Hey, we're moving...

Hey All,

Just a note to let you know we are moving our blog to http://plusroi.com/blog/.

Blogger has been a fantastic tool and I'll keep using it for other projects, but this particular blog is going to be better served in the new WordPress blog which is hosted on our servers and integrated a bit better with our site.

We are also just on the verge of launching our Online Marketing Insights newsletter, which means by default I'll have to get busier writing posts. Have a look and sign up if you're interested.

Cheers,
Rob

Online Advertising – Google Gets Intimate with Local Ads with Adwords Geo-Targeting

Something happened in December that should have gotten headlines in every small business paper, magazine and website in the world. Interestingly, instead, unless you‘ve read the Google blog or done a search on “local targeting” you’re probably unaware of what happened.

What happened is that Google added Google Maps functionality to its Adwords program.

Got a decent website for your electrical supply company in Seattle and don’t want to waste money paying for clicks from outside of Washington State? Simply select the Locations feature in the Target Audience options in your Adword campaign settings, select Washington state on the map and you’re done.

Ooops, discovered you need to add Portland because you’re getting lots of sales from there? Just go back in, select Portland and add it to your target.

Your own targeting criteria will be as unique as your business, but the results are the same: less overall cost for your Adwords program, much more qualified traffic to your site, and much higher ROI.

With only a tiny percentage of businesses taking advantage of these opportunities there is no good reason for you not to dive right in.

Worried that it’s too complicated? Well, it is a bit intimidating at first, but with an investment of a couple hours of keyword/competitive research and hour to set up your Google Adwords account you’ll be rocking.

Indispensable Google Metrics and Optimization Tools

In recent weeks we've been gearing up to relaunch our website for the next "iteration" of our business. This involves doing things like registering domains for "client-branded" A/B testing, getting professional help to redesign our site, and implementing the tools we need to optimize our site.

Much like the "shoemaker's children," we've been going without a number of the things I tell clients they absolutely must have in place for effective website optimization. Well, we're finally changing that, but as a young company we've still got to consider our budget (or lack thereof) for this effort.

This is where Google comes in. There are a couple of free Google tools that are mind-blowingly powerful if you leverage them well. Without further ado:

Google Analytics
While lacking the "Enterprise" power of major analytics providers (Omniture, WebSideStory, etc.), Google Analytics is incredibly robust for a free tool. Plus, it's easy to install.

Even non-techy folk (like myself) can manage to implement it by themselves across their site and blog if they take the time to read through the instructions. I had it up and running correctly on our 50 page site and blog within 30 minutes.

Our implementation even included setting up specific conversion goals that we wanted to measure.

Interestingly, for some companies, Google Analytics is a better choice than some of the more robust offerings. The reason for this is that you'll never get past the surface of what the Enterprise solutions offer unless you spend a LOT of time at it (trust me on this, I've spent days sifting through data in Omniture and WebSide Story).

Put it this way: unless you have a huge site or a large multi-channel effort (i.e. involving email conversion tracking and CRM), you're likely just fine or possibly better off with Google Analytics.

Google Website Optimizer
Google Website Optimizer is a tool you'll find within Adwords. It's primary purpose is to quickly and efficiently conduct A/B testing on landing pages for clicks to your site from Adwords.

Put simply, you can create two dramatically different versions of the page that people land on when they click on an ad you have placed in Google Search or on a content site. Then, you track the results until you can determine a "winner" of the test.

For smaller companies, this saves an awful lot of hassle in setting up a script to randomly display the different pages in the test and having the web team manage the process. If you can edit your web page and put it online, you have all the skills you need to do it yourself if you're willing to read the instructions.

Now, the beauty of this is that Google Website Optimizer works even if you're not paying to run Adwords. The folks at Google are quite generous in these regards. All you need to do is sign up for an Adwords account (free) and you've got access to it.

Also, using the Website Optimizer is not limited to Adwords. For instance, assume you have two strong opinions on what headline should be on a product page. This usually involves an argument and possibly (50% likelihood) the wrong decision.

With Website Optimizer, you can quickly create a second version of the page (with the different headline). Simply drop a snippet of code on both pages, set your conversion goal in Analytics, and you're off to the races. You can now run an A/B test on the page.

So, what are you waiting for?

The Case for Marketing Optimization vs Spending on Reach

For a while in my career I worked in a company where traffic was considered the main yardstick for success and growth. This became incredibly frustrating at times as, while there was always some more budget to grow traffic, getting budget for marketing optimization and conversion improvement was next to impossible.

While traffic is indeed a critical factor, it's the wrong one to focus on if you haven't invested in optimizing conversion on your web site.

Like George Kastanza figuring out a witty retort hours too late to use it, here I am some years later to present some pretty compelling math on why companies should focus first on conversion.

