Being able to move from your business objectives to appropriate web metrics is critical to ensure that your business meets its web objectives. Business objectives and web metrics need to work hand in hand since if you don’t have clear business objectives you will have trouble measuring them through your analytics package and if you do not have clearly defined web metrics that are appropriate to your business objectives then you will be reporting data that does not give a true picture of how the site is performing in relation to your business objectives.

Good objectives should be SMART. In other words they should be Specific, Measureable, Actionable, Relevant, and Time-based.

So how do you move from business objectives to web metrics? First you need to determine which business objectives your site actually supports. You will come away from this process with your website objectives which will be your basis for what you should be measuring on the site. It is important to note that there can be several website objectives that support a single business objective or one website objective that is used for multiple business objectives. The next step is figuring out what metrics provide insight into if your website objectives are being met.

In some cases the metrics may be very clear, but in others where the objective is not specific it may require really thinking about what demonstrates achievement of the objective. In these cases you must determine what event or events on the site constitute success of the objective. When determining the success events, make sure that every one of them are measureable.

Now that you understand the events that lead to your site objectives it is time to determine the specific metrics that will be used. For each event there are usually multiple metrics that can be used to track success or failure. You need to determine which ones are more relevant and should be used for reporting purposes. For example if a success event is an email signup, the metric could include the number of new signups per week, the percent of unique visits who sign up per week, or the percent of unique visitors who viewed the newsletter on the site and then signed up. In most cases it is a good idea to use multiple success events as the metric.

You should then define a monetary value to each success event or metric. It is not always easy to do, but it provides great insight into how each is performing in relation to the overall site objectives.

Lastly, there should be targets established so that the analyst knows what they are trying to accomplish. In addition the analyst needs to be empowered with the resources required to meet those objectives. The last step is putting all the data for the various site objectives into easy to read reports. For more check out my recent post on web analytics reporting that’s easy to understand.

One of the challenges of online advertising is getting your message out to the right crown. Google, Yahoo and MSN have geotargeting and dayparting solutions to show your ads to the right locations (based on IP address), and at the right times of day, but of the 3 only MSN has a basic demographic targeting tool. Facebook realized that they had a great amount of user entered data that they could make use of for ad targeting, so they set up their demographic targeting.

Let’s imagine that we have a nonprofit that wants to target young women that have an interest in politics, and work in the DC beltway. The first step is to head over to the ad generation site.

Not bad, 18 million people that we can target to. But let’s change those parameters so that we’re just reaching those who have identified themselves as women.

8.6 million is still a great number to target, but we wanted to go for young women, let’s try narrowing it down to just 18-25 years of age.

Facebook has an option where you can specify your political leanings, so let’s use those options to only show young women that have identified their political views.

2.7 million. How about just those that live in the beltway states – Virginia, Maryland and the District of Columbia?

Still over 174,000. This may be where we’d want to place our ad, and see what response we get from it. Of course, we can go more and more granular in ways that are just not possible on the regular paid search sites. How about if we specify that they must work in some capacity for Congress?

520 young women, who state their political preference, live in the beltway area and work for Congress. That’s absolutely amazing targeting. To be able to go down to virtually any level of granularity, based on education, favorite books, etc, etc, etc. gives you so many options when targeting your ads. If you understand your customers, and can segment them appropriately with the correct messaging, this should allow you to increase your conversion rates and improve your ROI, which should be everyone’s goal.

Web analytic technology has been around since the early 90’s. Over time the technology has become more sophisticated moving from focusing metrics like ‘hit’ and ‘page view’ to understanding the difference between a visit from a spider and a ‘human’ (well, their computer). And now taking it even further the ability to segment traffic and integrate with other systems (import campaign data or other CRM information). The amount of data available is almost endless.

The most common approach to analytics in a company was to assign a person, maybe a marketing person, maybe an IT person to be in charge of the analytics. Typically this person would figure out how to run a report that contained the relevant, generic data like visits, unique visits, hits (useless) and repeat visits. Then they would schedule this report to be sent to everyone each month or maybe even each week.

