwise overwhelmingly prefer to use a local app with a richer UI and features. Unless it is genuinely data for interactive sharing.

Sharing Data and Encryption

There are other ways to share data, after all. The failure of DropBox a few weeks ago (they accidentally allowed anyone to access any content in any DropBox for a four hour period) has made me wonder about the wisdom of having unencrypted data in the Cloud. I’m beginning to develop the idea that, just as I do with DropBox, I only put already encrypted data on it, or I use for insensitive data, stuff that I wouldn’t be unhappy to have leaked. But Google keeps everything in the clear… And that’s increasingly uncomfortable for me. We have client data. If we continue to march towards sharing data, I want an secure communication to an encrypted resource, not something merely protected by a single level of authentication and a secure communication protocol. I want the service to be unaware of the keys to unlock the data, so even if someone at the service provider forgets to lock the resource, I’ve still got a good level of protection on the data.

Google Apps doesn’t have that level of security now. I haven’t seen it discussed as a future option. And if Google has such disdain for Google Apps users that it won’t communicate about its’ most important new communications and search mechanisms… Well, I don’t think they care about my concerns for improved data security. So I think my business needs to move as soon as practicable, from a provider that isn’t even talking about services that I think are increasingly needed. That’s how connected Google Plus and Google Apps are. The behaviour of one is a likely predictor of behaviour for the other – and I just lost all confidence that Google understands what a customer is, and what their needs are.

“Google customer service” is apparently an oxymoron – or at least, looks that way from the lack of any statements that I can find, using Google’s own search engine to search their own web site site and blogs using the keywords “google plus google apps”, as of July 11th, 2011. As George Bush memorably said “Fool me once, shame on you. A fooled man can’t get fooled again. Erm.” I’m just fed up with Google. Really feel betrayed. Again. And Again. And Again. And that makes me feel like an idiot for trusting Google, again. I don’t enjoy feeling like an idiot – so I’ll be extending less and less trust to Google. But I will use them for their free services – those are absolutely fantastic value.

Summary

Understanding how Google treats paying customers yields some illumination on free services. My interpretation of silence is negative – I can’t construe Google’s silence in their own forums in any positive way. I fear for the security of my company’s data in Google’s Cloud – because it is held in clear and offers no opportunity to hold encrypted shared data.

Google should be reconsidering what it does with paying clients. They shouldn’t be the last to use a service, but amongst the first; and they should be communicated with. That’d create an incentive to pay to use the services. Not a disincentive.

I now detest Google about half the amount that I detest Microsoft. I’d still prefer to use an Android phone over yet another Windows implementation, Google Search Results barely over Bing, AdWords over adCenter. And I’ll probably be looking to see if I can find an autoencrypted file sharing service that uses AWS/S3 or something similar as a service, so I can remotely mount an “encrypted disk” and use it whether connected or not (rather like Apple’s iDisk but better shared, and encrypted).

Notes

Fuggedaboudid = “Forget About It”, said quickly and with nasality.

Advertising networks want to maximise their revenue. The way they do that is to pick the highest CTR, highest paying adverts. This turns out to be quite simple. Forget about pay per click, and instead calculate adverts as if they were CPM. That’s easy to do. If you have an advert that costs £1.00 per click, and has a 10% CTR, then it is showing at a CPM of £100.00 – because a 10% CTR means that you have 10 clicks per hundred impressions or 100 clicks per thousand impressions, and 100 clicks would cost you £100.00. Now rank your advertisers based on that calculation of the value of an advertisement and you have the first simple AdWords auction – it got more complex, but that’s still the basic premise of how it works, but with a “Quality Score” factor instead of just CTR.

However, while that maximises revenue for the advertising system, it can’t guarantee that it has maximised the revenue of the advertiser. The practical example was when we were working on Thomas Cook. If we offered adverts with “Book now…” then we got an increase of 40% more bookings and they mostly happened a week earlier, than if we used the copy “Browse…” or “Search…” instead of “Book now…”. But the CTR was slightly lower on the “Book now…” adverts. Which did Google choose to optimise? And what was the impact?