First, a couple assumptions:

• - The company in this example is selling a product (but the same principles work for ad revenues, B2B lead generation, etc.)
• - The company has not undergone extensive marketing optimization on their web site and/or promotions already.
• - There exists an opportunity to raise overall conversion rates by 5% or more (I have yet to work with a company where this is not possible).

In the first example below we have an example of a CPM banner ad where conversion is the only option as the ad is currently running at negative ROI (See previous post on Working the Math on Online Advertising).

As you can see, with the assumptions I put in, the ad campaign initial is returning only $.80 on each dollar spent on it. Despite the thought of "making it up in volume," spending more on reach on this campaign will only drive it further and further into the red.

On the other hand, if the company takes the time to run some insightful a/b tests on the ad and the landing page, all of a sudden the campaign is making money (ads and landing pages present significant opportunities for conversion improvement). Best of all, the campaign is now scalable. In this case, the logical thing to do is to NOW increase the reach of the campaign until it hits a point of diminishing return.

In the second example, let's have a look at web site conversion and the case for optimization versus spending on reach.

In this case we compare a 10% traffic improvement against a 5% conversion rate improvement (from 2% conversion to 2.1% conversion). "But Rob," you might say, "we make more extra revenue by paying for traffic and we get 10% more people exposed to our brand."

That argument would be true for the first month, but to truly understand the power of optimization, let's look at the math over a year:

• - Buying 10% additional reach costs $50,000 per month, which equals $600,000 per year, for a 160% overall ROI in the example shown.
  
  
• - Paying a consultant or agency $10,000 a month for 4 months is a good guess for what it might take to get 5% better conversion on a site with these kind of revenues (hey, I never said this blog wouldn't be self-serving sometimes). This assumes the site hasn't already been optimized.
  
  In this case, let's assume the 5% improvement kicks in 2 months into the process. What this math works out to is that a mere $40,000 spent on conversion will result in $400,000 in new revenues for a 10x or 1000% ROI.

That's pretty compelling math, isn't it? Better yet, you don't have to choose one over the other. By doing the work of improving conversion FIRST, then adding the spend on reach, you now get an example that looks like this:
Wow. Keep in mind that no real life scenario is as simple as the case I've described. That said, these are incredibly powerful principles and I hope I've helped show you why it makes sense to focus on conversion before opening the wallet for reach.

BTW, despite my tongue in cheek reference to hiring a consultant or agency, conversion and marketing optimization is definitely an area where you should develop in-house skills if at all possible.

I Love the "New" SEO

Once upon a time, SEO was simple: put a lot of good, relevant content on your site and you'd be rewarded with qualified traffic.

Then, as seems to happen with all online mediums (email, blogging, forums), trickery and "secret tactics" became the way to climb the search engine rankings. The big problem with this is that instead of being rewarded for adding value to the Internet, many SEO practitioners were being rewarded for polluting the Internet with spam sites, ugly gateway pages, keyword spam, etc.

That was about the time I stopped actively participating in SEO. I'm a "content guy" at heart and believe the long term path to success is to add value to the medium you are working in. I had no interest in learning how to "beat the algorithms" and I put my energies elsewhere (community development, word of mouth, email newsletters).

Well, what's old is new again and there's a bit of a renaissance going on right now in SEO. The algorithms have caught up to the SEO spammers and successful SEO now boils down to exactly 3 things:

1. Lots of good content, regularly updated (a great reason to blog about your business).
   
2. Inbound links to your content from high quality sites (which happen when you write great articles or posts and actively promote them to other sites).
   
3. Proper site structure, tagging, and coding (another reason to hate Flash-based sites).

That's it. Do a good job on those 3 things and your rise in the search engines is pretty much guaranteed.

Did I mention I'm back on the SEO bandwagon in a BIG way?

P.S. It wasn't until I quickly proofed this article that I saw the combination of today's date and my proclamation of "love" in the title. I guess you can consider this my Valentine's message to the folks behind Google's algorithms.

Design Considerations for HTML Email

I was asked by a potential client today if there were any good online resources to learn about special design considerations when you're creating HTML email newsletters.

After a quick look in Google I couldn't find anything that hit all the points I thought he should know about, so I ended up writing something up for him quickly myself. Not wanting to keep this information to myself now that I've taken the time to write it out, I figured I'd publish it in this blog.

There's not a lot of explanation with these points, but if you've been digging into html email/newsletter best practices, it should all make sense to you.

Here ya go...

• 600 pixels is the generally the recommended width for html newsletters, allowing for the full width to be displayed even in clients like Outlook where the software takes up considerable real estate on the left of the screen and previews the email on the right.
  
  
• It's critical to practice “vertical economy” for all elements of html newsletters. For instance, if the masthead or top banner is too tall, the headlines can very easily get pushed down “below the fold.”
  


• When creating comps for newsletter designs, it’s critical to bear in mind that html newsletters must be coded with old-school tables (not CSS). Text can be formatted with CSS if desired, but text encoded in this manner will not render properly in web clients like Gmail (Gmail strips out ALL CSS).
  