And each month or maybe even each week lots of people would receive this information maybe glance at it or maybe not and delete it. This is really not a valuable use of web analytics.

Web analytics can give you such important information. It can tell you if you are wasting your money on PPC, irritating your visitors with your site design or doing a great job of converting new visitors to customers. But you have to know what you are looking for.

Looking at pages of meaningless stats each month or week is a waste of time. Getting into the data and trying to find useful information is what you need to be doing.

Web analytics is complicated. Many marketers are not analytic by nature. Many IT folks who have been tasked with being the web analytic owners are not marketers by nature. So what is there to do about this? Unfortunately for some, there is no easy answer. There are two options. You can outsource all of your analytics or you can take the time to learn and interpret the data. This is where it becomes a discipline.

More on this next week.

Which of the following environmental nonprofit videos do you like?

This one from the National Fish and Wildlife Fund?

Or this one from the Chesapeake Bay Foundation?



The first video is… boring. the imagery is great, but with the style of the voice over, it comes off like a corporate training video, it just doesn’t work. The second video does, it’s fun, and it has personality, lots of personality. Is this just my opinion? Look at the count of views for each; the NFWF video has had 43 views in the almost 2 months it’s been live, whereas the other video has had over 1,400 views in the same time period. 43 viewers in 2 months is poor, especially for a professionally produced video. The Chesapeake Bay Foundation video was most likely pushed out to their members, through both YouTube channel subscriptions and other marketing efforts, whereas the NFWF video doesn’t appear to have been marketed, or if it was, the conversion rate is not good.

A couple of other things to note about the NFWF video:

• at 6 minutes, it’s too long to keep viewer’s attention, especially with the voice.
• they don’t provide their URL in the video, so they will probably lose some potential conversions

I think the lesson for nonprofits, and for anyone else, is that you can’t just take your efforts from other channels and assume that they’ll work effectively in all channels. You need to look at what works in each of these channels and tailor your message accordingly. Still, not everything will work, just keep plugging away and refining your message, and eventually it will.

On my personal blog, I received an email through a contact form, which until recently, had a “captcha” image that I could never read. Now that this security feature has been removed, I received a contact message from someone looking to buy a text link advertisement on my site. Seeing as I do not have any other advertising on the site, I thought the request was a little peculiar, but not out of the realm of possibility. My contact form also sends an automated, yet personal message to anyone who completes the form, and I received a friendly response email that made the person seem legit.

But when I went to search on the person’s name, I came up with several search results that indicate the person is a spammer, in fact, not even real. Blog authors argue that she is most likely a smart robot submitting spam comments both through posts and through contact forms, and one commenter even indicated that their email address had been collected from an image, not from a text scraper

Could this email be from a legitimate person who now has a reputation management problem, or a spam robot? I am reserving judgment for now.

This is a public service announcement to check out even legitimate looking emails by searching on the sender’s name for relevant information before responding.

Google has once again topped the list of best companies to work for according to Fortune Magazine.  It is the second year in a row that Google has landed on top after not appearing on the list in 2006.  Chief rivals Microsoft and Yahoo have both slipped in the rankings the last couple years as can be seen in the table below.

Of the companies to make the list, Google had the second largest job growth as a percentage of total employees at 60% although Microsoft added more new jobs 3,409 to Google’s 3,039.  While it may not look like Yahoo is all that appealing, the average salary and average bonus of Yahoo employees was higher then both Google and Microsoft.  The average salary was $116,250 and the average bonus was $13,671 at Yahoo while neither Google nor Microsoft made the list.  Those employees probably do not have the same level of job security, however, as it has been widely reported that there are job cuts coming at Yahoo.   

Each company makes the list with some of the perks they offer.  Microsoft pays 100% of their employees’ health care premiums, while Yahoo is in the top 10 for telecommuting with 20% of their employees being “regular” telecommuters.  Google has some of the more unusual perks such as $1,000 towards a hybrid or electric car as well as a discount program for employees in California who install solar panels on their homes.     

