Friday, January 22, 2016

SEO finally comes to organic Facebook posts

Noticed how your free Facebook posts have had fewer and fewer people view them? I don’t blame paid ads for this.

Facebook tells us that on average users would have 1500 updates to see in the news feed each day. Facebook is partially to blame for this as they’ve made so many actions potentially news feed worthy – friends liking stuff, for example. The platform uses interest based filters to decide what to show people. If Facebook thinks a user is likely to be interested in a post then they’ll show it to them. That's tough for marketers. If only there was a way to nudge Facebook's system to consider a particular audience as you're pretty sure they might be interested in your post.

The problem with the busy news feed is that it’s now full of brands, pages and your next door neighbour all keenly sharing stuff that you might like. Brands suffer the most because they work hard on “branded content” or other expensive, quality, content that meets their own branding rules while trying trying not to look like corporate messaging. Facebook natural reach is declining because it’s busy; not just because there are ads.

If you’re sceptical then watch this video from Facebook’s Eric Sodomka, which explains this in detail. Watch it before Facebook realises what’s been said and asks the Simons Institute to delete it.

So what’s this about SEO for Facebook?

Wouldn’t it be good if you could use tags to optimise a free, organic, post so that its targeted specifically at people within your audience who might like that tag?

Sure, right now you can use Facebook’s Interest Targeting feature to mark posts – but why would you? What Interest Targeting does is to restrict who can see the post to only those who match the interest. Essentially, all that Interest Targeting does is reduce the chances of your post picking up Likes and Shares.

Today, Facebook has announced that Interest Targeting will go and will be replaced by Audience Optimization. It talks to my SEO genes.

Facebook’s new Audience Optimization comes in three parts;
  1. Preferred audience - tags that encourage Facebook show the post to people who might be interested in the tag.
  2. Audience restrictions - a bit like the old interest targeting; you can say who might not be interested in the post.
  3. Audience insights - want to see how the content actually performed against your tags? This will show you.

To recap;

This coming week, for English language Pages, you’ll be able to call out interest tags for your posts and Facebook will optimise the delivery of those free posts to matching people.

You’ll be able to analysis how successful your interest tagging strategy is. You’ll be able to see how different subsets of audiences are engaging with your content.

This is powerful

Imagine what can be done with this. Off the top of my head;


Brands will be able to see which of their celebrity product champions actually resonate with their audiences and sports clubs will be able to see which brands might be the best match for them (or take the money anyway and apply audience restrictions to all-on going posts to reduce the impact of the sponsorship to fans who couldn’t care less about the sponsor.

Viral building

By deploying a series of test and learn posts marketers will be able to determine which interests in their audience are most likely to trigger shares and likes. Brands will be able to target vital candidates.

Media companies will be able to see which news angles work best with which audience subsets.

Linkbait and SEO

Using Facebook to help connect linkbait content with audiences with a high propensity to create links and other quality signals?

Facebook’s Audience Optimization for Publishers will allow brands (aka publishers) to refine that strategy; offering up insight on both the type of content to use and the audience to target.

I need to know more

It’s early days yet. Audience Optimization is rolling out as we speak and needs to be switched on by hand if your page has less than 5,000 likes.

It’ll take practice.

Facebook has already published Introducing Audience Optimization for Publishers and an Audience Optimization Get Started guide.

Here’s hoping your social media platform supports the new feature. Eh, Hootsuite?

Wednesday, January 20, 2016

Have we lost it? SEO integration with design and build

There was a time when websites were built by people who didn't have SEO high up on their agenda. It was common for brands, big and small, to create a whole site and then appoint an SEO agency as part of their launch strategy.

It was rubbish.

New sites went live with all sorts of serious problems. Huge amounts of money were wasted as big changes were made. Agencies were left looking pretty silly.

