Steve Hammer on March 5th, 2011

With a number of high profile brands getting attention in the media for questionable SEO tactics, it’s easy to see the suspicion of the career of SEO rising in the public opinion.  It hasn’t risen to Enron accounting levels yet, but hitjob pieces are starting to pop up everywhere, including a recent mashable article.  I’d like to raise a few points in defense of SEOs everywhere.  We aren’t murky trolls manipulating results for even more nefarious clients.   SEOs are trying to maximize signals for the search engines to serve the right result.

SEO is challenging and ever changing

Buy most accounts there are at least 200 signals, both positive and negative that search engines use to determine the most relevant result for a search.  More are being added all of the time, and the weights that are used change frequently.  Just keeping up with the changes is time consuming.

Take meta keywords for example.  Not long ago it was a strong signal that could be used to suggest searches for which a particular page should rank.  Abuse of this led not only to it’s deprecation as a valid signal, but as a possible negative signal in the case of mismatch between these keywords and actual page content.

Personally, I don’t spend much time at all doing SEO for this site.  I certainly know a lot more that I could do, but the time and effort isn’t worth the reward.

Good SEO isn’t working against Google

There’s a perception that SEO is about trying to trick search engines to rank a page, regardless of the searchers intent.  Nothing could be further from the truth.

Traffic on a website has a purpose, a benefit that is to be gained by the use of that page.  That benefit is shared between the user and the website owner.  If there’s a low benefit, there’s not much to be shared.  If the benefit his high a purchase can result, a conversion can occur, a client can be one.  There’s little point to low quality irrelevant traffic.

Certianly there are those that game the system and make decent money for doing so.  However, these businesses aren’t sustainable.  Easy come, easy go.  A trick might work for a while, but eventually will get caught and have a dramatic effect on the business.

SEO is about working with google

Take the following search for example.  “The president of the United States.”   This search doesn’t reveal much of the why, so there are many potential answers.  In my opinion the best result would be the official whitehouse.gov page.  It’s a very specific site with a lot of depth on the particualr subject.  It should have the most authority on the subject as well.

However that isn’t the first result returned.  Wikipedia is the first couple of results, with an entry about the current president, and a list of past presidents.  Other top 10 entries include an ehow entry on how to become president and the site for the band “The Presidents of the United States”

If I were an SEO for whitehouse.gov my job would be to make sure that the Google shares my opinion that whitehouse.gov is the most trustworthy, authoritative and comprehensive resource for “The president of the United States.”  Right now Google is wrong with the results they are serving and I need to send stronger signals that whitehouse.gov is the best page to serve.  That’s not gaming the system, that’s helping Google do a better job.  It’s right for the searcher, and it’s right for the site owner.

Maximize signals for the right result. That’s how I describe SEO.

The first and most important factor is actually being the right result.

It does get a murkier when you talk about competitive searches.  It’s harder to say who should show up first for a search like “eyeglasses” or “shoes.”

However, the best  SEO advice anyone could possible give is to be the best in that space.  Trust. Authority. Depth.  If it’s good for the user, it will be good for the searcher and search engine as well.

The Dark Side of SEO

The part of SEO that’s getting exposed are those that aren’t trying to be the best result, but are trying to game the system to appear to be the best.  It’s short term gain versus long term value.  It’s frustrating to get out ranked by a junk site that is clearly gaming the system.  However, they disappear often as quickly as they appear.   It’s actually Google’s incentive to do so.

In Google we trust

The bottom line is that we all need to believe that Google isn’t evil.  Search engines have to want to serve the best result every time.  As long as the two engines compete on result quality and startups continue to challenge, SEOs have to focus first on being the best result and then maximizing signals, not the other way around.

 

Tags: , , , , , ,

Steve Hammer on March 3rd, 2011

It might sound odd for an annual member of social media club and a former GM of a social network to write a post about some recent over selling of social media.  In fact, there’s a lot that’s still very exciting about this nascent industry.  However, there’s some parts that need to get brought back down to earth.

