Kerstin Baker-Ash

Google's Buggy Floating Maps

Google's annoying floating map issue could've started out life as an engineering bug, quickly snowballing into a real pain for paid search marketers. Not familiar with the issue? Basically, scrolling down any page with map results past the imaginary fold results in the non-premium ads on the right hand side of the page to disappear underneath the map.

It's annoying from a UI point of view - Google Maps follows you down the page, but weren't you scrolling down because the top half wasn't necessarily relevant to you? It's also very frustrating from a paid search marketing point of view. Some SEM folks have commented that their best conversion rates or ROI result in clicks on ads in position four - that disappears when the map disappears the ad. Suddenly, your SERPs go from this:

Turns into this:

At the bottom of the page, bloomsnroses.com completely disappears.

It's a head scratcher, only because it clearly isn't designed to help with UI - it's more of a revenue blocker. I can't see how that makes any commercial sense, but it's been going on for a while. Oversight, unfixable bug or buggy revelation?

Direct Online Marketing recorded a video of this back in October that shows the buggy floating maps results in real time:

 

Google AdWords Interface Bugs

Why not start the year off right with a slew of bugs in Google's AdWords interface? The latest bug hit just during final year report times, and seems to not be able to show data on certain accounts prior to October 2010 in the Campaigns summary tab.

Whitewashed for privacy, there's absolutely no information (clicks, cost, impressions) available in the Campaigns Summary Tab prior to October 2010. The information isn't available when downloaded, either. Getting nervous, I went to the change history to see if the campaign had been modified in such a way that the information would no longer be available.

Like magic, campaign information appears (it's a seasonally relevant campaign for January - April, in case you were wondering). I've got clicks, impressions, cost and no idea why this information isn't available in the Campaigns summary tab. No quick answer today either, because it's a bank holiday.

If anyone has information on why this is occurring or how to rectify it, I would be eternally grateful.

Persistence and Spelling

Really. It's staring you right in the face.

Kerstin Baker-Ash

It's not a hard name to spell - especially the two parts of the last name. So why is it so hard for people to learn?

Someone in the Toronto area seems to disregard spelling my first and married name correctly rather consistently. I'm not one for outing people on my blog, but I find it fascinating that this person visits nearly every day, spends a minute and a half on the site and always arrives after searching "kersten baker ashe":

When you're using Google Analytics on any site, you tend to look at the wider picture rather than granular, minute details. But if a big traffic driver to your site is a misspelling, don't overlook it. Brand names are often misspelled - it could be a typo or genuine ignorance of the correct way to spell your brand, but you need to cover all of your bases. 

Google frequently uses the "did you mean" more popular spelling in search results, negating the need to fully optimize your site (SEO-wise, at least) for your less informed or typo-prone customers. It's a crutch that's easily and freely used to cover a persistent misspeller, and you can largely rely on Google to direct people in the right direction if it's a common enough mistake. What about when you can't rely on Google's auto-correct benevolence to make sure your brand is covered?

Paid search. Yes, you'll have to pay. No, it won't cost you that much. Take a look at the traffic being driven to your site on brand terms. How are they spelled? What's the typical dwell time of a misspeller (in my experience, it's more than someone who has spelled it correctly)? How much is that customer worth to you in terms of repeat visits? Your boss and their ROI will appreciate investing in a couple of cheap, low traffic spelling mistakes.

Chances are, you're the only one that's going to be bidding on your list of misspells. In a world of less than wonderful business ethics, competitors often overlook the lucrative and cheap misspells in favor of brand coverage. So, maximize your brand reach and "all eyes are on you" potential by compiling a list and bidding on it. One thing to reinforce however, is that your brand is spelled a specific way; do not use Dynamic Keyword Insertion (DKI) because it makes your ad look cheap and hurried. Your customers are important to you - don't cheapen your brand by being lazy.

If your misspellers tend to spend more or engage with your site for longer, offer a bit of an incentive to them. Put your misspellings in a separate ad group or campaign, and create custom ads for what I affectionately call "monkey fingers". It doesn't have to be much; if you're saving 20% on the cost per conversion, offer your monkey fingers 5% off or free shipping. You won't create more misspellers by offering this ad, but you take the opportunity to show the customer you care about your business while reinforcing your brand and brand image.

