I have been experimenting with the Pay-Per-Click campaigns and I paid a rather expensive tuition to learn how to get some positive results.
I used Google Adwords and Analytics. To use the method below, you need some sort of comparable system to keep track of what exactly is going on with the PPC campaign.
In addition, I had both image (content network) and text ads. So far the image ads have produced the most income for my product.
One of the things I did wrong was take the advice on how to calculate the maximum bid for clicks from a Web site. It suggested that you take your budget and divide it by the profit margin for your product to determine the number of units you would have to sell to cover the costs of the PPC campaign.
You would then take this calculated number and divide it by your conversion factor (the ratio of one sale to the average number of clicks needed to make that sale).
This would give you the number of clicks you would need to make the sales needed to cover the costs of the PPC campaign. The basic math is shown below:
Product Margin: $8.00
PPC Campaign Budget: $240.
Sales needed to payback advertising budget = $240/$8.00 = 30 Units sold
Average conversion rate: 10 Sales/Over 252 Clicks = ~ 4%
Number of clicks needed to sell 30 units on average: 30/4% = 30/0.04 == 750 Clicks
Maximum Bid Price: Campaign Budget/Clicks Needed = $240/750 Clicks = $0.32
The thing wrong about this method is that your conversion rate really depends on the quality of your clicks. And by quality what is meant is how likely is the click going to be converted to a sale (the conversion factor which is 4% in the example).
By following the formula above, the less likely the conversion, the more clicks you need. The more clicks you need the lower the maximum bid price for the click and the less likely the conversion.
So if you follow this formula you wind up paying almost nothing per click but you have a huge number of clicks that you costs you money but don't generate income.
The method that seems to work is to first set your monthly budget to a relatively small number to give you freedom to experiment without breaking the bank. Then carefully select your keywords and content to target your market. This is very important because it will impact your sales results and can misdirect you.
Also, limit the number of both keywords and content so that it is manageable. I have found that there is no reason to make huge lists of both. Also, with a limited budget, you will spread the results so thin that you won't be able to tell what is going on. Besides what usually happens is that most of the click through will occur on a few keywords or content sites and a few of those will generate any income.
Google Adwords gives an estimate for the bid required to make the first page. I generally go with that initially and then tweak it and observe the results.
A word of caution, some of the keyword bids are astronomical -- I have seen as much as $15.00 per click. Unless you have a huge budget stay with affordable keywords.
When you start to see what is going on, eliminate the deadwood and add more keywords and content similar to the more successful ones in the current lists.
Another thing that I have found is that it is okay not to use your entire daily budget. In fact, that is a good thing. Keep adding the more effective keyword and content sites until the daily budget limit is hit.
If sales keeping increasing then top out when the daily budget is hit, then increase the daily budget by a few dollars. It may take a week or more of collected analytical data to make a reliable judgement of what is going on.
If sales increase with the increase in budget repeat the increase. As long as sales track upward, keep increasing until the daily click costs is less than the daily budget. If the increase in daily budget does not result in an increase in sales volume look at which clicks are producing more costs than revenue and then either remove those keywords and content sites or decrease the maximum per-click bid. Conversely, look at the keywords generating the most revenue and increase the maxumum bids slightly. This should move you towards the top of the page -- at least until the competition increases their maximum bids.
Again this should be done incrementally until the costs for the clicks is less than or equal to the revenue generated by those clicks -- preferably much less than...
When the daily costs per clicks falls below the the daily budget, add more keywords and content sites using the more effective ones on the lists as guides. Do this until the click costs reaches the daily budget or sales do not increase. If the budget limit is reached and sales increased up to that point, then increase the budget. If sales don't increase, again prune the deadwood and add more of the effective keywords -- this might mean removing some of the new words that were added.
Keep repeating this process as long as the results are positive with increasing sales.
Good luck!
