Gawkwire: Web Hosting and Internet News Resource

Data Center Efficiency:What about Customer Support?

2009-09-25T06:41:00-04:00

(WebHost Blog) There are a lot of definitions of efficiency, but I think my personal favorite is,” the ratio of output to input in any system.” If what you get is close to the amount of what you put in its efficient.

Energy efficiency is therefore the ratio between the amount of work produced and the amount of energy needed. When it comes to data centers there are a few extra incentives.

The most obvious is power is money. Not only does power cost, but the cost of power is rising and has been doing so for years. I remember an interview I did was Superb Hosting’s CEO Haralds Jass. This was a while ago, early 2000s I believe and back then he said the biggest problem will be power… and he was right.

The second incentive may not be obvious but it is self-evident to those who operate large amounts of IT equipment; the need to produce more work is also escalating.

I already have an article in the works for WebHostMagazine.com on the best practices of energy reduction.. and its quite lengthy; I have been meaning to make it smaller but I just don’t have it in me… so I won’t go over that here. But there is one thing I want to talk about that I only seen 1 other writer talk about (A great piece of work by Lex Coors at DataCenter Journal entitled Data Center Efficiency – It’s in the Design) and that is customer support.

I am going to take customer support a little further however. Virtualization is a great technology to have, but what happens when you have to deal with dedicated server customers? They are not necessarily going to want to share their server with others and although some may have multiple websites on their server, they may not be open to the idea of virtualization the server up (they might think its too much of a hassle, etc). One of the main problems with energy use is under use by servers. Four idling servers are using up power that could probably be better served doing something else. A web server that only serves one site running, on average, three hours of intense resource use then goes to something like 20% system resources the other 21 hours is a waste.

Packet Power mentioned charging for power like you would bandwidth. A good idea but it too is missing something; the customer service end. If you want to fully realize not just less power waste and still have happy customers, I think power should be a sellable metric and I think this opens up the way for power usage consulting. Technical support teams will have to learn how to implement best practices for power usage and servers and, more importantly, pass this on to the user. Now you might say, if I am charging my customers for power, would I not want them to use as much as possible? True you might want to do that but remember that you may not be charging them for all the power their system uses. For instance you probably won’t be able to figure out exactly how much power it costs to cool that one server or the total power loss from the central power bay to the server. Also you can gain more money or good pr by offering your customers a means to talk to technicians who can squeeze high performance yields out of every ounce of energy.

When it comes to the best practices in saving power, in the Web Host industry, it is not about making sure your house is in order… its about making sure your house and your customer’s is in order.

About David Dunlap
Over the past ten years David has been a prolific author of hundreds of blogs, commentaries and reviews found here on WebHostBlog.com, as well as WebHostMagazine.com and other sites around the Internet. David manages the daily operations at both WebHostBlog and Web Host Magazine & Buyer's Guide, and as the head editor, David uses his unique analytical skills to ensure that both sites maintain their integrity and tough, but fair minded, reputations. Prior to his active career analyzing the Web Host industry, David specialized in networking and communications for the U.S. government. David's expertise in traditional and search engine marketing has helped boost companies both inside and outside of the Web Host industry.

http://www.gawkwire.com/technology/data_center_efficiency_what_about_customer_support.html

SEM: Adding up changes – PPC Campaigns

2009-09-23T06:16:00-04:00

(WebHost Blog) We human like bold moves. How many of us have cheered out our favorite teams who put up the winning, homerun, touchdown, goal, basket, etc when there is no time on the clock and the air is thick with tension? Or how many of us have made a big change in life and thought, “I’m starting a new chapter,” ceasing to realize that the turn of each page is just as important?

Ok so maybe I am being a little more philosophical than I should be since I am going to talk about marketing, but just as well I think we know where I am going with this. When it comes to search engines the tiniest of fractions add up and depending on how they add up they can produce BIG numbers.