It's also possible that my mystery searcher uses autofill - yours could too. Their persistence could just be laziness, but you still want to make sure you're reaching them when they're trying to engage with you.

To my mystery monkey fingers, I'd love to know what you'd like to see out of this blog. Are you looking for more information on search engine marketing? On Kerstin Baker-Ash? I'm sorry you're not always finding what you're looking for, but I'd be happy to discuss with you what it is that keeps you coming back, day after day, looking for updates.

The Best Analytics Apps for iPhones

Own an iPhone? Analytics junkie? I've reviewed four analytics apps (at my own cost, no disclosure needed here) to help you choose the best one to get your fix.

Analytics App

One of the more robust tools on the iPhone market, Analytics App is a great tool to monitor multiple sites across a single login. That is, when it works. I've had login issues for the last couple of weeks - fair enough, they say that you can try refreshing the captcha on your Google account via Safari on your iPhone to get it to work, but I've so far been SOL. When it does work, Analytics App is stunningly easy to use and more than comprehensive. You want access to your custom reports created in Google Analytics? Goals? Event Tracking? Done. My one little gripe is that you can't drill down into geographic content. 

Since it is, in my opinion, the most comprehensive analytics app available and I've had it since February 2009, I haven't put in for a refund because I'd like to give them a chance to correct the issue. I still wouldn't bother downloading it if you haven't got the tool already; they don't think they'll be fixing it until late October.  

Touchlytics

The new kid on the block from New Street Software, Touchlytics has a few more bells and whistles than Analytics App does. I bought the Pro version ($4.99), with access to 40 different reports. One feature I really like is the ability to tilt the phone sideways to get a visual representation of your site's results - handy if you're with a client or colleague and need to show a segment that might not be on the presentation you've already prepared.

Touchlytics also has a few nice but not essential frills on the side. Shaking your iPhone updates the data (much like the Facebook app for iPhone), and you can doubly protect your data from prying eyes with another passcode lock (if your home screen lock isn't enough). Helpfully (and not frilly at all), there's a drilldown by country location, but it would be even more handy if this drilldown went down to regions or cities. You'd need this when you're launching a multi-city campaign with the same creatives and needed to tell the client what worked in Cheyenne didn't work in Boise.

The drawback? It's a bit clunky when you ask it to perform multiple tasks sequentially. I've found myself staring at a black screen for a few moments while the app shuts down and takes me to the home screen. Annoying if you're trying to show people lots of data, but not a dealbreaker - you can get right back in to the report you've just been looking at.  

Analytics Agent

Global Agent Inc offers two versions of its Analytics Agent, Pro and Lite (the latter being free). It's not clunky by any stretch of the imagination, but the instructions are written in such a way that I didn't feel comfortable shelling out the £2 to buy the Pro version. It would be a relatively sexy interface too, if the advertising on the Lite version didn't cover up the title of the report you were trying to access.

Obvious drawbacks include needing to sign in every time (even though I'd set the application to remember my password) and the fact that this was released before someone with better than average English skills gave it a once-over. In the Lite version, you're not able to select the date range you'd like to see, and I don't know if it's available in the Pro version because the features on the Global Agent site are written in Japanese. This is definitely an application I'd give a miss, even at the free price tag.

Ego

We all have big egos in search marketing, right? Ego will help you get your fix of followers, site visits and stats at a glance, but it's not comprehensive by any stretch of the imagination (fair enough, it's not marketing itself as such). With all your ego-feeding needs at the touch of a button, it's fair to say that this is the harried man's version of a bolder analytics package.

There's really only support for three major ego fixes though. You've got Twitter, where you can check how many followers you've gained or lost overnight and what your last tweet was - handy for remembering a particularly saucy night out and making a mental note to delete the tweet as soon as possible. There's Google Analytics, and it's dead simple to check how many page views you've had today, yesterday, this week, this month and this year - not anything beyond that though. Then you've got Feedburner, where you can check your stats to make sure your last post didn't tick people off enough for them to remove your blog from their feed reader. There's also support for Ember, Mint and Squarespace, but since these are fee-paying analytics or hosting services, there's not much use for those of us who don't subscribe or host our blogs there. Hopefully, they'll add TypePad (what this blog runs on) support soon. In the meantime, it's a fun little app with a sexy interface - just don't expect to get much meaningful data beyond visitors and followers.

