Nathan Zaru

I searched Google for the keyword "search". Interesting result... apparently they rank themselves as the 4th best result. 

This post originally appeared on my company blog
    
Competition thrives in the meritocratic world of tech.  No incumbent is safe, it seems, from a worthy newcomer.  In social media we see talk all the time of the new "Facebook killer" and the endless debates over Facebook vs. Twitter and (depending on whom you ask) Facebook vs. Google +.  Recently Pinterest, Instagram, and many others have added to the equation.  But there will not be just one winner of social media.  Rather, there will be a multitude of winners serving different functions each with overlapping networks.

For social network A to win against social network B, on aggregate users would have to prefer A to B.  But what if A and B do totally different things, does it make sense or is it even possible to declare a winner? It was fair to compare Myspace to Friendster, Friendster to Facebook, and Facebook to Myspace because they provided roughly similar features and functionality.  Check out Wikipedia's list of social networks.  It's long, and looks incomplete to me.  More importantly, very few (if any at all) are trying to do what Facebook does exactly.  Most are either trying to serve a unique function or provide a singular function Facebook already has but much much better.

The companies that will win are the ones that understand the future secret to success in social media: the network size needed to support a platform is in direct proportion to the number of features and functions it provides.  This means you don't have to have 800 million users anymore to be a dominant social platform.  StumbleUpon and Instagram are not winning because of their massive user base.  They are winning only because they each do exactly 1 or 2 things and they do them much better than anyone else.  In fact, you can bet Stumble Upon and Instagram users are on Facebook and use it propagate their activities from other social platforms.  So instead of competition as Facebook vs. Instagram or Facebook vs. Stumble Upon, it's a symbiotic relationship where both companies benefit from the cross pollination of activities between platforms.  

The history of internet economics also tells us that there will be multiple winners in social media.  Six or seven years ago it would have been tempting to declare Amazon the winner of the eCommerce sector.  Likewise for Craiglist and the online classifieds industry.  But Amazon didn't kill eCommerce, it became the preeminent standard.  Craigslist didn't win online classifieds, it proved the model and the industry.  Facebook has paved the way for social media.

So who will dominate social networking? More companies than you might think.  But rather than providing similar functions, the companies that win will define new ones and hardwire mutually beneficial relationships with other social platforms into their technology to maximize their exposure.  It is indeed a good time to be in social media.

This post originally appeared on my company blog
    
In my post on Facebook advertising strategies I showed you the important Cost Per Acquisition (CPA) equation.

This calculation is absolutely vital for evaluating the success or failure of any campaign you use to acquire people, customers, subscribers, visitors, etc.  But what if you want to predict the value of a campaign before it even starts so you can weigh your options against alternative strategies?  Unless you are paying on a CPA basis (i.e. you know exactly how much an acquisition will cost) this equation does not work because it is only useful after you measure the acquisitions you gained from a given campaign.

It's hard to guess how future initiatives will perform.  The good news is you can use historical conversion rates you already have (CR) and the cost per click you are willing to spend (CPC) to figure out approximately how much an acquisition will cost.  It turns out that a relationship between CPC and CR is equivalent to expected Cost Per Acquisition.  

I'll save you the proof, just be sure to use dollars for CPC and a decimal for CR.  This CPA calculation is exact if you are looking at the cost you did pay for a visitor and the conversion rate you did observe.  But this equation is more powerful when used as an approximation of CPA for future campaigns.  By just looking at the cost you are willing to pay for a visitor and the conversion rate you might expect, you can instantly figure out how much a customer will cost.  Then you can compare how much that customer will cost to the expected value of a customer from that channel to determine if that campaign is worth pursuing.  To take it to the next level, you can compare expected CPA versus expected CPAs from alternative channels to determine your best source for customers.  This works for any resource  -- social networking, search advertising, affiliate marketing, even SEO -- because each initiative requires some sort of monetary output and results can be measured by new inbound clicks generated from that channel.

I've long been a fan of Isocket.com because I really like how they are transforming the online advertising industry.  Earlier this year their Customer BFF Ryan Hupfer asked me to spill some of my beans and talk about how to run effective advertising campaigns.  I agreed and was happy to get this down on paper, it has served me well so I hope other marketers out there might find it useful.  

To see the full text, check out Get That Click: Strategies For Successful Display Advertising

This post originally appeared on my company blog
        
Marketing channels are instruments, and you are the musician. The master musician plays both in time on in tune, so your multi channel marketing message will have the greatest effect when it is synchronized and congruent.

Synchronized. Are you strategically planning your marketing campaigns to coincide across different mediums, or are they just haphazard points in space? Stack them together.

Congruent. Are you communicating the same message across different mediums, or is your voice divided? Match your messaging.

A concert is the united performance of musicians within an organization.  What does your concert sound like?

I first made a Facebook ad in April 2007 when I managed my colleague's campaign for university office.  The Facebook advertising platform was brand new and our opponent did not know how to respond.  It turns out our victory that year stands as the only loss our rival political party took in the past decade.