Over-promise, under-deliver happens too often in social media

It’s pretty easy to go out and talk about some incredible ideas that can be spread by social media. It’s easy to imagine the pitch: “If twitter can overthrow a government, what can it do for your challenger brand?”

The problem that I have with the pitch is that it’s fundamentally flawed.  People care about freedom.  Oppressed people will wake up and dream about how to remove their oppressors.  No one that doesn’t work for your company woke up yesterday and wondered how to grow your market share.  If people share, it’s because they think their friends can benefit, not because they care about a brand.  If there’s no understanding of basic motivations, there’s little chance of success.

Just add some viral to it

It’s another expectation of social media.  Viral.  Everything can go viral.  Viral is not easy, and you can’t just sprinkle a little on top of a mediocre idea.  There are lots of tactics that can enhance sharing, but viral is another beast entirely.

Viral implies exponential growth.  In each generation, more people have to share than the prior.  It’s simple math.  If 100 people per sharing generation don’t generate at least 1% sharing, it’s going to die out quickly, no matter how hard you seed.  Both sides of that aren’t that easy to generate.

The lifetime of a share is incredibly short

The time lines and news feeds of many people are incredibly full.  With hundreds of friends and follows, even a few over sharers can drown out a very relevant message.  Facebook tried to fix it with an algorithm, and that can limit what you see as well.

Social media right now is as attention deficit disorder as it gets.  If it doesn’t compel action at the very moment a friend is exposed, it’s completely lost.  Personally, most activity is on either the newsfeed or the timeline.  I don’t go to profiles anymore to see a users personal story, I consume in real-time.

Measurement is spotty at best

In many ways the measurement of social media is an incredible challenge.  Many people use PR like metrics of impressions.  These don’t mean that anyone actually saw  the message, and result in potentially inflated numbers.  You could say the same about follower accounts or fans.  The number has no indication of engagement.

On the other side would be direct clicks on embedded links.  This is has the potential to be well understated.  Much like scanning the headlines, there are plenty of times where a message is received but the call to action is reserved.  This is particularly true of retail products, where there’s no need to click to complete the ultimate goal.

Lots of people aren’t actually on social media

Sure there are 500million accounts on facebook, and we keep hearing about the massive percentage of pageviews that are consumed there.  What we don’t have a feel for is how many of those accounts are actually active, and what those people are actually doing.   It’s the equivalent of saying how many people use email.   The purposes of email vary wildly, and lots of people have more than one address.
Lots of people just don’t understand it.  They don’t have the need to constantly connect or be bombarded with information.  My wife for one thinks that tweeting something rather than just enjoying the moment is patently absurd.
I’ve seen recent data that indicates that only 4% of smartphone users engage with a location based service like foursquare.  Keep in mind that most users of cell phones don’t have smart phones.  That’s a small sliver of a minority product.

The real purpose in my mind

Everyone needs to remember that social media isn’t really a new behavior, it’s a new channel for an old behavior.  It’s a public sharing of word of mouth.  People talk about what’s interesting to them.  They’re friends with people that add value.  They share things that help.  If no one was talking about pipe tape yesterday, creating a facebook page isn’t going to help.

Simply put:  It’s the content.  It always has been. No matter than channel it always will be.  Give people something they care about and everyone wins.

I’m still bullish on the potential of social media, but I’m tired of the other half of the bull.  Follow me on twitter, and I’m going to spend more time debunking overblown social media predictions.

Tags: , , , , , , ,

Steve Hammer on March 1st, 2011

You’d have to be living under a rock not to have heard the constant stream of crazy coming from Charlie Sheen.  It’s truly impressive to see that some has done some much coke that they think it should be named after them.  Much like Paris Hilton, Mel Gibson or any other celebrity meltdown, it’s hard not to watch the trainwreck.

Sheen was made for Twitter

One of the differences with this current rant is how quotable it is.  It was made it a series of faux New Yorker Cartoons.  From “I have one speed, I have one gear, GO” to “I am on a drug, it’s called Charlie Sheen” it would seem as if the rants were made for sound bytes.  And sound bytes were made for Twitter.

