Friday, December 01, 2006

Microsoft Adcenter 0.1 - What were they thinking?

I continue to follow the different ad platforms from the major search engine companies. I'm still shocked that Yahoo and Microsoft were not able to launch a better ad platform after pooring milllions of dollars into development. Both companies understand usability, but it seems that they have ignored usability and speed in the new ad platforms.

Here's my top list of ad platforms:

1. Google Adwords - by far the best plaform. Improvements are made monthly and Google continue to impress me.
2. Ask Paid listings - I've only just logged into this. It looks good, but lacks quite a few features
3. Adbrite - interesting solution.
4. Text-link-ads - another interesting solution
5. Yahoo Search Marketing - Better than the old Overture platform, but it failed to reach my expectations.
6. Microsoft Adcenter - Microsoft - What were you thinking?

After reviewing these platforms, I'm confident that it will take a few years before the other players catch up to Google.

Wednesday, November 08, 2006

A potential description to Youtubes and Googles copyright problems

I believe somebody will come up with a solution to the copyright problem.

Here's one potential process:

  1. A copyright-owner goes to, uploads their content and activates the copyrights.
  2. The copyright association partners with major text, video, sound search engines and crawls all the content in the world.
  3. When the search engines find illegal content, an automatic e-mail is sent to the website owner and the hosting company.
  4. In the e-mail the website owner and the hosting company is introduced to a set of tools that enables them to automatically remove the content, identify the person who uploaded the content and temporarily blacklist the IP-address of the uploader. The system also provides the siteowner with tools to monitor the illegal uploader.
  5. The website owner can also implement a micropayment solution that's completely integrated with Paypal, Google Checkout and other payment systems. This way, all websites can offer copyrighted content, make some money on it and pay the content owner in a reversed affiliate program™.
Any comments?

Thursday, October 19, 2006

McDonalds with innovative blog initiative

Update! McDonalds actually follows this campaign up personally. This post did not qualify! Hurray for McDonalds. This might be an example of a successful blog marketing campaign!

I'm impressed. I used to go alot to McDonalds, but after the movie SuperSize me, I rarely go there. However I do love their milkshakes. I live in Stockholm, but I'm frequently in Oslo, Norway.

This picture is taken from McDonalds in Oslo, where I used to go alot.

So again, I'm not a fan of McDonalds food, but one positive aspect is the travelling aspect. Since I know McDonalds has really strict guidelines regarding their food, I always know that McDonalds serves the same food all over the world. So to me - McDonalds makes me feel safe when eating in foreign countries. Even though it's junkfood, I know that it's safe compared to those noname restaurants. You can read more about the McDonalds initiative here.(in Norwegian). An interesting part of this marketing campaign is that McDonalds is trying to stimulate the conversation by paying money to bloggers. That means that I make money for blogging about mcdonalds. They are walking on a thin line here. Bloggers might react with rage since McDonalds is trying to commecialize blogging.

Wednesday, October 18, 2006

Google replaces "I'm feeling lucky" with "shop" button

Consumer research released by last week showed that when tested head to head against Google, users preferred’s shopping search results by a 55% to 45% margin.

Some time ago, information about Google's "I'm feeling lucky" button leaked from Google, showing that almost noone uses the button.

If Google's goal is to give relevant information to the searchers, they would replace the "I'm feeling lucky" button with a button that says "Shop".

This could improve the search results for the people looking to shop. It would increase the conversion rates for online merchants and Google would make much more money on ads showing up when people hit the "shop" button.

On the otherhand, Google might confuse the users by changing a button. I think it would be a smart.

Tuesday, October 17, 2006

Google Analytics vs. nettracker and other web analytics packages

Alot of discussions about web analytics turns into arguments concerning page view, hits, unique visitors, visits, etc. The focus should be on the trends. I've taken stats from Nettracker and Google Analytics and compared the two different stats. Here are the results.
As you can see the trends for both statistics packages are the same. On average Google Analytics Visits number is only 58% of the values that Nettracker reports as visits.

It doesn't matter which one is correct as long as the trends are similar!

Tuesday, October 10, 2006

10 ways to improve Google search

Larry Page once said: "The ultimate search engine would understand exactly what you mean and give back exactly what you want".

I believe Larry Page, co-founder of Google, is wrong!

I think it's possible to go beyond giving people what they want. I believe that a search engine that gives me "more than I want" would be better than Larry's ultimate search engine.

Larry wants the search engine to meet your expectations. I want a search engine that exceeds your expectations.

So here are 10 ways to improve Google search. Somebody has probably written all of this before and Google(or somebody else) is probably working on all of it as we speak, but I'm writing it anyway.