Yup, Google picked the higher CTR adverts, which de-optimised revenue for our client. So we stopped allowing Google to choose which adverts to optimise. It’s the only safe defence.

So, very early on, we made a policy decision that we don’t allow the advertising networks to choose which adverts they want to show, unless the client purpose didn’t involve a strict ROI target (e.g. a brand awareness advertising exercise). And we now inspect adverts to see which ones have a lower ROI, and lose them – not based on the CTR, but based on sales.

adCenter doesn’t have the option to show adverts equally, on rotation. That means that adCenter will optimise its’ own revenue, at the expense of an advertiser, by removing the advertisers ability to collect a statistically useful sample of results, unless the advertiser goes to daft lengths (e.g. sequential tests rather than A/B or multiple parallel).

If you are interested in optimising for the new adCenter, see if you can register your interest in having Microsoft permit you to have equal rotation, so you can assess which adverts work and which fail. Note that there are other requirements if you do this – if you have multiple keywords per AdGroup, for example, the performance of each keyword needs to be roughly the same, or you may be selecting adverts to display for high volume low conversion rate keywords, when the right driver to select the adverts is a low impression volume high conversion rate keyword. Making web advertising effective is a bit more complex than just equal rotation – the optimising method has to include how many times each advert is showing for each keyword.

This lack of control by the advertiser has been noticed by other adCenter users. I was going to add this comment (or one rather like it) to the adCenter discussion group, until I found out that to do so, I’d have to stop using this computer, and use a Windows machine. I’d rather not, thanks. I have picked this operating system for a lot of reasons – and Microsoft’s tedious obstacles don’t make me more likely to use Windows, it makes me less likely to do so.

I should point out that this observation about optimisation is not unique. In fact there’s quite a lot of other people making this or a very similar point. Here’s some:

• Use Split Testing to Create Killer Ad Copy
• Ad Copy Testing: Systematically Improving CTR – Microsoft’s own blog for advertisers recommends split copy testing
• Is it possible to make ads rotate evenly? – Microsoft adCenter discussion forum with some strong points made by advertisers

This is an issue that Microsoft needs to take seriously, It can have a substantial impact on performance, and the value of the tried and true technique of split testing is significantly reduced when Microsoft chooses which advert to show. It may help Microsoft’s revenue, but not that of advertisers. And offering “just create a new AdGroup because they start without optimisation” – is just not do-able. We tried. Merging the data from multiple attempts so that you capture the period before Microsoft has optimised is a data horrorfest, especially on a large account. And testing like this is yet another management overhead on a smaller fraction of advertising than Google – Microsoft need to reduce the administration overhead, not increase it, relative to AdWords’s cost of operations.

When the networks merge, the combined total should probably be about 20-25% of the AdWords total. Now that’s of a size to start getting properly interested in optimisation, and one can justify some management time to organising and sorting it out.

Because we’re anticipating increasing client spend on the enlarged network, we’re looking at what techniques we can use to optimise. That’s what we do… optimise stuff, after digging into it. So I wanted to comment on something that I thought would be useful to do, in the larger network. I went to the Microsoft adCenter user discussion forum to contribute. And I got this, when I tried to submit a comment:

Microsoft adCenter discussion forum requires MSIE, through non-display of CAPTCHA on non-MSIE browsers.

Microsoft is making use of their advertising platform pretty difficult. Firstly, I don’t like this whole “you must use MSIE to get the experience”. That’s so 1997. These days, there are browsers that offer very usable interfaces. Oh look, I’m using three of them now – Chrome, FireFox and Safari. It’s not as if Microsoft even said “MSIE9 is so far ahead that we’re making that the standard” (because that’d be a pretty stupid thing to say, anyway), because MSIE8 *can* be used – and that’s certainly no more advanced than any of the other major browsers.

Nope, Microsoft have made de-facto that to effectively use their web interface for their advertising system, you have to run their operating system and their browser. It’s so… retro non-chic. It’s a marketing approach out of some 1970′s vendor lockin agenda.