  
• Images must be used carefully if emails are to be remotely readable prior to images being displayed in the email client. While well-branded companies like Apple can arguably get away with having image-based emails, it’s generally considered a bad idea.
  


• My recommendation is to build the email in a manner where headlines and titles are readable without images being displayed. Alt tags, of course, also become quite important.
  


• My personal preference for a nice, readable layout is a 1:3 two column layout with the narrow column on the left side (150 pixel left column, 450 pixel main column). Set up like this, the left column is very effective for call-outs, table of contents, appropriate ads and promotion.
  


• Generally, the most practical approach to ongoing newsletter production is to have a professional designer mock it up, get a competent coder to do the initial html templates, and then have the editorial or production staff place the articles in each issue (possibly with a little help on the graphics).
  


• Also, in general, the simpler the better for layout. While the temptation is always there to build it out to the nth degree, a simple layout that can be quickly and easily edited for each issue makes ongoing deployment WAAAY more quick and practical. Also, while the layout needs to meet branding guidelines and represent the company well, a straightforward, consistent layout is also in the best interests of the reader.
  
  
• Expect to engage in a significant amount of testing to ensure the initial template renders properly across a wide range of email clients (including gmail, aol, yahoo, hotmail, .mac, mac mail, outlook 2007, windows mail, etc). Its kind of painful, but this time spent up front will save a ton of time troubleshooting down the road.
  
  
• Many of the email service providers provide wysiwyg templates, but depending on the complexity of your layout and the proficiency of those administering it, Dreamweaver can be a more practical option (my preference).
  


• Generally, I lean towards having shorter articles in the newsletters. If there is a longer one it is best to go as the last article. If there tend to be a number of longer articles, it’s usually best to provide short excerpts combined with links to the complete article on the site (which should be part of your SEO strategy as well).

Happy coding!

Working the Math of Online Advertising

Ever struggled to figure out whether it makes sense to make a specific online ad buy? There's really only one way to figure it out, and that's to run the math on it.

Expect your "best guesses" to be way out of whack the first time you try this excercise. That said, doing an estimate and then a report after the campaign is the only way to get the insight and learning required to make better and better ad buying decisions.

More importantly, if you have ad buys that you do on a recurring basis, you MUST do this kind of analysis to figure out if they are making you money or not.

So, in a nutshell, here are the rows you will put in your excel spreadsheet:

1. CPM (the amount you pay per thousand for your banner ads).
2. Cost per impression (1/1000 of the above - somewhat redundant but it helps with our formulas and to illustrate a couple of points).
3. Click % (the percentage of people who click on the ad - this ranges from almost nothing to a couple percent for a highly targeted vertical ad).
4. Cost per click (cost per impression divided by click %) *Note: for Google Adwords, where you are paying based on cost per click, this is where your spreadsheet starts.
5. Landing page conversion % (this is the percentage of people who arrive at the landing page from the ad and take your desired action).
6. Cost per conversion (Cost per click divided by landing page conversion %).
   
7. Value of conversion (this is your net profit if you are selling something or an assigned value if you are undertaking lead generation or other activities).
8. ROI (Value per conversion divided by cost of conversion)

Once you arrive at ROI, there are a few factors to consider. For instance, if there is branding and/or residual value in your campaign, you will want to factor this in. Don't expect to achieve positive ROI in all your campaigns, but be prepared to aggressively test and optimize the ones that are close to positive ROI.

Once you successfully found an activity that generates clear ROI, you have uncovered a beautiful opportunity and need to explore how you can duplicate and/or scale up your efforts. This is the key customer acquisition strategy of most of the biggest companies on the Internet: Find a few ROI-positive acquisition methods and scale them up like crazy.

There are a million variable factors in this model, but keep it simple to start and then expand on it as you get comfortable. For instance, you may want to assign the "value" of the action as your average LCV (lifetime customer value) instead of the value of a one time sale.

The beauty of using the model is that as time passes and you get real results to enter in the spreadsheet, you'll get more and more accurate in your projections. Some of my clients keep a tracking document that stores both their original projections as well as the post-campaign actual results for each campaign they run. While this requires that you put your ego in check (a lot of the projections will be very, very wrong), it eventually becomes possibly the most important and accurate decision making tool in your business.

So, while it will take you a few minutes to set up the first time, this exercise is very straightforward and becomes incredibly valuable if you do it for each campaign you run. Also, as you can see, it is an excellent tool for and "apples to apples" comparison of different types of advertising (AdWords vs CPM for instance).

If you get really creative, you can even find ways of incorporating contests, trade shows and sponsorships into this model of ROI comparison (this gets tricky but is definitely valuable as you get comfortable with this kind of model).

BTW I actually thought this was going to be a quick and easy post before I started, but as I got further in it became clear that this is an area that will require some more digging.

More to come (including an example)...