Okay so lets get started. Examples, we will use a simple PPC Adwords campaign, search engine ranking on a key term, annnnnnnnnnnnnd lets go with a banner landing page. (I am just going to cover the PPC one today)

PPC Adwords

A few years back I was in a session led by Curtis R. Curtis. I believe it was at HostingCon. The important part was one of the speakers discussed PPC Adwords. He asked us how many people spent say more than $50 a month, then $100, then $1000, then $10,000, and one person there actually spent $20,000. Now mileage of course will vary when it comes to small changes, the company that spends $20k a month who receives a 1% savings will save a lot more than a company that only spends $500 a month, but the principles are of course the same. For this exercise, I am going to use a company whose budget is $1000 a month and they are running it for a year, so a $12k a month budget.

The basic principle of Google’s Adwords is simple; make a text based ad, targeting certain keywords. The amount of money you are willing to spend on a click + how many people click on the ad = the ranking of where it will show up on the search term pages. That said, so you wrote a really good ad and you spend $4 a click. The 1-3 positions for that main key term targeted by you add are spending $7 a click. So we’ll say you are somewhere around the 10th spot, however, web surfers like your ad better and click on it. After a few days of clicking your ad moves up to #3. You are spending less money than the others, but your ad was rewarded for its content so you are in the 3rd spot and are saving $3 a click over the #4 spot. In this, extremely basic, example you saved $3 a click and with an ad budget of $1k that would mean you save $750 a month. This is an extreme example of course to illustrate what our objective is.

The other side of the coin is Google likes putting your ad in areas that are similar to your keyword(s) to maximize your ppc dollar. Your main keyterm might cost $4 a click but Google puts you on something similar so lets say the original keyterm was web host deals, Google, being clever thinks hmmm host deals, host and hostess? Awesome I will put this ad with the hotel keywords as well. Now you might get a ton of clicks. But when they click through they see that you are selling web hosting and not party or hotel hosting deals. So they leave, but you still spent the money on them clicking on the ad. Since Google is trying to save you money, we’ll say this particular one was costing you $1 and was clicked 200 times before you caught it. You were up $750, but now you are down $200 so, so far you have saved $550. Not bad, but a very simple step would have saved you $200.

Little Thing #1:Negative Keyterms

There are two forms of keywords, normal and negative. Normally means this is what I want my ad to appear on (this and its derivatives). Negative keywords are those you don’t want your keywords on. Last time I talked with Ken Jurina at Epiar, he said that Google allows for 10,000 negative keywords, Yahoo allows 250, MSN allows 1,022 (with a cap of 100 characters apiece). Ken and Epiar both are the authority when it comes to negative keyterms and PPC campaigns in general. For some info on them check their rare negative keyword skills at Epiar.com.

You made your campaigns you have multiple ads and you have spotted your target keywords. Two things you need to do is brainstorm your keyword terms and even specific words within your phrases, and figure out what other industries they are associated with and add them to your negative keywords list. Each day you should check your PPC stats and see if you got clicks from odd keyword placements, add those to your negative keyword list. Each one can save you one cent to hundreds of dollars and it shouldn’t be too hard to get a few hundred. If you actually want 10k of them then sign up with Epiar (If you have a big budget for your PPC campaigns, then you should go to Epiar or another quality consultant).

Little Thing #2: Ad swap

Review, rewrite, refresh! I am going to say this every time, over and over again. Any ad campaign online that you do you need to make sure your ads are fresh. Depending on the venue this can mean swapping them out everyday, every other day, or every week. The thing is we are not even looking at big changes, we are going for small ones first. Capitalize every first letter on the second line. Swap out one word something more actionable. Swap the first and second line. Capitalize words in your domain name. Whatever, there are billions of combinations for any single ad and you should try going through them. Those who do this I salute you, now the next step. Do not base your swapping on one keyword phrase and do not base your success on clickthroughs, it should be based on conversions.

So you have 10 ads ready to go. You put two up at 50% each and you track how many clicks and the conversions on those clicks. You notice one is rocking across the board the other one isn’t doing very well. Swap out the poor performer put in one of your new ones. Check it again. You notice one of the keywords (not your main) in your list is doing really well on the new one, but the new one isn’t doing so well for all your other terms. Put that keyword into its own placement with the good performing ad and deal with it separately. Going back to your main, drop the underperformer and swap it with a good one. The new ad has low clicks but you notice its conversion rate is through the roof. No time to be sentimental, the first ad had a great one but its time to swap it for the new ad.