Anyway, the point of frustration was passed today, when I realised that not only do I have to use a Windows system to interact with the advertising platform, I also have to use a Windows system to interact with the community forum. It didn’t use to be that way. A year or so ago, I could post message. Now, though, I have to complete a CAPTCHA – which, regrettably, is not visible on Chrome, Safari or Firefox, and has no recommendation as to the plugin I need to adopt. Nope, if I use a non-Windows system with a non-MS browser, they just don’t care to hear from me.

With no CAPTCHA form visible, how do I submit the form with a CAPTCHA? That’s just a design and test fail.

And there’s another thing…

If I’ve just typed a comment to your blogging platform, couldn’t you had the courtesy to let me know that I needed to be running MSIE *before* I start writing? I mean you do *know* that your “community” now requires MSIE to be used before a comment can be added? In my case, that means usurping a Windows machine used for compatibility testing and adCenter operations, so I can contribute to a discussion. I don’t feel that generous to Microsoft. I have better things to do with my time than switch machines, in order to post a comment in a discussion thread.

But that’s not even the serious thing…

If I’ve just spent five minutes typing in a reply on the forum, Microsoft knows quite a bit about me. Potential details include stuff like the pacing of characters, the previous activity of the Microsoft authentication service by that ID, the IP address, possibly backtracking the IP address, and details of the packets to indicate the type of operating system I’m using, all those browser header details, and various other clues that I’m probably a human. So if you can’t tell from my logging in and all that typing, that I’m human – how the hell are you going to tell from a single click that my advert clicking is or is not fraudulent?

How about it Microsoft? Can you actually detect bad clicks at all, or do you just guess that some percentage are bad? The amount of spend we’ve done on behalf of clients, and the ROI, has never meant that it is worth checking the adCenter fraud detection system. I think I’m getting interested now.

And why can’t Microsoft make its discussion and feedback systems accessible to other browsers, like any other competent technical organisation? What makes Microsoft so technologically incapable that a *discussion* forum requires a technology limitation on the platform used for access? This is just so primitive. I am *so* glad that I decided my organisation would not be using Windows as a standard operating platform. This level of lockin is just tedious obstructivism. At least have the courtesy to tell me when I log in that “some features of this discussion, such as replying to comments, are not possible unless you use Microsoft Internet Explorer 8 or 9″? Then I won’t waste my time trying to engage.

Microsoft has not convinced me of the Windows Advantage. It has convinced me that they are technologically primitive, and that using Windows systems would be a technological backwater. That’s the complete reverse of what they probably thought they were doing when they made this design decision. The signal I get is “if you use Microsoft products, you’re locked in, unable to choose what is best for your business, and we don’t care at all”. That’s not a good message to people who aren’t using Windows now. It is not going to convince us that using WIndows is a good idea. Forcing users to do something is weaker for branding, than exciting them.

Now, that could cause a problem managing these accounts. I could use Google’s support for POP and IMAP and read the Gmail accounts for each of these into a central administration account. That’d mean quite a lot more work to set up each account – tripleing or even quadrupling the administration involved. But I’m a geek, and old UNIXy kind of geek. I know some stuff about nearly anything to do with operating systems… and mail systems were something I did some work on, over the years…

Within the mail system, that is inside the Gmail that is part of Google Apps, a plus sign after a name is a qualifier, not a different address. So if you send me email as “jeremyc+(make up something)@merjis.com” it still reaches me at jeremyc@merjis.com. IIRC, this usage is an old, often omitted or badly handled part of RFC822, the ancient mail system specification, from around the 1970′s. However, the Google Accounts (and consequently Google Plus) part of Google hasn’t worked out that the Gmail part of Google has made this work, and treats Google Account identities with a plus sign in the middle, as completely different entities, though, AFAICS, none of them can ever receive an email, at least when part of a Google Apps domain.

Google Apps still hasn’t merged its’ idea of identity with much of Google, so although my main identity is managed by Google, I have still had to re-use or create identities to play with Google’s stuff, outside the Google Apps domain. I don’t normally need to know which identity is which by given the different Google Identities differing user names, because the email addresses are checked by filters in Gmail and automatically allocated to client tags – so I can see at a glance which AdWords Account (or whatever) is receiving the email.