Things got better. With that much money in the equation progress was inevitable. Four things happened;

  • Clients / brands understood the importance of technical SEO in the build
  • Web developers improved their SEO skills
  • SEO agencies dealt with senior decision makers
What about the fourth? An important but often forgotten consideration is this;
  • Google got better
Search engines can now cope with a wider range of web design and coding techniques. I still wouldn't build a site in AJAX and frames but Google is better able to understand your JavaScript, for example, than ever before.

What's the problem with SEO and web build?

SEO changes. The stage in the web design and build project that SEO is generally thought about hasn’t changed.

Back in early 2014 I wrote about Innovation in SEO and the "technical build". My argument back then was that getting the coding right, worrying about URL structures and similar wasn't enough. The build strategy in today's SEO world had to consider how the site would operate now that we're all publishers.

Examples I gave at the time included;
  • Having an easy and appropriate location on the site for linkbait publishing
  • Landing page strategies for retargeting, affiliates, etc
  • A page retirement strategy built into the CMS
It's now 2016 and I see very little progress. These considerations, and others related to current SEO techniques, are more important than ever. Sites are going live with an SEO disadvantage and CMOs and CIOs are beginning to waste their money again.

When was the last time an SEO took a look at a core revenue driving page of your site (a product page, for example) and said "That would be an excellent page for outreach"? Or "It'll be easy to make that page a key part of the publicity campaign"? Have you been asked by the boss to prove that the blog is driving sales?

Are we really all publishers now?

Yes, whether it's Facebook updates or your Fitbit sharing your weekly step total in an email to friends; we're all creating content.

When it comes to SEO this means we've so many more brands competing for attention. Just take a step back and look at all the companies still pumping out infographics. Look at the brands working with bloggers on co-creation projects. There's been tremendous growth in branded communication and content.

This all equates to publishing. Publishers (games, books, music...) are on the same evolutionary path as marketing agencies. It's about growing audiences and making money from those connections. It’s too early to say where we’ll meet – transmedia campaigns based around IP franchises with profitable commercial models and loyal fans, perhaps – but there will be a meeting. Just look at publishers like Buzzfeed or The Drum who already have their own agency-like agency assisting content units.

There may be some who haven't come to the same conclusion as the rest of us. They may not think we're in a messy battle for attention. It doesn't matter as even this small tribe recognises that modern SEO is very much about getting certain audiences interested in and aware of certain content (even if they’re just doing it for the links).

What does this mean? What do we need to do?

It means that we need to start making web sites - digital assets for our company - that have a fighting chance during the coming years of "peak attention".

Brands need websites that will be relevant and useful. Having a web site with a blog annex isn't enough.

Sure, you might have a cool video of a fashion vlogger trying on some of your hats but do you have the means to turn that into a custom landing page, social enough to be interesting and yet ecommerce savvy enough to sell hats? No? Just going to publish the video on a blog post, link to some of your hat product pages and email some fashion bloggers to see if they care? That's not going to be enough.

In this example, if you had had the opportunity to integrate modern SEO in to your new website then your hat product pages should have been able to display the video. They should have been able to act as a hub or anchor point for the story. If you were, somehow, able to get fashion bloggers interested in the video then such an approach would hugely increase your chances of earning those vitally important editorial links to your product page.

This is hard

Yes, this is hard. This is one of the reasons so few people do it. A range of skills and collaboration is needed.

Another reason why so few sites are built in this way is because SEO is not a consideration while this sort of site design is being thought about. It's rare to find SEO minded people in the room so early on in a new website's lifecycle.

But it should happen

My hope is that SEO can get back to where it needs to be when it comes to site design and build.

New projects like AMP may help - sites being expanded with a Google defined mark-up designed to load quickly in mobile.

The importance of SEO will help. Savvy client-side decision makers will help too and there are plenty of those (fighting against old company structures and silos, generally).

Image credit: Kevin Dooley.

Monday, January 18, 2016

Why Google's new Visit-in-Person query type deserves some attention

SEO is always changing. One of the reasons why true SEOs are rare is the skillset changes. The changes are sometimes subtle which require instinct as much as hands-on experience to detect and master. I think Visit-in-Person queries are one of those subtle, but important, changes.

