Ask the SEO Guru

Question for the Week of May 29th, 2008

I had a drive by comment on my last post today and I felt that it is a good enough question that it should be a topic and not just a reply. It is something that I feel doesn’t get discussed nearly enough in the seo community so this is my chance to do more than just complain about it.

This comment was made in reference to my last post about an interview I gave last week.  Interview is here Eric Enge @ Stone Temple

And the post is here Page Rank - where it's been and where it's going

Here’s the question:

Actually, the heart of watching the numbers is to ensure you hit your objectives, but the need to demonstrate ROI certainly SHOULD be the primary purpose, at least in most cases,  but is actually whatever the client tells you it is. Believe it or not, most clients come to a website promoter demanding, insisting on an expecting – placements. And  if a client comes to you telling you they want o be number 1 for anything, then it is usually difficult to attempt to educate them on looking for ROI without talking down to them and making them feel a little silly for not thinking of it themselves. 

It's sad but true. I usually simply do what they ask and try to develop the kind of relationship with them over time that enables me to be more forthcoming without offending or talking down to them. I also find it easier to establish that kind of relationship while they are paying me as opposed to paying one of my competitors.

As more and more people become more and more educated about the web and what it can do, that is changing slowly. Painfully slow. Even so, I still rarely equate numbers with ROI specifically and rather think of it as simply objectives.

So, whether it is ROI, tracking referrers, conversions or just wanting traffic levels to increase the perceived value of your online real estate, they are all objectives and indicators need to be watched if you expect to know when you are on the right track and when you’re done.

I always watch indicators and as for the term “important indicators”, that’s a subjective term. If you think they are important – you’re right.

You can find stats and analytic programs off the shelf to track, store, alert and suggest based on it’s own data for pretty much any metric you can dream up. I’ve run across a LOT of would be marketers who run so many stats and analytic programs that their life is little more than checking stats and running reports. I suggest you get the stats you want but only for the purpose of hitting your numbers. When you spend so much time tracking customers that you don’t have time to sell something, you lose!

So, step 1 is to decide what you want and why.

Let’s first look at just basic traffic and referrer tracking.

For simply tracking traffic levels and referrers,  there are LOTS of free stats stuff out there that will give you that. Awstats, and Webalizer are both free programs that come with just about any hosting account or server account. Then there is Extreme Tracking. One of the oldest free stats programs out there and does fine for basics if you don’t mind displaying a goofy little button on your site that anyone can click and view your stats without a password, (not recommended for anyone serious about making a few bucks online). There are ways around it but it’s against their TOS and I believe if you’re gonna play, play fair.

Doing a search for free stats programs will give you plenty of options to slog through but I have had experience with each of the ones I mentioned and they are easy and simple and do an acceptable job for the basic stuff.

Now let’s talk about the ROI tracking.

I know this topic gets thrown around a lot in SEM circles and virtually every post regardless of quality or credibility gets that knowing nod from the members with little challenge. I too am guilty of letting broad and often nonsensical, ”ROI is everything” statements stand without speaking out, but now we’re in MY house and I can say what I want without the flames, so ------

The Reality is You Can NOT Prove or Demonstrate ROI to a Client.

At least not unless you have control over or intimate knowledge of, cost of goods, pricing, inventory levels, floor planning, interest rates, taxes, shipping costs, hiring and salaries, bonuses and commissions, returns, missed deadlines, customer service efficiency, and a few other details  but I’m going to assume you get the point.

Any one of those things could have as much an impact on ROI as how many hits you get to a webpage or any kind of hits to orders ratio. You can demonstrate how he paid you $10,000 and within 180 days you generated $100,000 in orders BUT, orders aren’t profits and that $10,000 was likely a fraction of the investment he expects a return on.

Sure, you can show a client that you had a 40% conversion rate and prove that you had 100 hits and 40 orders but if the CEO is spending all his money shooting up pot and snorting blondes, he’s going to go broke and you lose a client. Period.

Few clients are going to give you control or knowledge of those other types of metrics or tell you how much pot they’re shootin up.

There certainly are people out there hiring SEO’s who damn sure do know how to determine ROI, just as sure as there are a lot of people hiring SEOs who don’t but both types of clients are still businessmen and they still have their own ideas of what they need to do to determine their own ROI .

So my point is simply that I believe the smart thing to do is let the client tell you what metrics they expect you to give them and then find the program or system to monitor and report the metrics that give the client what he expects. To me, that enables you to control your own income objectives and ROI.

That said, if you can’t generate revenue through your efforts, then eventually you will lose the client. If you can generate sales as a direct result of your work, then even if the CEO is sharing his money with three ex-wives and an IRS agent, he is still more likely to continue with you because a bad business with sales is still much better than a bad business with no sales. Odds are very good he will still eventually go broke but there can be no question that the more sales, the longer even a bad business can survive.

Keep in mind that for the last year my focus has shifted dramatically from providing SEO for individuals to wholesale link building and content generation for other SEO companies, so I don’t have the time to focus on analytics, (occasionally someone makes it worth my time), like I used to. Plus, for the last 6 or 7 years I’ve mostly used my own custom program for tracking and analytics but when I did use commercial stuff regularly, here is what I used and what I learned about important indicators early in the game.