That was convenient for creating a lot of AdWords accounts, and have me automatically receive the emails, rather than having to set up IMAP support in a lot of Gmail accounts. But it was, to a substantial extent, relying on Google to be confused about identity. If they weren’t confused by email addresses, I’d have had to create other email addresses and do the IMAP thing to drag all the email to a central place for monitoring. More work for me, but I’m vigorously lazy – I’ll put a lot of work into understanding something, so I can achieve what I want, for less effort each time I do it.

I finally worked out what happened when I was invited to join Google Plus. The email went to jeremyc@merjis.com – a Google Apps account. On clicking to join, I got a new tab in the browser, and I was immediately logged in, with my name (not my email address) showing. However, I had been switched to the account jeremyc+money@google.com[*], which is the main Google Account that I use to interact with AdWords.

So, jeremyc+money@merjis.com is a Google Account, not a Google Apps Account. When I got the Google Plus invite, it automatically used the identity that was logged in as a Google Account. However, people are now sharing via Google Plus with jeremyc@merjis.com, and that’s not a Google Account with a Google Plus identity. So even though I receive all the Google Plus messages addressed to jeremyc@merjis.com in the account under Google’s management as jeremyc@merjis.com, I can’t read it, because Google Plus knows that the registered entity jeremyc+money@merjis.com is not jeremyc@merjis.com – and jeremyc@merjis.com can’t join Google Plus, even though jeremyc+money@merjis.com is a member of Google Plus, and jeremyc+money@merjis.com shares all its’ email completely and fully with jeremyc@merjis.com.

So… If you message me using Google Plus, I can’t respond, because you haven’t shared with me, although there is every appearance that you have, and I receive a message about you having shared with me. And that’s because although I’m me, there are a lot of different me’s maintaining complete identity separation including different passwords for access, except for total sharing of email. Confusing, eh?

Footnote – it isn’t really “jeremyc+money@merjis.com”. The *whole* reason for using another identity is because we’ve managed individual client spends of US$500k/month – so I don’t want hackers to know which Google Account is the MCC owner. If they hack the account, it causes a lot of misery and pain. So it is deliberately set up to be private, hard to guess, and with an ugly password. I’ve done that, because I don’t believe that Google does enough to protect AdWords accounts with large budgets (read “large budget” as as “larger than the average Joe could cover with some kind of credit card insurance”). My bank, on a much lower budget, makes me use an authentication device as much as twice per transaction on sub-$100 transactions, whereas Google seems to operate a very simple username plus password system for millon dollar budget accounts, which makes it relatively easier to hack than a bank…

Anyway, returning to the main web site, at last, you can do a *limited* Google Search As If You Were In Another Country Or Language… I say “limited”, because at present we’ve only got the UK, US, Germany, France and Japan covered, in English, French, German and Japanese. If you have a language or country you need, drop us a line or add a comment on the blog and we’ll add it.

So, how do we know this tool works?

When we first developed the tool, we phoned search engine marketeers that we knew in the USA, South Africa, Japan, Australia and other parts of the world, to check what they saw when they did a search, and we compared it with what our tool was showing. They were the same, subject to the usual Google variations of ranking that different data centres deliver.

If you want to get some independent evidence, and you’re not in the UK, give us a call, and we’ll tell you what we see – so you can check whether it’s really useful :)

So, what’s top, right now, for the search for “SEO”, in each of the UK, US, Germany, France and Japan? Here’s the tool results:

As you should be able to see, we’re pulling up different adverts, and different organic searches. If you just change the country code – it’s not enough.

So… How can this tool be improved for you?

What is the Plus One Button for websites?

It’s a small piece of Javascript to insert on web pages that are candidates for Google Search Results. On each page, you need to refer to the JavaScript file, and then invoke the script.

<script type=”text/javascript” src=”http://apis.google.com/js/plusone.js”></script>

<g:plusone></g:plusone>

As you can see from the brevity of this code, it has no idea *who* has dropped the link, other than someone with authority to modify the site. This is in strong contrast to Facebook, where the ID is an integral part of adding “Like” buttons. The implication is that Google doesn’t care who drops the code – they care who clicks on the button it displays, and probably, they care more about which page was clicked.