However, until sometime on March 1, Charlie Sheen didn’t have a twitter presence.

Somewhere around 3:00 PM that changed.  @charliesheen burst on to the scene.   He picked up thousands of twitter followers moments after being discovered.  With in a few hours, he already had in excess of 250,000 followers.

The first tweet

Once a critical mass had been reached, the first tweet is posted.  Cryptic with a bit of catch phrase, and an attached link.  “Winning..! Choose your Vice… #winning #chooseyourvice“  The link features a brand prominently placed and gets some 150,000 views.   Truly impressive for an account a few hours old.

The trick – It’s advertising.

Credit is due to Phillip Hotchkiss, the chief product officer of Klout, but the president of an advertising platform called adly posted that he and twitter signed Charlie up.  It’s about Celebrity Endorsements on Twitter.  The motivation for the new account just became clear.

250,000 people just signed up to get spammed and support an out of control celebrity.  Way to go Internet.

What can we learn from this?

As social media practitioners, there’s a lot to learn from this.  Here are some quick bullets:

  • People want to see the spectacle.  The unfiltered access to crazy was just too much to resist.
  • Strike while the iron is hot.  Two weeks from now might have been too late to start tweeting.
  • Keep it quotable.  The 140 characters of twitter are the new soundbyte.
  • Be authentic.  Charlie might be crazy, but you can’t deny that he’s really saying what’s on his mind.
  • Remember that everyone has an agenda.  Mr. Sheen not only wanted to tell his story, he wanted to make money too.

If you have other lessons, leave them in the comments

Tags: , , , , , , , ,

Steve Hammer on February 18th, 2011

It just took a few words for me to sound like a corporate drone.  As  a start-up veteran, It was painful to realize that I uttered the following words:

“I’m looking to get a resource to…”

My head is saying that I have some money to hire someone, and I’m ambivalent as to how they will work.  Contract, permanent, or outsourced.   This wasn’t a situation where technology was the answer.

No matter what happened I was still going to need a person.  How utterly dehumanizing to look for a “resource”

Human resources is a department, not a description

Human resources is a term that describes a lot of functions.  That department doesn’t just recruit they handle benefits, compensation, performance and lots of other stuff.  The right way to look at them is to say that they handle the resources for the humans that work at the company.  They should be viewed as the department that acquires resources that happen to be human.

I couldn’t function without my team, and these days a team really goes beyond who reports to whom.  I expect a certain level of dedication and passion that goes beyond a simple role.  I’ve pushed back against narrow job descriptions and for letting smart people do what they do best.   They make my dreams a reality and add to the richness of ideas at the company.  And then I call a potential hire a phrase that denies their fundamental uniqueness.

I felt like a complete jerk after I said it.

My pledge

I’ve pledged never to call myself a guru, ninja, swami, or any other self-aggrandizing title.  I’m now making the pledge that I won’t refer to a person as a resource ever again.

Who is with me?   Give me your pledge in the comments.

Steve Hammer on February 2nd, 2011

After reading the news that OKcupid was acquired today for $50 million, I couldn’t help but think back to my experience at Downtoearth.  I suspect that commentors that have a long memory will remember Match’s experiment there and say that it is evidence that it failed.   I’d know better than anyone that it’s far from the truth.  Downtoearth was forced to change, plain and simple.

The arguments that users will eventually trade up into paid dating sites was our main business model.  We made this argument when our existence was threatened and were told that downtoearth held match back from a principled stand against free dating sites.  The upside of an introductory model didn’t outweigh the downside risk of people trading down to free dating – at that time.

At that point we were doubling traffic every month, and then we had our marketing budget pulled.  We had to do something that didn’t compete and that was the genesis of Stir.  That’s not a failure, that’s an extreme strategy tax. Given the freedom, we’d still be operating.

Two years later things have obviously changed.   Match has the confidence that their product is truly superior to free dating sites, a contention that that I wouldn’t have argued even back then.  OK cupid executed some brilliant PR and SEO by mining their data, lowering their cost of user acquisition.   It would even seem that the seeds of our trade up argument started to take root.