Thursday, June 4, 2009
Sunday, May 3, 2009
Invention as a Process
I think that most people have the misconception that an invention is a single idea for solving a common problem. Unfortunately, in my experience I have found that invention is a process that solves a series of problems. Often implementing the single idea will require solving problems dealing with manufacturing, distributing, advertising and several other problems associated with commercializing the product. That is the bad news. The good news is that like many other skills it can be acquired through deliberate practice.
If a there is already a commercial solution, a good place to start is to list all its short comings. If there is no current commercial solution there may be many reasons. One might be that the cost of the solution to the end user exceeds the pain associated with the problem. However, this is not always the case.
Perhaps the commercial opportunity associated with solving the problem has not been recognized. An example that comes to mind is the ketchup container that can be stood on its lid. The problem was that the viscous ketchup would take considerable time to flow out of the bottle.
The problem although widely known and the venue for many comedic situations -- real and dramatized -- was not solved until relatively recently. Despite the fact the solution for the apparent problem was pretty simple.
If no commercial solution exists then look at the way the "thing" is done now and list the "thing's" deficiencies. An example would be using a tool designed for one task to do another (i.e. a carpenter's hammer used to crack nuts).
Once the list of the short comings is complete, it can be used to generate the new product features. The features should be ranked according to severity of the associated pain. It is assumed that the bigger pains would generate the most financial return. In other words, the consumer would be willing to pay more to have the problem solved.
After the new product's features have been listed with the most painful at the top, a series of potential solutions should be generated. Do not stop at just one solution per feature, generate as many as possible. Also do not be limited to just one genre of solution -- all mechanical, all electrical, all chemical etc...
Try and generate as many solutions using as broad a range of methods as possible. Take several days, weeks or months as needed to create as many solutions as possible. It is usually the case that the ideas will slow to a trickle or dry up completely. When that happens go on to ranking the solutions.
After the list of solutions is created, a list of methods or designs to implement each solution should be constructed.
As before the design lists is evaluated for ease, costs, feasibility, etc... The whole product life line should be considered. That is, from the time of the start of making the product to the end of its life when it is disposed of by the consumer.
It is unfortunate that there are no absolutely pure products in the sense that there is a trade off between desirable positive consequences and the unavoidable negative ones. A judgement should be made considering the long term effects of making, selling, using and disposing of the product.
The rest of the process is somewhat tedious but involves ranking the solutions for the most net desirable effects and the lowest cost. It make take several iterations to find an satisfactory solution.
If a there is already a commercial solution, a good place to start is to list all its short comings. If there is no current commercial solution there may be many reasons. One might be that the cost of the solution to the end user exceeds the pain associated with the problem. However, this is not always the case.
Perhaps the commercial opportunity associated with solving the problem has not been recognized. An example that comes to mind is the ketchup container that can be stood on its lid. The problem was that the viscous ketchup would take considerable time to flow out of the bottle.
The problem although widely known and the venue for many comedic situations -- real and dramatized -- was not solved until relatively recently. Despite the fact the solution for the apparent problem was pretty simple.
If no commercial solution exists then look at the way the "thing" is done now and list the "thing's" deficiencies. An example would be using a tool designed for one task to do another (i.e. a carpenter's hammer used to crack nuts).
Once the list of the short comings is complete, it can be used to generate the new product features. The features should be ranked according to severity of the associated pain. It is assumed that the bigger pains would generate the most financial return. In other words, the consumer would be willing to pay more to have the problem solved.
After the new product's features have been listed with the most painful at the top, a series of potential solutions should be generated. Do not stop at just one solution per feature, generate as many as possible. Also do not be limited to just one genre of solution -- all mechanical, all electrical, all chemical etc...
Try and generate as many solutions using as broad a range of methods as possible. Take several days, weeks or months as needed to create as many solutions as possible. It is usually the case that the ideas will slow to a trickle or dry up completely. When that happens go on to ranking the solutions.
After the list of solutions is created, a list of methods or designs to implement each solution should be constructed.