This is a constant process - it is very important (to the point of if you are not going to do it there is no point in running PPC ads). It’s really not that much work either. The corrections take maybe a 10-20 mins to do per ad and you should already be tracking your ads anyway. This trick though has been known to save companies thousands and can really make your PPC campaigns pay for themselves with conversions.

Little Thing #3: Decrease your bid

In marketing, there have been many stories about companies that increased or decreases the cost of the products and ended up either making huge gains or classic blunders. One can hardly forget Cadillac and their ill-fated venture into economy cars. The car was not only failure but it knocked Cadillac out of the top luxury car spot in America because shoppers were now confused on whether Cadis were luxury cars any more. The thing is sometimes you can adjust the cost per click on your ads. After a few months of them being on there, drop the bid by a cent or so. Wait a week. If nothing happens drop it another cent. Going back to our example, it can get a maximum of 250 clicks on the main keyphrase (if you haven’t used Adwords, I am greatly oversimplifying this, in practice you get a more diverse amount of clicks and 1 cent can really save you a lot more than what I am about to say). our company has decided to be bold and reduced their bid by 10 cents. The “gamble” worked and their position is unmoved. Instead of spending $1,000 for the month and getting 250 clicks they instead spent $998.40 and received 256 clicks. One minor change is all it took.

Little Thing #4: Check your organic traffic

This one actually slips a lot of peoples’ minds. say you are have your Google ads targeted at keywords that you aren’t doing so well in (a noble and worthwhile goal) and then one day you end up placing high on a keyword in the organic search. If your organic position is better than your ad position, its probably a good idea to make that keyword a negative keyword for your Adword campaign. Why pay for a click when you can get them for free?

Well that’s four quick areas that can save you some cash with PPCs. All told, all of these can be a few hours a week and I would think a few hours is definitely worth the potential gains.

http://www.gawkwire.com/seo_sem/sem_adding_up_changes_ppc_campaigns.html

Gender Bias and Sexism in IT

2009-09-22T06:20:00-04:00

(WebHost Blog) I don’t often discuss topics like this on the blog, in fact I believe this is the first time, but an article I read made me think. The article in question is from TechNewsWorld entitled Is Sexism Rampant in FOSS? What I find funny about this title is that it almost connotates that sexism is only in the open source community and not rampant in IT in general.

FOSS of course is an acronym for free and open source software and the article was based off of findings in an article by Bruce Byfield and a second article by Yuwei Lin as well as comments from the open source community.

A blogger from Slashdot named David Masover made a pretty good point as to the differences between the corporate world and FOSS, “I don’t think this is specific to FOSS. I think, if a problem exists, it applies to all software development, and it’s just easier for management to force people to behave in a corporate setting: behave, or you’re fired. If you tried that in FOSS, people would just fork the project.”

Some of the factors for women not getting into FOSS include lack of mentors, discriminatory language, gendered based environment, male-dominated competitive view, lack of a female perspective, and no sympathy from women peers. To this one blogger wrote, “there is not a lot in there about sexism.”

I extremely beg to differ. All of the above points demonstrate varying levels of sexism., mainly a hostile work environment. A hostile work environment that has many tie-ins to gender is sexist. I don’t care who you are, or how much you can tolerate, if your work environment is hostile you become uncomfortable and then there is only two things to do, fit in or leave. If you fit in then you end up becoming a part of the problem for others. If you leave, well that only demonstrates the point.

In IT as a whole, I think the main problem of sexism stems from the lack of role models. One of my friends has plans for getting her PhD in computer science so she can become just that, and what is interesting is in many of her classes she is the only female. She doesn’t have anyone she can relate to, no one to talk to about how she can move forward career wise, or has practical experience in the upper echelons within companies. These are things that most men take for granted.

From what I have seen, hostile environments make it difficult for people to have longevity at a company, which means they will never get to the higher levels of management, and therefore will not become role models.