Click on the icon that the code displays, and your ID is added to the list of users that are interested in this page – and the icon changes from white to blue, to indicate that you have clicked it. If you “Plus One’d” a page, then it gets marked in the search results, as shown above. Note that the tooltip is not changed – when you +1 a result, the tooltip still shows that you should click to recommend the page, but that will actually remove the recommendation. Probably a transient oversight at Google, rather than intentional.

Here’s this very article, towards the bottom of Google search results for “Google plus one for websites”, before anybody I know had “plus one’d” the page:

You can see that I’ve +1′d the page. It shows up under the listing, below the URI, and it knows that I’ve voted – but gives no clue how many *others* have clicked. At the point that this was taken, there were a handful of +1s for the article that were at least an hour old.

It appears to take at least fifteen minutes for the “plus one” information to be updated in search results. I have to admit that I became bored checking – so somewhere after 30 minutes and before four hours, the search results had added a friend, who I’d ask to make a timestamp-noted entry. This is *much* slower than the update to the count of “plus ones” shown on the page – that count changed in seconds. I’ll guess that the delay before turning up in search results is to add the social graph data of who I know that has clicked, so my search results pages show data about people I know. I suspect that this process takes some time… so if you’ve just put up a post and emailed your mates to +1 it, don’t be surprised if you see on-page activity a long time (internet time) before the Search Results update to reflect that.

Amusingly, while I was checking to see who else that I know had used a plus one on an article about a plus one (interesting behavioural change – I went looking *much* further into results – will others look deeper to see if their friends have marked stuff, and will that be a transient behaviour until the SERPs behaviour becomes clearer?)… I found that the preview had updated to include the latest summary within the article and the clip showing the position. Here, recursively, is this article showing a clip of Google showing a preview of this article showing a clip of Google’s search results showing this article…

The implication is that Google snatches previews at the time of interest and marks them up, as a much faster process than the social data is updated. Now… Why is the social data apparently slowed so much, relative to the preview? I don’t know, and I’ll be thinking about that for a while.

What Failed Before, and Will The Plus One Also Fail?

Failures aren’t the worst thing that can happen. Failures tell you when you aren’t doing the right thing. So long as you don’t do something that kills the business, you can learn from failures. A failure to learn from the lessons you should be learning from a failure, now that is a real failing.

What has Google tried before, and why didn’t they work? And are there lessons from those attempts about what would make Google’s “Plus One” work, or be yet another fail marker?

Google’s SideWiki

Perhaps the most comprehensive attempt to solicit users’ direct feedback on a page or a business, is the Google SideWiki. This is barely known about, even in search engine marketing circles, much less by the wider public. It lets any user add an annotation to a web page. If the web master knows about the SideWiki, they can respond… But I can’t recall ever talking to a webmaster who knew about SideWiki before I mentioned it, much less ordinary web users.

The problems for SideWiki are quite large. There’s no on-page presence (if you have the toolbar or Chrome, there’s a minor marker in the decoration frame for a web page, but nothing in the area that users focus on). The SideWiki is intended to be present even for sites that don’t want to participate – which is why it has such a peripheral vision presence. It’s a parallel web space so only the search geeks or the deeply aggrieved find their way on to it. Parallel page specific annotations are a rich source of information for Google, with a user indication of liking or not liking the content, and it can provide a parallel stream of keywords and some clues as to the importance of a site or page (e.g. the Apple home page has a lot more SideWiki comments than the Twitter home page – working out why that is meaningful is left as an exercise for the reader! Try using the SideWiki for this page to discuss this? Your attempt to use it may partially explain why it has been so little used and known.)

Considering Google’s first rule in their Ten Things (“Focus on the User and all else will follow”), I’d have to call SideWiki a brave, but doomed, effort. Collecting additional data on page value is a great idea. But doing it in a way that users aren’t aware of, and can’t possibly comprehend, without a marketing information campaign to support it, and proper tools to engage with it for site owners and regular users – well, it’s destined to be roadkill. SideWiki is, I suspect, a victim of Google’s Engineering Lead business philosophy – a fallacy supported by the success of their first effort in this space, Google’s technologically and philosophically profound switch in ranking the results in search engines.