Again, step 1 is to determine the objective and what data you need to track the numbers to make sure all your indicators are pointing north.

Most important in my opinion is software that does more than just collect referrers, IP’s, words from search engines, etc. What you really want to know is where a hit came from on a page, how long the hit stayed and which link they clicked to leave the page.

What you’re looking for is a hit getting to a page and what referrer they came from. If it was a search engine, what keyword/phrase did they use to find that page? If it was another webpage, (especially forums/blogs/social sites), what was said to make the link relevant enough to get the user to your page? Finally, how long they stayed on the page until they clicked a link or the back button?

So know you know someone came to the page and why. Does the content they get when they get there relate to the content that made them come in the first place?  Always remember that once you hit a page, there are only 3 ways for them to leave.

NOTE: there are some pretty tricky browser scripts out there that offer options, but that stuff is beyond the scope of this discussion.

#1. Close the browser. This happens VERY rarely. So rarely in fact, I don’t feel it worth discussing  in this post.

#2. Click a link that takes them somewhere else. (remember, YOU control these)

#3. Hit the back button. By far the most common exit from a page.

#2 and #3 are the ones we are interested in now.

You can’t track a back button hit, (I wasted a lot of time and money trying to develop something that would--- and failed.), but you can track that they didn’t hit anything else. There may be something available now that I’m not aware of because like I said, I’ve been using custom software of my own for a long time now.

If you could track the back button, you could estimate how much time the average person spent on the page before leaving. That way you could fairly accurately calculate how much the visitor for which referrer read before they left. Then just keep altering the text until they start clicking the links you want them to click instead of the back button. Because by far the biggest reason people hit the back button is because they don’t “feel” like they got what they were expecting when they first clicked a link somewhere to get to this page.

The Back Button Boogie

Even though I was never able to figure out how to track the back button hits and how fast, I’ve been doing this kind of work a long time, I’ve worked with a LOT of people and software and I’ve asked a LOT of people a LOT of questions. From all of that experience, (experience is what you get when you’re expecting something else.), I’m able to make an educated guess with a high degree of confidence. You have about 3 to 5 seconds to convince a visitor that they found the page they were looking for when they hit that last link to get to this page. If you don’t convince them, they do the back button boogie and they’re gone.

So, if I can’t track the back button but I CAN track any other link they click on the page, I watch the referrers closely and make sure that the majority that get to a specific page all came from closely related terms. If I see I’m getting hits to one page for shoes and tennis rackets, I know I’ve screwed it up and start looking for ways to separate the two options and get shoe people to shoe content and tennis people to tennis content. Even if it’s not feasible, (did I mention I have custom software?), then I choose the highest number of related referrers and talk to them and just accept I’m getting the back button boogie for almost all the others.

Once I am confident that I’m getting targeted visitors, then I structure the content so that I’m basically saying in 5 seconds or less, HEY, YOU SMART PERSON OF THE WEB PERSUASION, YOU FOUND THE EXACT RIGHT PAGE AND HERE IS THE HUGE BENEFIT YOU GET AS A REWARD. Then offer them a series of navigational links to lead them into a conversion process and to be able to gather data on what these people looking for this thing has responded to.

Usually the links I offer are the basics. Click here for pricing, click here for specs, click here for discount coupons, etc.

Now when I get 100 people to the page and 40 of them click one of the 3 or 4 links I offer, I know two things.

#1. 60% of the people did the back button boogie

#2. I said the exact right thing to get 40% to move them towards a buying decision

Now, just keep repeating the process until you get the order. If they click informational links, then give them information in a feature/ benefit/ call to action format and then offer them  another 2-4 link options based on the referrer, (what they just clicked on from your own previous page).

At each step, keep looking for that 40% number. I suggest 40% because over the years that has consistently worked out to be about a 5-10% conversion ratio from all the hits to the domain. I always shoot for 100% on each page and I always look for ways to get closer to that 100% but there comes a point of diminishing returns and you find your time better spent elsewhere.

There is a LOT more we could go into. Analytics is certainly an art that those who master it profit well from but the basics I’ve outlined here are the kind of stuff that anyone can do. I’m sure of that because if I can do it -- ANYONE can do it and doing it takes your skill set, and revenue potential, to a whole other level beyond search engine placement and traffic generation.

I’ll be honest, I rarely do this level of work for individual clients any longer because it is very labor intensive and few clients understand it well enough to justify the fees I have to charge to be able to do it. But I certainly do it for my own stuff. At least my stuff I care enough about. I should be doing more of it and so should you.

Below are some stats programs I’ve used over the years. I don’t recommend one over the other because to me, it is mostly a matter of preference. They all do a lot and do it pretty well but they each left me with the motivation to build my own.

Web Trends     Omniture      Netiq  

I urge you to consider these carefully before making a financial commitment. Good analytics can get expensive and you need to be sure that the program you choose will do the job you want it to do.

The next ones are free but sometimes free costs far more than simply paying for what you need.

google analytics, (formerly Urchin)  

awstats  

webalizer

extreme tracker

I already gave you links to these sites above. Jeesh! You're too lazy to scroll up?? You really are an SEO huh?

Peace Y’all

G

Course you broke your little sister’s bike. You knew damn well you were too big to be ridin that thang! Now just wait till your father gets home boy. He’s gonna let you have it.