The Bottom Line

I would hesitate to recommend Analytics App because it's not working for me right now. It's too bad, because it's simple to use and doesn't send me into home screen frustration if I ask it to switch between accounts. The winner by default (and by good design, it must be said) is Touchlytics, who also do a Lite version for $1.99. The functionality is superior, you can extract meaningful insight and data, and the frills add a cherry on the top of an already fantastic analytics app.

Paid Search Sitelinks

First reported in August and September by Darrin Ward (USA) and iCrossing (UK) respectively, Google are now allowing some advertisers additional real estate at the top of the pops in paid SERPs. While this is largely a helpful step for PPC advertisers, what isn't discussed at length is the implications on natural listings, and the impact that has on SEO.

 

It's important to note that there are negative implications for brands invited to participate in the test. Firstly, the tracking needs four new layers added to a campaign to test the efficacy of sitelinks off one brand term. This can be a pain if you set it up with older tracking software which then requires individual tracking tags on the four different pages. A headache, but not an insurmountable obstacle. 

You can't measure brand engagement on a home page when you deeplink searchers to a microsite within the main homepage. The advertiser provides Google with a list of ten possible sitelinks; while they respect the order you give those sitelinks in, Google reserves the right to change your list without notice depending on performance. These are brand searches, not additional query strings with pets, motors or business tacked on to the back - it's also possible that searchers will not feel like their particular query (let's say content insurance) is being met, even though they searched on the pure brand term. Let's not discuss additional revenue generation methods by Google, it's something akin to beating a dead animal with a nuclear weapon.

On the positive side of the argument, you have the potential to give your brand extra prominence (in paid SERPs at least, more on SEO later). More users can engage with different strands of your brand - pets, motors, business, etc - without having to search or navigate through different parts of your site. Let's also not discuss the conversion potential for pure brand queries in paid search without additional sitelink help, either - see dead animal/nuclear weapon reference above.

Does this hurt traffic to natural listings? Potentially. When pages in the natural part of SERPs are pushed further and further down, this can't help with a case for SEO and brand optimization. Forget about outranking on key generic terms, either - if there are three paid ads at the top of the SERP and one of them has sitelinks, you can pretty much forget about appearing above the fold on most computer screens. To be fair, paid sitelinks are largely within the domain of brand terms, not generics. However, the rollout potential is dangerous when you consider the work that goes in to jumping even a one place above a competitor in natural rankings.

It's far too early to comment on the efficacy of Google's newest PPC beta test, but it will be very interesting to see if it increases brand engagement on different areas of a site or drives searchers to exactly the same place they wanted to go in the first place - the home page. 

Exporting More Than 500 Results in Google Analytics

If you're a Google Analytics super-user-junkie, you'll sometimes run into problems exporting large keyword lists for your own entertainment or actual client work. There's a quick and slightly dirty solution to this (thanks to Matthew Attwell for the tip):

Firstly, go to the bottom of the Traffic Sources > Keywords list:

 

You'll likely have more keywords than the "allowed" 500 - look at the end of the URL:

Where it says &trows=500, simply change it to 50000:

Don't expect the visual result (i.e. more than 500 keyword rows) to change. Instead, export the document to a .CSV file:

Unfortunately, 50,000 rows is the maximum that a comma separated value sheet will allow - once you've changed the limit, you'll be able to export a relatively full keyword list. 

Simples.

Update (23/06/09): Rob mentioned in a comment below that changing the trows parameter didn't work for him. If you're having problems getting it to change, set the limit parameter to 50000 instead of trows.

What Did I Miss?

I've been operating outside the blogosphere for a while. Holiday, catching up at work and getting over what seems to be a rather protracted case of jet lag has consumed every waking moment for the last couple of weeks. I won't go into a long holiday recap - it was fabulous, I saw my family, danced the hora at my friend's wedding and drove (rather, James drove) over 3200 miles from Toronto to Philadelphia to Orlando and back again. I ate at a haunted restaurant in Savannah, nearly tossed my cookies on a centrifuge ride at Epcot and drove through Louisville, Kentucky on Derby Day (no Mint Juleps for me).

It looks like a lot of meaty search news happened while I've been away or under a two-week-big pile of work. Wolfram Alpha is launching on the 18th of May. Google presented Search Options and Wonderwheel. Twitter caused some geek-rage when it changed the way we see @replies.