The “Engineering Lead Marketing Fallacy” is that if you make a great product, people will find and use it – but the reality is that if you have something good, you need to educate, inform and persuade users that it is good; this may be a rich vein for another article on another day, though; just think about Sun, until I write that article. I’m deeply familiar with this fallacy, having been a techie myself, before my Damascene conversion.

Google Stars

In Google Search Results, you can “star” a listing. This means that it is a page that you like. Starred listings are preferentially popped up to the top of your results pages. Like SideWiki, that is great feedback for Google on pages that work well for users. Pages that are well liked should get lots of stars – an easy way for Google to automatically find the hot stuff and the useful stuff. They are directly related to Google’s Bookmarks service – and social bookmarking is one of the unsung other keys of Social Media (it used to be sung, back when delicious was new).

Why doesn’t “Google Stars” work? When I talk to users they haven’t noticed the little faint outline of a star, or couldn’t be bothered to work out what it might mean. This star is an unusual icon, faint, and unexplained. Only the most curious users work out what it is for. And Google does their best to make sure that users *won’t* engage directly with the Google Star. That may be a surprising statement, but you have to consider the user experience (UX) when considering why the Star doesn’t and can’t work very well to deliver user feedback.

What’s Google’s proud claim and ambition (number 3 in the ten things)? That they deliver the best page of search results so you spend as little time as possible on Google. This does not encourage users to spend time investigating the details of search results. Any system that Google develops that requires time to parse the page, infer meaning and interact with the results, is counter to that (so far) successful formula of reducing the time spent looking at Google’s own search results rather than looking at the content that you wanted to find. Because of this emphasis on shifting people off the search results page as quickly as possible, Google Stars will be a minor “also ran” in terms of soliciting feedback, but probably slightly more effective than the almost invisible SideWiki.

As a subset of this mechanism, is Google’s “spam marking” annotation. Same problem. I find out that the page is spammy when I’ve been to it. When I use search (rather than being an SEO and staring at search results for ages to inder why Google does what it does), my attention is not on cleaning up after Google’s spam cleaning efforts, but on getting to the page that helps me do whatever it is I’m doing. So the user focus is not, at the time of looking at the page, focused on the question “is this spammy?”; anything as onerous as a check box, is too onerous to work with and is inappropriate within the normal context of a user interaction with search.

Effectively using the Star or Spam Marking within Google Search Results means having to return to Google, perhaps rekeying the search, and the user doing so while intending to mark a result as being annoying. Too much effort for most people. You’d have to *really* hate a web page to take that much effort. Or remember that if you are using Chrome, you can use the faint star at the right hand edge of the input box (URL/search field) to Bookmark/+1/Star the page you are on.

Google’s Search Wiki

Wikis are great, for certain classes of problem at least. The previous version of our company main site was written using a search engine optimised wiki we developed in 1994 and ran until a catastrophic incident in late 2010, but that’s a story for another time…. The point was that a Wiki collects lots of stuff, organise it and let users inside and outside the company work on it.

Google used to use a Wiki in search results. Google’s SearchWiki was effectively another parallel web – another Google system that had potentially one web page for every web page out there. This was before the success of Facebook, when MySpace was still a power online. SearchWiki looks like “Social”, smells a bit like “Social” – and it failed. There’s no sign of the user generated content attached to search listings any more. I think this experiment fell foul of the same problem that afflicts Google Stars – but even more so. The problem being that users on a search page don’t want to interact with the search page, they want to go to a useful web page. The Wiki compounded that problem by making the interaction into an essay. Users like simple UI choices – “do I like this page”, “do I like this company”, “do I never want to see this again”? Only a tiny fraction of users like writing essays. More prefer 140 character limits (c.f. Twitter) and more yet prefer just a “Like” button.