Why Facebook Advertising?
Facebook advertising has changed substantially since then, but one fact remains:  it is the most efficient display platform to target specific audiences on the web.  Marketers are taking note as Facebook advertising is bigger now than ever.  Money is going away from Adwords/AdSense and going toward Facebook. This is a huge shift.  The problem though is the Facebook platform does not have adequate tools to administer and optimize your campaigns. I  have also not seen enough literature covering the bigger questions of Facebook marketing:  why are you advertising, who are you trying to reach, what do you want them to do?  This is what I want to address.  I will assume you have at least a basic fluency with the Facebook marketing platform.  If you need to review the basics just check out any of the mainstream tech blogs.  I will be limiting the scope of this article to ads that take the user away from Facebook to an external URL, be it your company website or some other web property.  Whether you are focusing on purchases, visits, registrations, or page views, acquisition marketing is the name of the game.  

Power of the Platform
The power of Facebook advertising is the versatility of it's targeting capabilities.  Want to advertise to male dog lovers in Canada?  Done.  Interested in female soccer players in Japan?  No problem.  Looking  for 22 year old college students in Germany?  You can get to them with just a few clicks.  These segmentation capabilities are unprecedented.  If you are searching for a core audience, or if you know who they are but are looking for more, you should consider Facebook advertising.  The key to Facebook advertising that is really cheap and really effective is a comprehensive and scalable tracking, testing, and optimization strategy.  

Track Your Activity
You cannot rely on Facebook to do this for you so let me save you a lot of trouble:  track everything in an external campaign log.  I use a spreadsheet but feel free to use something more fancy.  Each ad has many more data points than you might think.  You need to track campaign name, ad name (headline), ad copy, image, interest segment, demographics, bid price, go live date, kill date, and agent name (advertising manager) for every single ad you make and every subsequent update.  This can get very very complicated over time.  Facebook analytics is setup such that it makes the most sense to make conclusions on the campaign level.  So to make it easy and more accurate for yourself, fix the interest segment to the demographic you are targeting.  Make this unique to each campaign.  What this means in practice is that each campaign should be targeted to a unique audience.  Any variations on an interest segment or demographic require their own new campaign.  If you vary the interest segment and demographic targeting on the ad level you will kill your time with unnecessary overcomplexity and will go bald from tearing your hair out.  Track everything in your campaign log.

Test Different Ads
You have to test everything -- ad headline, copy, image, interest segment, demographics, and bid price.  This is principally because your preconceptions on what might work best probably will not.  You are a scientist, Facebook is your lab, and advertising campaigns are your experiments.  Sounds complicated but the concept is simple:  hold all variables but one constant and test relentlessly.  For reference you check out my A/B testing strategies or brush up on the scientific method.  The more targeted audiences and ad versions you test, the more surprises you will get.  For a given campaign (i.e. fixed interest segment, demographic, and bid) you can test different ad headlines, ad copy, and ad images.  Be sure to only test one variable at a time.  The results will tell you what advertisement design will work for a given target audience, and this is the valuable data you are looking for.  Repeat this for as many relevant  targeted audiences (i.e. campaigns) as you can come up with.  Bidding strategy is easy.  First, never use CPM, go for CPC.  You are focusing on acquisition so you need to know the precise price of a click.  The recommended price of a bid is a function of targeted audience you are looking for, and thus should vary on the campaign level if you follow my instructions from above, not on the ad level.  But the secret of Facebook advertising is that it is a marketplace just like any other so prices fluctuate according to advertiser demand.  Thus, the "Suggested Bid" does not mean much.  Start low, below the suggested big, and increase incrementally until you start getting clicks.  If you loose clicks keep increasing your bid, and stop once you get demising returns or it gets too expensive relative to your CPA (more on that to follow).  Re evaluate your bid periodically to make sure the market price did not reset so you are not overpaying.  Track changes in your bid on your campaign log as they happen. 

Optimize For Conversions
The goal of Facebook advertising should be to optimize your ads for cost over time so that you can acquire customers as efficiently as possible.  But the best metric Facebook has for you is Click Through Rate.  While this does tell you how well your ad performed relevant to the targeted audience, it says nothing about the performance of the visitor once they get to your website.  So you have to do more.  First and foremost you need to use tracking codes in the click through URL in every ad.  You can review my social media tracking strategies, but ultimately you need to have your goal action on your website tied to your web analytics system so you track how many conversions each ad brings over time.  You also need to think about where you land your visitors coming from Facebook.  Do you send them to the home page, a deeper index page, or a page specially built for incoming Facebook traffic?  In either situation, you need to have whatever you promised the visitor in the ad they clicked available and featured prominently to minimize bounces.  You can also try a super advanced technique:  generate dynamic landing pages according to the tracking code the visitor is carrying.  This will be a technical challenge, but generally speaking increased customization yields increased goal conversion rates.  This all leads to the most fundamental metric, the cost per acquisition of a customer.  This is how you calculate it

This is the ultimate metric you need to use to evaluate the effectiveness of a campaign.  Comparing apples to apples, you can properly conclude which campaigns are working and which are not.  This successful campaign data will tell you who your targeted audience is and, by extension, the demographics and interest groups you are able to efficiently acquire.  These are your customers.  

You should follow me on Twitter for more

I used to be square.  But college and the job markets told me I had to be round.  For a while I was trying to be round.  Now I know it's okay to be the square that I am.