It’s hard not to be a bit bitter and very envious.  OKcupid cashed out and I’m out of the industry.  Even so, I’m growing business massively at my current employer and that is something I’m proud of.

Tags: ,

Steve Hammer on January 31st, 2011

When I look back at my time at Stir, there are quite a few lessons that are worth relating.  This post is the first in a series of lessons from a great experience.

We had an assumption when we started stir about how we’d make money.   Stir was focused on connecting people with places they would like to go.  We know there were some high dollar transactions in this marketplace.   Ticket reselling through outlets like stubhub can be quite lucrative, and there are affiliate programs that make claiming a share of the sale reasonable.  We also know that there were some large transactions for bottle service in clubs.  We thought we could be an extension of the promoter system and generate income there as well.

We had a great target picked out as well.  We thought we’d go after the young person that frequents bars and clubs.   They would be the heavy users that would deliver lots of activity to the site.  All we would have to do was convert that action into purchase activity and we thought we’d be fine.

Our big issue was the disconnect between our target and their purchase behavior.  The person that’s out every night falls into one of two camps.  Either they are a true regular on the “scene” and they already have plenty of people selling them bottle service and tickets, or they are looking for the best deal and don’t spend a lot at all.  The second group wants to hang out with the first, but they don’t buy the product in the same way.   One could argue that the second group doesn’t buy at all, as they’re usually getting in for free and grabbing as many first drinks as possible.

The spenders had an alternative, high touch channel, and that left us with the far less lucrative group.  We could have had 100% market penetration of that group and we would have still found it very hard to make money.  Not that we understood that at first.  All we could see was that we were growing our traffic and activity, but hardly selling anything.  It took plenty of interactions with the clubs to understand that we were pretty unlikely to have a lucrative path.

That left only advertising compensation.  We weren’t clever enough to create groupon, but even if we did we would have not had the sales force to execute there.   Low intent traffic isn’t particulary lucrative.  Facebook can make some money because of it’s epic size, but since Stir was niche site it didn’t have much of a chance.

I’d like to think we could have made the pivot to make money off our audience over time, but we really didn’t have that luxury.  It’s still one of the biggest mistakes that we made at Stir.  We didn’t have a real business model.  There’s a few startups (Twitter and Facebook) that grow huge enough to buypass that epic gaffe, but most have to figure it out far earlier.

In the end Stir executed about 80% right, but with a 20% business model.  There are plenty of companies that do pretty well with an 80% business model and generally mediocre execution.

Tags: , , , , ,

Steve Hammer on January 19th, 2011

My wife called me out one day on my desire to always be the hero.  It’s a very strong word, and it was clear that it was pointed out as a character flaw more than an admirable feature.  Rather unsuprisngly, she’s quite right.

If a friend has a problem, I’ll do what it takes to help them out.   I’ve spent hours helping friends out with their SEO or market positioning and never asked for a single thing in return.  When I was building stir, it was all about the noble product we were building more than the personal return that was possible.  It’s not uncommon for me to be up until 2 in the morning working to fix a site problem or develop the right bid.

Maybe that explained the problem with our business model on Stir, but that’s an aside.

One of my former bosses once criticized my negotiating skills.  The strange part for me was that I got deals done, and they were generally successful.  However, to him, I wasn’t squeezing the the last dime out of my potential partners.

It’s still my goal to build a sustainable business.  There’s no doubt that I could make a quick buck here or there and build something for the short term only.  There’s just no joy in that for me.

My view on the world is simple: Deliver value.  Two simple words that mean a lot to me.  To deliver value you have to understand what people truly consider valuable, and find a way to produce it.  It’s about solving problems.  It might be incorrect to assume that if  a truly valuable problem is solved that rewards will be there.  Even if it’s wrong I’d rather be remembered for what I created and gave than what I took.

I don’t think that makes me a hero.  I think it just makes me human.

Tags:

Steve Hammer on January 12th, 2011

At current, I have what most people would call an unusual schedule.  I leave my office around 2 every Monday and Wednesday to pick up my daughter.  From there she spends some time at home while I work.  It has to be one of the best parts about search engine marketing for work-life balance; as it can be done from virtually anywhere.