To see them all check out the full list of 100 best companies to work for.   

I just finished reading a great book by Daniel J. Solove titled: “The Future of Reputation: Gossip, Rumor, and Privacy on the Internet”. In the book he talks about Reputation, Privacy and the legal issues surrounding both. In there he has examples of real life law suits, where people have been photographed or videotaped doing something in public, and it’s been posted on the web and gone viral, in many cases causing problems for the people involved.I started thinking about this again today when I read the story about a Chinese couple that kissed on a train platform last September, only to find out that last week someone had taken the CCTV footage from the platform and uploaded it to YouTube and other video sharing sites. Should that couple have the expectation of privacy? Sure, they were kissing in public, but at the most they only intended anyone else waiting for a train to notice (should there have been anyone else there). What they didn’t intend was for a global audience of thousands to see it, so they’ve sued the train operator for the fact that the footage was released.

Reputation Management wise, was this the best thing that they could have done? Now that they’ve sued, the press all over the world has picked up the story, however the press that I’ve seen has actually been fairly responsible and not posted either details on the couple, nor links to the video, which means that they’re not either pushing people to the video, or helping it to rank higher. This is good news for the couple, as it limits their further embarrassment. With this lawsuit they can maybe get a change made to the laws of their country to prevent this from happening in the future.

But what happens if the law goes too far? Where’s the line between privacy and the public good? Should a flasher have their picture posted on the web? Photos of license plates of cars cruising the wrong area of town? Video of an traffic law violators being pulled over? Something to think about in this CCTV globally linked world…

Because it is so easy to do just a little more research on the internet, many purchases are not made on the first visit to a web site. It is very common for someone to find exactly what they want to purchase, but not buy it immediately. They go to a few more sites to see if they can find a better price, a better deal or maybe free shipping. When they don’t find a better deal, they come back to your site and make the purchase. This is called a latent conversion. By ignoring latent conversions you may be working with misinformation in your web analytics.

One big impact of ignoring latent conversions is tha it makes your PPC campaigns seem like they are not as effective as they really are. For example, some can click on your ad, come to your site, find what they want but not make the purchase immediately. Instead they bookmark the product page, leave your site and later come back and make the purchase. Analyzing this case it would be a direct visit that leads to the conversion. But that is not correct. That sale should be credited to the PPC ad.

Most smart analytics packages look for this and do credit the appropriate ad and keyword with the conversion. The key is that either you need to be using a javascript based analytics package or you must use a persistent cookie. Since this data is cookie based, it will not be 100% accurate but you still should pay attention to this data.

It can give you insight into information that you are missing on your product page or other potential roadblocks to conversions. For this reason it is a good idea to segment out return visitors who make a purchase and review the paths that they take through your site.

I mentioned that the latent conversion data won’t be 100% accurate. That is true of most all revenue/conversion data in any analytic packages. That is not a criticism. It is a fact. So if you plan on comparing your web report sales data to your ecommerce database there are a few things to take into consideration.

Regardless of whether your web analytics package is javascript based or log file based there are factors that impact what gets written into the log files (either web server log files or log files generated by the javascript) which cause the analytics package to miss some sales.

It is also possible for your analytic package to overcount revenue data if someone hits submit a few times. You may want to read the information on how your analytics package handles this situation if you plan on trying to compare your sales numbers in your ecommerce database to your sales numbers in your web reports.

One other factor that analytics packages don’t generally handle is returns. The purchase may happen on your web site, but the return may happen through a order management system. This will cause your web reports to show more sales than your ecommerce database.

Web analytics are great for uncovering trends, determining which PPC and other marketing programs are effective and understanding a lot about the user behavior and experience on your site. They are not a replacement for your ecommerce database when understanding the full picture of your online sales.