As before the design lists is evaluated for ease, costs, feasibility, etc... The whole product life line should be considered. That is, from the time of the start of making the product to the end of its life when it is disposed of by the consumer.
It is unfortunate that there are no absolutely pure products in the sense that there is a trade off between desirable positive consequences and the unavoidable negative ones. A judgement should be made considering the long term effects of making, selling, using and disposing of the product.
The rest of the process is somewhat tedious but involves ranking the solutions for the most net desirable effects and the lowest cost. It make take several iterations to find an satisfactory solution.
Saturday, March 28, 2009
NLP Technique and an IT Project
NLP refers to Neural-Linguistic Programing. This is basically a thought management technique that deals with using internal mental images and language to alter behavior. Some of these techniques are applied in the area of communications.
The communication example I heard compared a conversation to two people walking. For this couple to stay together they have to start from a common point, head in the same direction at the same pace.
I took this to heart. I can't count on the number of times I made an assumption that the person I was talking to and I were starting from the same point. If I am lucky, the person I am talking "at" will pull me up short. If I am not so lucky, they will let me walk off with the delusion that 1) they knew what I was talking about; 2) they agreed with me and 3) they would somehow support my efforts in this area.
Prior to hearing the NLP communication analogy, I would just attribute their subsequent lack of support as deliberate sabotage or passive aggressive behavior. After hearing about the NLP analogy I made a deliberate effort to change my behavior and tried to improve my listening skills. I approached a conversation as a dance. As Franklin Covey recommends: I first tried to understand the topic from the point of view of the other participant, then adjusted my approach correspondingly and it quickly paid off.
So what does this have to do with an IT project you ask? Let me explain, I inherited a database system that was based on a software platform that was neither scalable or expandable. In short, we hit the end of the road. We needed to change the software platform and add features to the database system.
Prior my "enlightenment" I would have approached the project in probably a less effective manner. I would have done the work based on my interpretation of the end users needs with minimal effective listening. Any problems that came up I would blame on the end users lack of cooperation. After all, that was common operating procedure for me and a lot of my colleagues.
Now I approached the project from the point of view of the end user. The first decision I made was to start the new system so that it looked and felt almost exactly like the old system. What I tried to do is decrease or eliminate any learning curve associated with moving to the new system. So that was done. A consulting firm wrote the software to almost exactly match the old system's user interface. We had also set up two computers that periodically synchronized the databases to keep them both current. This allowed us to work on the system without disrupting the work of the end user. We would ping pong between the machines to introduce upgrades. When the new improvements were approved by the end users both computers were upgraded.
The end users loved it because it did not create more work for them or interrupt the work they needed to do now. After all, a new system that increases work load or makes it harder to do the existing work in not expected to be warmly accepted.
Next to upgrade system features, we asked the end users what they wanted next. With the list they provided, we prioritized the development work based on the criteria of: relative urgency, ease of implementation, and benefits. The features would be developed and installed on one of the computers and submitted to the end users for sign off.
I must add one clarification on the end users group: it did not contain all end users. It was composed of the group creating the information in the database. However, this database was being used by several other company divisions and the end user group was getting feedback from these divisional users.
The project was a success as defined by the end customer's satisfaction level which was relatively high. In addition, the project was relatively low budget and was very near its final form in less than 6 months. The work continued for quite a while after that because the users kept requesting new features.
Also I think all of us involved in the actual development work were pretty proud of the results we had collectively achieved. And once we settled in, the process was virtually friction free.
The communication example I heard compared a conversation to two people walking. For this couple to stay together they have to start from a common point, head in the same direction at the same pace.
I took this to heart. I can't count on the number of times I made an assumption that the person I was talking to and I were starting from the same point. If I am lucky, the person I am talking "at" will pull me up short. If I am not so lucky, they will let me walk off with the delusion that 1) they knew what I was talking about; 2) they agreed with me and 3) they would somehow support my efforts in this area.