With access to something like adwords editor, there is even the ability to work offline and upload changes when connectivity returns.  I can literally work anywhere, and at any time.  For a funky schedule it doesn’t get much better.

Web analytics are also in the same boat.  Some of the top platforms already have iPhone apps that can allow me to check site traffic virtually anywhere.  Almost all of the information is right there at anytime.

I think there are only a few reasons that I travel in to the office anymore.  First, I actually like the interaction with other people.  It gives me ideas to push things further, and it helps me to express my efforts.  And second, there are some hardcore analysis that requires interaction with a data warehouse.  Once I cross the line from online only activity and need to know about offline behavior is where a more dedicated resource is required.

Search Engine marketing wasn’t something I chose for the balance, but it really is an advantage for the job.

Tags: , , , , ,

Steve Hammer on December 1st, 2010

A bit ago I became aware of project honey pot, a clever way to dangle traps for malicious spammers.  These traps allow them to gather data on some of the worst violators of email address harvesting and other nasty techniques.   Personally I think these guys are doing a great thing to reduce the clutter of the inbox, and in some cases the clutter of comments.  They do offer feeds of their findings which can be used to help filter these bad actors from internet traffic.

My question is: Why hasn’t google or bing created a similar idea to combat webspam?  They could actually buy out some of the off topic directories, link farms and other so called bad neighborhoods and convert them over to harvesters for link brokers.  If they didn’t want to buy them they could certainly create a few thousand and do the same thing.  If they know how to identify a bad neighborhood they could make them pretty easily.

It’s kind of like doing undercover operations.   Some times you have to hang out in the bad part of town to figure out who’s behind breaking the rules.  It’s even more useful for the search engines because they have made the rules and also have to enforce them.  They are the only ones that know exactly what they consider over the line.

I welcome comments here.  Non-spam ones of course.  Would this work?  Would they get sued?  Is there an even better way to run the sting?

Tags: , , , , ,

Steve Hammer on November 24th, 2010

There’s a considerable amount of time that’s spent for most search engine marketers in managing bids.  As a result there are some fantastic tools out there automate and assist with the process.  They can use very sophisticated algorithms to determine an optimum bid and take in to account time of day, location and tons of other factors in the process.

Usually these managers optimized based upon the value of a single transaction, and this to me is the real failing.  If your site is doing it’s job well, repeat visitation can be a very significant drive of revenue.   To bid properly, these repeat visitations need to be taken into account.  This creates a customer lifetime value that can be far in excess of the value of any single transaction.

A simple hack is to inflate the value of a transaction to model some of the repeat.  It’s common to see this in affiliate marketing for dating sites.  Some sites will pay 100% of the value of the initial signup, because they have determined that a large percentage of their users will have a second or third renewal.  It’s a hack, because some percentage of those users won’t be new users, so in essence they’ve been acquired twice.

It takes an actual model to get it right.  The model needs to take into account not only the simple number of times that a user repeats, but the percentage of repeat users in that channel, changes in transaction size for repeat users and countless other factors unique to every business.  The model can be quite complex, but is essential to have good bidding.

The most significant factor can be repeat visitation and transactions through non-brand search terms.  One would expect that combination to be a very high percentage of new customers, but that isn’t always the case.  If this is off, it’s easy to significantly overbid these terms.

It’s also not enough just to run the model once and then forget about it.  There’s real cost savings opportunity that comes from having a good social or email program that can reacquire customers rather than have them reenter through more expensive channels.  As those retention factors start to change, the model needs to reflect it as well.

Customer lifetime value modeling isn’t easy, but it’s what it takes to do industrial grade search engine (and internet) marketing.

In summary:  Transaction bidding will frequently underbid for traffic and limit growth.  Simple repeat models will overbid for traffic and reduce margin.  It takes a real customer lifetime value model to get it right.

I’ll show a sample model in a future post.

Tags: , , , , ,

Switch to our mobile site