Google Plus One

There have been a few other attempts by Google to get users to interact about sites and web pages. They’ve all, to my knowledge, been variations of the above themes – hidden content in a parallel web or user engagement in search results pages. Note that user engagement does work when well done – Google’s own search results are a partially a consequence of user engagement with the content (title, snippet, URL and the actual site all play a part, as does the user interaction with all of those). The failure is not in the effort to engage users, but in the absence of the first rule (“Focus On The User”) in the interaction models – I believe these exercises have been focused on improving search results for all users, at the expense of focus on the individual user as contributor. Clicking a “Like” button, or a “Plus One” button, is a contribution – it’s just non-textual.

Google’s Plus One Button is different from all these previous efforts though. It’s on the web site, not the search results. It’s where users are focused when they are doing something. I think it is the first thing that Google has done, probably since they invented the whole idea of Page Rank, likely to actively engage with user opinion. This is a markedly different activity from any effort so far. But will it work?

So long as the interaction model is kept to a minimum (“I like this and would want to see it again” as a checkbox) and on the page that the user is considering, and if it turns up in search results (“you liked this page when you saw it”), it’s reasonably good.

The idea that I’ll know that people I know have marked a page – that’s good, too, and that puts it closer to the Social Media space. Knowing that someone I respect likes a page, helps me to decide that I want to visit that page too. This may have a profound effect on SEO – and given that user preferences are having increasing importance in ranking, who’d care to bet that this won’t also become yet another Google AdWords Extension – allowing you to advertise more visibly when you have a preferred page in the listings?

What will stop “Plus One” from working?

Several things can stop “Plus One” from working:

• A good distraction from FaceBook and LinkedIn or GroupOn. Distract people from paying attention to “Plus One” when it is new, so that webmasters focus on some other issue to deal with.
• Google’s Engineering Fallacy – just because this is probably the best interaction model so far devised for improving search engine results, doesn’t mean that it will win. It needs marketing and communication efforts. It needs Matt Cutts to write about it and encourage webmasters to embed it. It needs the Webmaster Tools guys to make space for it and report the results in WMT. It needs to actively let webmasters know when their sites are loved (and relatively unloved – like Google Analytics Benchmarking data). And most of all it needs to be social – I *care* when Bill Slawski, or RustyBrick (Barry Schwartz), or Kim Clinks like a web page. I don’t care when it’s some random Joe (unless there’s several thousand real random Joe’s, because that has its own message about a page).
• It needs to be non-spammy. Google has mechanisms to understand whether real users are interacting; the votes should probably use the same kinds of invalid click detection systems used in AdWords and Search Results. However, reporting to those to webmasters may need the same kind of obfuscation that conceals exactly which clicks were invalid, in AdWords, to continue the security-through-obscurity approach used in click fraud detection mechanisms.

At the moment, I don’t see how how I respond to a friend Plus-one-ing a page. So it is limited in terms of social media. I can like, or not. I can’t interact. Unlike Facebook, where I can comment on something that someone has found. That may yet prove to be the greatest weakness. Being able to see my friends liking something – that’s good. Being unable to respond and correspond, a core weakness.

And the name… I can easily grasp “liking” a page. But “Plus One-ing” it? Smells geeky, even nerdy. Hard to say. Hard to write. Hard to put into a grammatically correct page unless treated as an adjectival-noun thingie.

Is there any sign that Google Gets Marketing Communications?

A bit. But the most prominent usage of the “Plus 1″ graphic on the announcement page is just a link to the giant graphic, not a way to plus one the plus one announcement… That’s hidden with the other social markers below the article – and not even first in the list. Placement itself means something. Come on, Google guys – you rank stuff by importance. If you put “plus 1″ towards the end of the list of social markers, how important is it? Learn, for goodness’ sake, from your experience.

I’ll probably update this section as I see signs of life-after-engineering. When I see some type of audience communication rather than simply an announcement.

Summary

Good effort, but not enough, yet, to be completely convincing.

Possibly worth getting your braver clients to sign up and see what happens.

More important for Google’s de-spamming of search results than for social interaction.

May fail, because it is only partially social – no response mechanism to friends.

Where’s the social list drawn from? Ah. Now that’s an interesting question… for later.

Potentially disruptive – but not clearly so.