One of the biggest Google updates ever was Florida. It hit just before Christmas and destroyed the rankings of retailers. How could retailers, almost universally, all loose the search positions? The Florida update was the one which brought the idea that people like to research before they buy to Google. The Flordia update helped create the three core query types Google used for 10 years.

  1. Navigational Searches
  2. Research Searches
  3. Commercial Searches

Those three core search types have different names today but they’re still around. There’s also a fouth – Visit-in-Person.

Page 72 of the formerly confidential but now public Google Quality Tester Guidelines spells the differences out for us.

Know queries (which we might have called Research before) are those for which the searcher is trying to find something out – which tablet to buy, for example.

Do queries are those in which the searcher is trying to take an action – buy an iPad, for example. There’s a subset of Do queries known as Device Action and these are incredibly important to know about as well. Device Action queries are those commands give Google to control your smartphone. Google gives the example [open facebook app] as an example.

There’s Website queries in which the searcher is looking for a site or a page. I speculate “Navigation” isn’t quite as an appropriate term as it was before as users may wish to navigate to apps on their phone.

Visit-in-Person queries are exactly what they appear to be. Google’s illustrated guidelines hardly seem necessary. These are search terms which suggest the user will make a physical visit to a real world location.

It’s easy to see how this might influence SEO going forward. Here’s a whole chunk of searches in which a digital asset isn’t the ultimate destination. These searches won’t drive traffic to your site.

Visit-in-Person queries are easy to understand but they’re difficult to study and predict. Using Google’s guidelines we can list all known queries with Visit-in-Person intent.

  1. [chinese restaurant]
  2. [gas stations]
  3. [pizza]
  4. [yoga class]
  5. [coffee shops]
  6. [movie showtimes]
  7. [car repair]
  8. [dentists]
  9. [bank of america atm locations]
  10. [starbucks near me]

It’s interesting that Starbucks needs a “near me” qualifier on it to be considered a VIP whereas [chinese restaurant] doesn’t. It’ll be the brand effect.

Neither Starbucks nor any nearby Chinese restaurant will be aware of the search in their own site analytics. Starbucks will be able to see the search term driving impressions but not clicks in their Search Console provided one of their URLs is included in the mobile results. That’s an interesting SEO decision – do make the effort to create pages, with sufficient authority, to rank well enough to allow you to track a query?

One of your main traffic drivers, perhaps a former Website/Navigational search could become a VIP search overnight. That’ll knock your numbers for six. I’ll use [pizza] as an example here. Google’s data tells them people are looking to buy pizza from a shop and not look up the recipe. That would be tough to swallow if the keyword has been a big traffic driver for your food blog or magazine site.

The implications go on. VIP searches are matched to locations, not websites. You may not even need to have a website to benefit from them.

As it happens [burger] is also a VIP search (at least from my location – all searches are location dependant). The results include a small restaurant around the corner, which doesn’t have burger in its name, which doesn’t have the keyword in its Google Local description and for which the word “burger” appears in only a few reviews. So once again we cycle back to the importance of UGC even though the very Google Quality Tester Guidelines that explain Visit-in-Person searches urge quality testers to be cautious with UGC.

It’s also worth noting that many venues had the word “burger” in their UGC and weren’t selected for the VIP results. In fact, one of Google’s three suggestions was over a mile away whereas venues only a few hundred metres away were omitted.

It’s important not to write Visit-in-Person searches off as just Local searches. They’re not. They’re more specific than that. There’s a stronger, clearly defined, offline intent with a Visit-in-Person search. It’s not the same thing as looking for a local florist to order online from.

It’s also important to consider that Visit-in-Person query might be the first, not the only, internet to real life mapping Google supports. Connected Cars may allow for more and Google does have an OS for cars. Just as the ‘Do’ category has the Device Action sub-category; we may Visit-in-Person grow to include specific actions or means of locomotion.

Wednesday, January 13, 2016

How badly are bloggers poisoning the well?

It’s still fashionable for brands and agencies to work with bloggers. How long will it be until it may be harmful to do so?