On a work-personal note, I'm up to bat for two conferences in the next four weeks; SMX London, where I'll be speaking about writing killer search ads and landing pages, and SES Toronto, talking about internationalizing your campaigns and sites. In the next day or so, I'll go more in depth about what you can expect from the session, as well as pointers to my awesome fellow presenters' teaser trailers. I'm über-excited (as usual), because the calibre of conferences is definitely on the rise, and I get to share a panel with people I completely look up to (without the girl-crush element) in the industry.

Back to the grind, work and presentations to crack on with!

HSBC: No "goodservice" in Sight

In the past few years, companies reaching out to a new generation of consumers have oft-used "search for" slogans in their above the line communication.Orange's poorly planned "I am" campaign has presumably taught us that in order to make sure you're searchable on the terms you're telling people to search on, you have to do a little behind the scenes legwork beforehand. Sadly, one advertiser that has followed Orange's example of putting the cart before the horse is HSBC. 

HSBC's new outdoor campaign suggests that instead of calling their 0800 number (that's like a 1-800 number for the Yanks and Canucks among you), you should search on "goodservice". The first result is poor; Google suggests an alternative spelling, and returns two top results from Canada, as well as a "goodservice" result from a Portuguese site. HSBC could've got around this by using paid search as a bridge until their SEO ramps up, but instead Audley Travel are they only ones showing up in sponsored links:

Fair enough that Google suggests I've made a spelling mistake, so I searched on "good service" with a space in between:

HSBC doesn't pop up anywhere. In fact, the first instance where they do show up is... nowhere. They're not to be found in natural or paid search results. Duncan Bloor first wrote about HSBC's goodservice campaign nearly two months ago, and they did seem to be using paid search to prop up their abysmal SEO. However, someone has massively dropped the ball and either let PPC budget run out or not ramp up a successful SEO strategy.

Thanks to my colleague, Jonathan Keane, for spotting this search #fail.

I Don't Believe in Broad Match

Again, another slightly misleading title. I broadly don't believe in broad match with very few exceptions, none of which fall under user relevancy. I strongly believe that setting everything at broad match is reckless, and suggesting it to clients isn't best practice. This isn't meant to be a sweeping swipe at Google, but I've been left cross-eyed by the numerous reasons for broad match and wanted to get my opinions to the contrary down on a piece of paper (or blog) to make my thoughts clear.

Broad Match Sell #1: ~25% of search terms each month are unique queries.

Great. Humans are different from each other. Wonderful. What that statistic fails to include is the fact that 1% of search terms represent nearly 30% of all queries, according to Search Tools Consulting. What does that mean? Your generic or brand name keyword is singlehandedly responsible for a third of your traffic. Another five eighths is taken up by keywords already in your account (if you've done a long tail keyword buildout - and if you haven't, well... see me after class). We're talking about five percent of potential queries. Potential queries which could be covered largely by phrase match:

Unique query: "summer 2009 travel"

Phrase match: Summer 2009"

Result:

So a couple of search terms come up, but Google and the advertisers largely know that I'm looking to travel in Summer 2009 and have responded to my query appropriately. 

Broad Match Sell #2: We're not going to talk about cost. Nope. Shhh. My fingers are in my ears. I can't hear you! Lalalala.

Next time you're confronted with a "unique relevancy" pitch, try this. Ask them how much more it costs. You'll usually get a roundabout flurry of phrases that sound like "user experience" and "search relevance" but nothing to do with your original question about cost. Why? Because it does cost more. That's the honest truth. Exposing yourself to search queries which aren't entirely relevant to your search term does one of two things:

1. The searcher sees your ad, decides it isn't relevant to their query, and creates an impression without a click for that keyword. Quality Score suffers.
2. The searcher sees your ad, decides you might be able to help them, and clicks through on your ad before leaving the site quickly - you didn't fulfill their expectations. Quality Score suffers, and you pay for a click which never had a chance of converting. 

Wonderful. Two negative outcomes, no positive outcomes in this particular scenario. Just for some graphic representation, take a look at this search I did on "flights to Canada":

  

Fair enough that Ask is appearing, well, everywhere these days in an effort to gain some market share. Let's pretend they don't have a long line of credit with the bank for a moment. I'm not going to click on the ad, but hey - someone might. They've created a situation where their Quality Score is affected by a non-click impression or irrelevant cost for a click that has about as good a chance in converting as Billy Graham does to Judaism. Just so that it doesn't look that I'm taking swipes at Google only, a little broad matching debacle from Yahoo!:

Low Quality Score = higher cost per click. The question to ask your Google rep is why you would want to be paying for irrelevant traffic or bringing down the Quality Score you've worked hard to achieve. Hopefully, their answer is a concession. Yes. You will pay more. No, it makes no sense to broad match on travel when you only offer packages to Botswana.