Everyone’s heard of Writer’s Block, that condition whereby an author just can’t string 2 sentences together to save their life. Maybe it’s time to update that condition to Blogger’s Block? As we’ve talked about in the past, when you have a blog, you should post regularly, and blog on a schedule where possible. Pre-blogging, where you write your posts well in advance of their due date, helps immensely, as you don’t have to worry about last minute blogging. In fact, Scott Adams of Dilbert talks about doing the exact same thing in his recent book “Stick to drawing comics, Monkey Brain”. When he started out drawing Dilbert he was drawing the comic several months ahead of the publication schedule, now he’s only days ahead…

So what happens when you’re down to the last minute, and you have our newly redefined problem – Blogger’s Block? There are several things that you can do. The easiest thing to do is, of course, nothing. However, that’s not a strategy that’s going to be conducive to the long term growth of your readership, so instead, here’s a list of real (fairly quick) solutions for your problem;

1. Link Post: Do a post where you link out to posts or articles that you’ve recently read that you think are great. Why not expose your readers to them? If you’re worried about losing your readers, don’t. Instead you may find that they’ll like you for exposing them to some great new stuff, and they’ll come back to you to see what else you recommend




2. News Post: Similar to the link post, but this covers news about a particular topic. Maybe it’s your industry, or maybe it’s related to a topic that you posted about the previous week. Simply go to your favorite news feed site and search until you find articles that you like




3. Finish The Half Finished Post: If you’re like me, you have a nice selection of half finished posts. Take a look at them and see which may be worth finishing. Or you can do what I once did, and write a post about your half finished posts, giving reasons forwhy you won’t finish them. This actually ended up being a very popular post.




4. Rebuttal / Agreement Post: Read your feeds and see if there’s an article that you feel strongly about. Post your opinion about that topic, using the original post as the starting point, before you know it, you may have some back and forth going between you and the other blog that can be beneficial to both.




5. Analogy Post: Think about your activity that day. How can you relate it to a topic that you cover? For example, today I played soccer in freezing temperatures. Our regular field was unplayable, so we went elsewhere. That field had another group of people on it, with a permit, so we located another field, which worked out great for us. If I were writing a blog on technology issues, I could relate that experience to the need to have redundant backup servers in case of emergencies.




6. Picture Post: Sometimes a picture does say a thousand words. One blogger I know throws a “Caption Me” picture whenever he has Blogger’s Block, this both gives him time to come up with something else, and it involves his audience.




7. Historical Post: If your blog has been around for a while, it’s good practice to throw up a post with links to your best historical posts, these can either be ones that you’re particularly proud of, or the ones that drew in the most traffic (which aren’t always the same posts).




8. Blogging Inspiration Post: Erm, yes, as if you hadn’t guessed that this would be the last one.

In June, I warned an industry colleague not to purchase a compact digital camera at a deep discount because of the reviews and reputation of the online store he was considering. I found the reviews from ripoffreport.com, which is one of the largest “consumer advocate” sites on the Internet. The issues of bait and switch that caught my attention at Ripoff Report were well documented on other sites as well.

I am no stranger to Ripoff Report. In 2005, I took my chances with, and completed offers to receive a “free ipod” through Incentive Reward Center, even after reviewing scam alerts and poor customer service rants. As I completed offers, I took screenshots of each step, screenshots of the terms and services, the privacy policy, and collected any contact information I came across. I figured when all was said and done, if I did not receive an ipod, I would have spent $50, mostly on useless products and free trials, and then I’d go out and buy a new ipod at retail. Depending on how wronged I felt, I was going to report them to the Better Business Bureau, their local Chamber of Commerce, and anyone else who might remotely listen. A couple of minor hiccups and decent contact with customer service, and I received my top of the line ipod.

But what is it like to be on the receiving end of a complaint on Ripoff Report? One SEO agency (NOT US!), who also offers reputation management  services, received a complaint on the Ripoff Report. And they handled it by blogging on their own site about Google’s love of certain sites, including Ripoff Report and Wikipedia. It has been picked up by other blogs, positively, and they’ve handled it nicely. Though they’ve brought additional attention to the issue, they have also effectively put it to rest. It is a calculated risk that companies have to take when responding to user-generated content, that bringing greater attention to the issue will in fact mitigate it.