1. Improve inbound link factoring. Instead of just valueing inbound links from quality websites like PageRank does, give higher scores to inbound links from websites that my friends and professional network value. This shouldn't be that difficult using data from my list of MSN-friends, the people I send mail to through gmail and the google toolbar.

2. Ask clarifying questions. When a user searches for Columbia, asks you: "Did you mean Space Shuttle Columbia or the country Columbia".

3. Include conversion factors in the ratings. Increase the search scores of sites with high conversion rates. These sites probably have desirable product offerings that are more relevant than sites with low conversion rates.(Google is probably already measuring this for a lot of companies)

4. Increase the search scores of sites with loyal users. A site that people are loyal to is probably a good and relevant site.

5. Introduce a human side. Connect the Google search results directly with Google Answers. Can't find what you're looking for? Get a quality answer to your search from a real person that's an expert on what you're searching for

6. Add e-mail notification features. Include the option of e-mail notification when a person responds to your search

7. Add Social search. Add functionality similar to where google's search result pages are enriched with information about the sites from people I trust.

8. Analyse sentences. Give the user options in the search results to let the user narrow down on the search. See how the search engine does it.

9. Switch the "I'm feeling lucky" button with "I wanna shop". Usability studies performed on indicate that people like the two options better than Google's options. According to Google studies, nobody uses the "I'm feeling lucky" button on Google anyway.

10. has some nice people search functionality. I can actually type down the name of a person in Norway and get address information and which companies that person is affiliated with. Pretty neat.

Feel free to comment and improve. Some of my suggestion categories probably overlap. I will try to improve the list as I get comments.

Wednesday, August 23, 2006 and Google Adwords combined

Another smart move by Salesforce and Google. In a lot of organization there's a divide between sales and marketing. With this new initiative by and Google, the gap between sales and marketing becomes smaller. It's now possible to compare the effectiveness of a cold-call and a Google Adwords-campaign. What a blast.

Check out the video that explains how it works

Sunday, July 09, 2006

A quick guide to Google Analytics

This brief guide to Google Analytics has 4 main sections. These are:

1. Google Analytics Executive overview

This section contains import numbers from a birdseye perspective. This report should be used by the CEO, the director of Marketing and other executive that is interested in how it's going for the company website.

Quick overview
The report gives you a quick overview of how many visitors your website has in any given period. You can also see how loyal your audience is. How many people visit your site for the first time and how many people return for a second visit. If the percentage of new visitors is high, you're doing a good job in attracting new visitors to your site. This is good, but if it's too high, it means that people don't come back! It's important to see the balance of driving new visitors and getting customers to come back.

Where do my visitors live?
The executive report also provides you with a map of where your visitors are from. This can be used as a measure of which people in your marketing department are doing a good job. If you spend your entire marketing budget on marketing in the US and all your traffic originates from Japan, you should start to wonder!

What sources generate traffic?
The last part of the executive report is a report on traffic sources. This gives you an idea on what activities are generating traffic.

If alot of your traffic comes from Google, it's a sign that your search marketing personnel are doing a good job. If your main source is "Direct", it might mean that you have a strong brand and that a lot of people go directly to your site. If your traffic is fairly low, high direct traffic numbers means that you're not very well-known and you're not very visible anywhere.

If you want to drive more quality traffic to your site, you should check out Google Adwords and Overture. If you need help with this, you should check out Here's you'll find a list of people/companies that are Qualified Google Professionals in your country.

Another basic advice is to make sure your site follows Google's webmaster guidelines. This might be a technical challenge, but you should require your vendor/your IT department to abide to these guidelines. If they don't - switch vendors/personell:)

2. Google Analytics Marketing optimization

This section of Google Analytics should be used by the marketing people that want to drill down in the report details. Here you'll find information

3. Content optimization

not done yet

4. E-commerce analysis

not done yet

An Investment Approach to Marketing

This article explains how to use web analytics to help guide your marketing spending. It contains step-by-step instructions on how to get the most out of your marketing dollars.

A major goal of any company, large or small, is to maximize the value of their marketing budget. To do this, many companies strive to calculate how much a new customer is worth in terms of future sales – the customer’s “lifetime value” (CLV) – in order to determine the ROI of their marketing efforts.

Calculating Customer Lifetime Value

As I stated above, the customer lifetime value approach tries to estimate how much each customer is worth in dollars over the lifetime of their relationship with the company and thus discover exactly how much a marketing department should be willing to spend to acquire each new customer.

However, this approach has many challenges and often ends in failure. Why? The main reason is that in most cases these CLV models become too complicated for practical usage and therefore have little value.