Broad Match Sell #3: It can be a branding experience.

Okay, I'll give them a bit of "branding experience" leeway for a second. This sits straddling a broad match and a trademark post, the latter of which you all know my opinions on. Just because you can broad match on terms doesn't mean you should. It also doesn't mean that you are perceived as any more relevant to the user, or that you will brand yourself by appearing on searches. Sit back for a moment and think about the last time you engaged with a brand. Here's mine:

Andrew Girdwood tweeted about a tea that he was drinking. In this social media world, it's the equivalent friend telling you about a product that they really like over your morning coffee (or tea). I went to the site, looked around, liked what I saw, and made a purchase. Yes, I have searched on "interesting tea" before, but the engagement with the brand occurred only because someone in my social circle was talking about it. Unless Google/Yahoo!/MSN are willing to give up any empirical evidence to the contrary, I would say that appearing on a broad search does little in the way of a brand lift. Just ask Ask.

Remember, Google and the other engines are sales people too. While it is their job to help you optimize your accounts and get the most out of your campaigns, they're also revenue targeted. With that in mind, make sure you weigh all your options when you explore your match types, and make a decision that is good for you.

The Holidays Might Be Over, But Are Yours?

Living in a world where technology has evolved in service of our business and personal needs, you'd think brands would understand that changing ad copy over the holidays would be beneficial to their brand perceptions. It surprises me how far pay per click marketing still has to come to be regarded as a profession; right now, it seems like it's still chock full of half-baked attempts by offline marketing managers interested in "that internet thing".

As simplistic as the process of ad scheduling (dayparting for you old school types) is, it seems to have escaped a number of retailers this year. In the spirit of the holidays, I'd like to offer a very quick and easy way to ensure that 2009 is the year you start looking like a rockstar to your board.

Step 1: Create two campaigns - one for pre-holidays and one for after. This is especially important if your business is expecting to put on a Boxing Day or January Sale.

Step 2: Create a list of keywords and copy them across both campaigns so that your keyword inventory is identical. This only makes sense if you don't have any seasonal-specific keywords (we'll get to sale keywords later); don't copy "Christmas Turkeys" over into your post-holiday campaign - chances are the people that are searching for that keyword might be looking for something else.

Step 3: Write some ad copy. Keep it relevant to the season - Christmas Special: Socks on Sale, Guaranteed in Time for Christmas is great for your pre-holiday searchers, but make sure you create an equally relevant ad for your post-holiday traffic. Something along the lines of 20% Off All Socks - Boxing Day/January Special! usually works a treat.

Step 4: In "Edit Campaign Settings", opt in to ad scheduling (i.e. turn it on!):

Step 5: If you've never turned Ad Scheduling on before, you'll have to specify your time zone. For single time zone countries like the UK, it's relatively easy. For US and Canadian marketers, make sure you select the right time zone for your business (if you have a physical location in Houston, it makes zero sense to set your time zone to pander to Yankees on the East Coast; make sure you set it to Central). Setting up different campaigns for different time zones is another post altogether, so we'll keep it relatively simple. Remember - you cannot change your time zone once it's been set. Make sure you don't accidentally set it to Tapei (GMT +8); it can get a tad confusing.

Step 6: Christmas Day was on a Thursday this year, so you can set your ads to go off  without interrupting your turkey conquest plans. Name the time you need your pre-holiday campaign to go off, and it'll go off. Then, all you need to do is make sure that you do the same for your post-holiday campaign - except this time you schedule a time to turn it on.

This is a pretty simple way to protect your brand reputation online. For instance, I'd be very unlikely to purchase something from the retailer at the top of this post, just because they haven't shown me that they care about their products or their brands enough. I know it's not advanced PPC techniques, and there are a load of programs out there that can switch on and off by specified date - this is just a ridiculously simple, band-aid fix to make sure you look like you want your customers' business post-holidays, even if you're still sipping mojitos in Varadero.