Could be emulated, perhaps better, by Bing using Facebook’s social graph and their Like button data. How? Well, if Bing did this, then made a “plus 1″ boost results for a while, “until the algorithm settles down”, then SEOs would want to get a Bing button everywhere and people to click on them – and that would probably help clean up Bing results, increase Bing usage, and is another reminder on web pages of Bing’s presence in some markets (still just an also-ran, here in the UK). So, disruptive but not exclusive… and open to a competitor taking it higher, further, faster.

So – what’s it like? See this…

… and do, please, click it. :)

"Flam" - Flattery Spam example, praising the comment policy as if it was an article.

Next up, has the nature of spamming changed? I believe that it has. On this blog, and clients’ blogs and generally across the internet, I’m seeing a lot more “flam” – flattering spam. “You’re article really helped me with this difficult topic, thanks, I’ll be linking to you and subscribing to your feed” – stuff like that. Contributes nothing but a feeling of pleasure… and sometimes even sneaks past Akismet, if the author has managed to lull the suspicions of bloggers.

Of course, the acid test is whether my attempts to deflect spammers to the Comment Policy page have been effective.

Results

This WordPress blog's Comment Policy attracts most spam, now.

After seeing the trend towards “flam”, I’ve gone back throgh comments here (and on client sites) and made sure that every comment is substantive. That is, it contributes to the material in the posting and would be helpful to other readers.

That in turn, raises the question of what to do about those rare cases when someone says something useful, but then leaves a keyword loaded username. IMO, the best thing to do is to re-write the name as something more sane. After all, the comments here are no-followed, and it is, bluntly, rude to post under a keyword rather than your own name – not to mention being outside the guidelines for marketing laid down by the Advertising Standard Authority and described in the Code of Advertising Practice.

Recommendations

If you run a blog, there are three levels of protection to consider, basically depending on the volume of spam you get – which in turn is a reflection of your sites’ popularity with search engines (more on that, in another article I have in progress).

We have a general recommendation which is that marketeers have to reduce the barriers to engagement. Philosophically, I’m opposed to the idea of making it more difficult for a user to add a comment to a blog, to defend from a spammer. It’s just wrong. The burden should be pushed to the spammer, not the user. So, wherever possible on any form, we avoid CAPTCHA (which is, itself, subject to easy attack from Mechanical Turk style systems – use automation to find the CAPTCHA and just push the images to paid-by-activity humans who’ve signed up for the service).

If your blog is not popular (up to tens of visitors per day), then basic tools like Akismet are absolutely fine – though too infrequently used on small blogs, IMO. You’ll probably get no more than a one or two “flams” per day – and the rule is “if the comment could have been applied to any article you’ve ever written, it is probably spam”.

If your blog is moderately popular – hundreds of visitors per day – then Akismet or other crowd-sourced or Bayesian detectors are probably not quite enough; consider using a honeypot to distract spammers to a page or posting where you know that the overwhelming majority of comments will be spam. This, too, should keep the count of suspect spammy comments that need attention down to a handful per day.

If your blog is seriously popular – well, you can probably afford to have a few annoyed users who can’t get through the defences. Very popular blogs might want to consider both a crowd-sourced and/or Bayesian and CAPTCHA solution of some sort. That’s not as extreme as some US magazines who have ludicrously tough comment defence, added in pursuit of finding candidate new subscribers, rather than spam reduction, I suspect, and an illustration of the maxim about reducing the barriers… I cant be bothered to sign up, when they make it so difficult.

That “reducing the barriers” stuff is really important. Take a look at Bob Cialdini’s Influence – an old book, now, but still germane as it deals with fundamentals of human behaviour. Until you’ve given something of value to someone, they are unlikely to give you much of value. So asking for a full name and address and other contact information, just to drop a comment, is fabulously in excess of what most people will be comfortable to do. In the wake of the Sony hacking, users should also be becoming more aware that leaving details on publicly addressable sites is dangerous. And thats compounded by poor security practices from large organisations who confuse users’ ability to detect safer and less safe places to submit details (see my earlier article on Google’s absurd inability to help users identify which are trustworthy Google URLs and which URLS can be abused by scammers).