The problem is with disclosure and authenticity. Bloggers aren’t keen on either and there’s danger that either the public or Google will notice. Google, in particular, is a stat crunching monster. Once the algorithm decides that mathematically the chances are that links from blogs will have been arranged or incentivized and therefore should be a negative quality signal rather than a positive one?

I’m not just picking on bloggers in this post. Back in 2014 I wrote a piece that made it clear agencies and client-side teams were encouraging this behaviour by not refusing to take part. If a blogger takes cash for a piece and “forgets” to put nofollow on the links for Google or to clearly state “this is an advert” for human readers then all too often that’s seen as a bonus win.

That same year I wrote tips for evaluating blogs that tried to steer people to safety. In particular, the Colleen/Jac test was designed to see whether the blogger actually had something unique to contribute or whether the platform was just sprouting reviews, rewiewbait (posts written to score freebies) and paid-for placements.

Since 2014 I don’t see much of an improvement.

I appreciate bloggers, more and more, want to make a living off their blogs. Blogs want to be treated like businesses. That’s fine. I just wish more blogs could actually be given business rates!

What’s the average CPM for a remnant or RTB ad brought from a big brand publisher these days? A few pennies? Let’s pop over to the Passionfruit Marketplace and see what bloggers are charging. I like Passionfruit and they’ve given bloggers a better steer on what to charge than the current prices most bloggers ask for.

The average of the 5 most active Beauty bloggers, by month;

  • $17 for a banner ad shown to a 1/10th of the (undisclosed) traffic
  • $9 for a small sidebar ad shown to 1/6th of the blog's 101,000 monthly pageviews
  • $85 for a button ad shown to undisclosed traffic
  • $400 for a square ad (smaller than a MPU) shown to undisclosed traffic
  • $15 for a small banner shown to 1/5th of the blogs 6,000 pageviews. Weirdly the blogger says you're not allowed to promote anything with the banner.
Using the two blogs who are kind enough to disclose traffic figures that makes an average of about 9000 pageviews a month (not uniques) for an average of about $105. That's a $86 CPM.

A $86 CPM is ridiculous.

There are good deals and less intelligent deals on Passionfruit. The goal of the above is to indicate just how rare traffic reporting is – bloggers much prefer social reach – and to indicate how expensive it is to advertise on blogs if CPM is your metric.

Most brands and agencies don’t buy ads on small blogs though. They’re much happier (and better off) using advanced platforms in a real-time bidding systems to show the same people ads for one thousandths of the cost.

What about disclosure and knowing when to nofollow links? Using a platform (who’ve asked to be nameless) who email offers and opportunities to bloggers, I went through 5 “You get to keep the product” offers and found between three and six reviews from blogs for each.

What percentage of bloggers disclosed they got the product as a freebie?

What about nofollow links? Counting only the blogs that linked back.

It’s great to see that %100 nofollow on an incentivized deal. That tells me either the bloggers – more likely the agency – knew the risks. It’s also worth noting that links were not an asked for requirement in Offer 5. That’s the review deal with the most follow links. That’s how it’s supposed to work. If links are editorial then you’ve more chance of scoring that positive quality signal.

It’s also interesting to see that bloggers seem to be better at nofollow than they are at disclosure these days. That’s probably for two main reasons; the bloggers don’t want to risk Google anger whereas still too many bloggers don’t understand disclosure rules.

If you were Google’s algorithm, though, you’d be able to read the clues clearly here. Sites that look like blogs have a high chance of linking out and, increasingly, those are links that shouldn’t count. The only way bloggers can ensure blogs remain positive quality signals is to keep those follow links rare, precious and appropriate.

I’m a blogger. I see hope for the hobby. Travel, beauty and fashion bloggers need to work on their own personal brand as much as their blog – they need to charge celeb rates rather than platform rates. Gamer and gadget blogs have other ways to earn cash – gadget blogs, in particular, should do more with affiliate marketing. Gamer blogs have communities to grow.

Image credit: Thomas Hawk