The basic calculation appears quite simple upon first glance. You merely establish the customer lifetime value by estimating the average profit per sale (total revenues minus marketing, advertising and fulfillment expenses, divided by number of sales). Then you factor in how many times the average customer will purchase from you over a certain period of time (e.g. 3 years).

On the surface this seems like a quick and easy calculation. In reality, each step in the process is filled with challenges and presents opportunities for discussions and doubts. Many companies stumble due to the lack of data.

Others stumble because of senior peoples’ different opinions on which factors the CLV model should contain and how to calculate them. Disagreements usually arise when you try to:
  • Determine the how much of the fixed costs should be allocated to each acquired customer

  • Estimate the percentage of customers you lose

  • Adjust for how many customers you would get with no marketing activities.

  • Adjust values for cannibalization between channels

  • Estimate CLV adjusted for seasonal changes or other complex factors

  • Include the value of referrals in your calculation

  • Segment customer groups or calculate one overall number for the company

Simplicity is Genius

Instead of getting into long discussions on how accurate the customer lifetime value needs to be, one should focus on simplifying the model so that ordinary people can actually use it to make smarter marketing decisions on a daily basis.

Jack Welch, former CEO of General Electric, once said:

"Simple messages travel faster, simpler designs reach the market faster and the elimination of clutter allows faster decision making."

In order to simplify the model, we need to accept that CLV calculations, as with all other marketing forecasts, can and will be inherently inaccurate. So instead of seeking to create a better CLV model, we should choose a simpler method of determining which marketing mediums to use.

One simple method that you can use is marketing cost per sale. The marketing cost per sale is calculated by dividing your total marketing expenditures in a certain time period by the total number of sales generated in the same period. This calculation gives you some insight into the return on investment of each marketing dollar while avoiding the complexity of CLV. It is fast, simple and thus powerful.

Simple Messages Travel Faster

For example, if your marketing budget is $2,000,000 per month and you sell 5,000 widgets in the same month, your average marketing cost per sale is $40.

A lot of seasoned marketing professionals will debate the accuracy of this simple model. When this happens, it's important to stay focused on the actual goal of the model, which is to provide marketers with an easy-to-use tool that they can use to make smarter marketing decisions on a daily basis.

Practical Usage:

Monthly marketing budget: $2,000,000
Number of widgets sold per month: 50,000
Average marketing cost per sale: $2,000,000 / 50,000 = $40

If I invest $50,000 in a banner ad campaign on (Norway'slargest newspaper) and get 2,500 sales from the campaign, the marketing cost per sale is $20.00.

If I invest USD $20,000 in a banner ad campaign on and get 2,000 sales, the marketing cost is $10.00.

If I invest $25,000 on Google Adwords and get 6,250 sales, the marketing cost is $4.00.

If I move the $70,000 I spent on banner ads over to Google Adwords, it would give me 17,500 sales instead of the 4,500 sales I obtained with the $50,000 and $20,000 investment in newspaper banner ads.

Of course, there is a limit on the search volumes but until you reach that volume, the goal is to move your marketing budget gradually over to the channels that have the lowest marketing cost per sale. The idea is that if I can identify a marketing channel that gives me more customers for the same budget, a natural reaction should be to allocate more marketing money into that channel.

How to Get Started

In the Medium ROI report in your web analytics software(i.e. Google Analytics, Clicktracks, Webtrends), it is possible to accurately measure the results of different online marketing activities (banner ads, affiliate programs, search engine marketing, e-mail marketing). You can tell exactly how many sales your latest e-mail campaign generated and how your banner ads do compared to your search engine marketing efforts. Then by simply dividing by the marketing costs for the different activities, you can estimate the marketing cost per sale in each channel.

Since it's very difficult to measure the effect of offline advertisements (TV,radio,etc), we can make the assumption that the cost per sale from these channels is equal to or greater than the average cost per sale for all marketing activities. Many people will argue that this assumption is too inaccurate and/or biased but, more likely than not, the cost per sale for offline activities will be much higher than the average cost per sale for online activities.

Cut Costs and Report results

At the end of every month, you should take a look at your marketing allocation chart to see whether you managed to move money into more cost-effective channels. If you continue to gradually move your marketing funds towards the low-cost channels, you'll see marketing savings for your company. Your monthly marketing allocation might look like this:

All numbers are fictive numbers

A good salesperson knows what it takes to make a sale. They know how many phone calls they have to make to set up a meeting and they know how many meetings it takes to make a sale.

A good marketer should also know what it takes to make a sale. They should know how many searches/impressions have to be made to get a qualified click and they should know how many qualified clicks it takes to make a sale.


Marketing at Wharton

Jack Welch quote


KP's blog