And now I pose the question; in an industry where talent has changed the face of our medium the Internet, time and time again, how much talent have we wasted, how many incredible programmers, managers, inventors, etc have we lost due to bias?

http://www.gawkwire.com/business/gender_bias_and_sexism_in_it.html

Botnet Causes Click Fraud

2009-09-21T06:27:00-04:00

(WebHost Blog) The Bahama botnet, so coined from Click Forensics, has found a means to mask their clicks and traffic as legitimate clicks and because of this click fraud will be seeing a surge.

It is really an elegant solution. First off its a botnet, so these clicks are coming from a variety of IPs which gets past the basic click fraud filter (too many clicks from the same IP automatically gets flagged as fraudulent). Second the botnet masks the source of the clicks. So instead of the PPC networking seeing the clicks come from some non-disclosed location in the Bahamas, the network sees the click as coming from a university, government office, or even libraries. Thirdly, this is not a strong arm tactic, the intervals between fraudulent clicks is interspersed, thus the network could click an ad then wait an hour to click it again or it could wait six minutes or 38 minutes, whatever.

The means by which the infection spread was also equally elegant. Remember not too long ago that the New York Times was tricked into putting a malicious ad on their web site? That had a hand in this. Also the Facebook virus scare dubbed Fan Check likewise had a hand in this. The malware used in the virus removal kits for the Fan Check “virus” and the malware used with the Times ad are eerily similar to the malware used with the Bahama botnet.

The problem with an attack like this is, its very hard to determine what is fraud and what isn’t. I would liken it to spam. Some spam is very easy to catch just like some click fraud is clumsy and easily found. Its when the fraud begins to mimic normal human patterns is where you have the difficulty and if the code for the Bahama botnet becomes more refined it may be nigh impossible to separate fraud from truth.

http://www.gawkwire.com/technology/botnet_causes_click_fraud.html

Marketing Jargon: What is Intuitive?

2009-09-10T06:36:00-04:00

(WebHost Blog) Intuitive is a word that gets thrown around a lot. You will see it in press releases, reviews, sales literature, and pretty much on anything that carries some form of interaction. Intuitive interface, intuitive controls, design, layout, well heck just take any word and slap intuitive on it and you can be your own marketing team.

Webster defines intuitive as “directly apprehended” and “known or perceived by intuition.” I am sure that last definition really helped.

Unlike most marketing jargon, I do not utterly distain intuitive. Although I think it has been used way too often, it is a word that should be the basic goal of one of the most important pieces of a Web Host’s offerings and that is the control panel. It should also be the basic goal of Web Designers when it comes to their design.

What is Intuitive? I define intuitive as any set of functions that do not require users to learn a new skill set. For control panels, if you take any computer user and sit them in front of a control panel they should be able to use all the functionality of that control panel without having to look at a manual or use the help icon. That is intuitive. Now I know what you are saying, how can someone with an extremely limited use of the computer actually use all the functionality of a control panel that a system administration would use? Parallels solved this problem in their Panels 10 SMB edition by adding user roles. If an employee only knows how to use email then all they get is email on their control panel.

For Web designers a web site is intuitive if a surfer can find what they are looking for without having to double back. That is intuitive. This includes but is not limited to the inclusion of multiple avenues of search/browsing (categories, tags, search bar, wiki-links, etc), landing pages for ads and news items, reducing the amount of clicks needed to get to descriptions (ie removing things like clicking on a product summary to get a product page and from their getting a full product feature list, then clicking on another link to bring you back to the buy page, and so on), charts and graphs for comparing multiple items, most popular sections, calendarized itemization, and the list goes on.

Intuitive maybe overused, it can end up being senseless marketing jargon, but if you provide services or products on the internet, then intuitive better be your standard.

About David Dunlap:
Over the past ten years David has been a prolific author of hundreds of blogs, commentaries and reviews found here on WebHostBlog.com, as well as WebHostMagazine.com and other sites around the Internet. David manages the daily operations at both WebHostBlog and Web Host Magazine & Buyer's Guide, and as the head editor, David uses his unique analytical skills to ensure that both sites maintain their integrity and tough, but fair minded, reputations. Prior to his active career analyzing the Web Host industry, David specialized in networking and communications for the U.S. government. David's expertise in traditional and search engine marketing has helped boost companies both inside and outside of the Web Host industry.

http://www.gawkwire.com/business/marketing_jargon_what_is_intuitive.html