Prior to hearing the NLP communication analogy, I would just attribute their subsequent lack of support as deliberate sabotage or passive aggressive behavior. After hearing about the NLP analogy I made a deliberate effort to change my behavior and tried to improve my listening skills. I approached a conversation as a dance. As Franklin Covey recommends: I first tried to understand the topic from the point of view of the other participant, then adjusted my approach correspondingly and it quickly paid off.
So what does this have to do with an IT project you ask? Let me explain, I inherited a database system that was based on a software platform that was neither scalable or expandable. In short, we hit the end of the road. We needed to change the software platform and add features to the database system.
Prior my "enlightenment" I would have approached the project in probably a less effective manner. I would have done the work based on my interpretation of the end users needs with minimal effective listening. Any problems that came up I would blame on the end users lack of cooperation. After all, that was common operating procedure for me and a lot of my colleagues.
Now I approached the project from the point of view of the end user. The first decision I made was to start the new system so that it looked and felt almost exactly like the old system. What I tried to do is decrease or eliminate any learning curve associated with moving to the new system. So that was done. A consulting firm wrote the software to almost exactly match the old system's user interface. We had also set up two computers that periodically synchronized the databases to keep them both current. This allowed us to work on the system without disrupting the work of the end user. We would ping pong between the machines to introduce upgrades. When the new improvements were approved by the end users both computers were upgraded.
The end users loved it because it did not create more work for them or interrupt the work they needed to do now. After all, a new system that increases work load or makes it harder to do the existing work in not expected to be warmly accepted.
Next to upgrade system features, we asked the end users what they wanted next. With the list they provided, we prioritized the development work based on the criteria of: relative urgency, ease of implementation, and benefits. The features would be developed and installed on one of the computers and submitted to the end users for sign off.
I must add one clarification on the end users group: it did not contain all end users. It was composed of the group creating the information in the database. However, this database was being used by several other company divisions and the end user group was getting feedback from these divisional users.
The project was a success as defined by the end customer's satisfaction level which was relatively high. In addition, the project was relatively low budget and was very near its final form in less than 6 months. The work continued for quite a while after that because the users kept requesting new features.
Also I think all of us involved in the actual development work were pretty proud of the results we had collectively achieved. And once we settled in, the process was virtually friction free.
Friday, March 27, 2009
The Unexamined Life Results In Low Quality
Another seminar on managing non-conforming materials in the medical device industry got me thinking about the root of the real problem concerning quality.
I have worked in engineering for nearly 3 decades and I have seen the same behavior many times as a result of new paradigms and procedures. Initially after the introduction of the new process culture, things improve. However, as time passes the improvement degrades and eventually the same level or lower quality results. And when I refer to quality I mean that the resulting product meets the intended goals.
I have a hypothesis concerning this phenomena. When people learn they go through the phases: conscious incompetence, conscious competence and unconscious competence. Upon introducing the new material people are put into the conscious incompetence phase. They have to think about what they're doing and learn the new material. As time passes, they become more familiar with the new stuff but still have to think about it to do it. This is the conscious competence phase. However, in the final stage people can perform the new dance in their sleep and often do. Basically thinking virtually ceases and the reasons for the procedures are forgotten and quality begins to decline and the rates of defects increases.
The problem is that the workers have fallen asleep on the job. I have observed this many times in new business paradigms, quality improvement processes, new products, etc... any time that there is a change that "wakes-up" the employees.
Perhaps a solution would be to have an automated system that measures some facet of the product and varies the procedures periodically to keep the workers awake. The automated system would do a statistical analysis of product variations and ideally select the process modifications to optimize the process.
I have an image of a robot controlled by this system running around on tank like tracks with a swinging appendage on the front that terminates in a big combat boot. The robot would run around and stimulate the workers at the "seat-of-the-problem"...
Just a thought.
I have worked in engineering for nearly 3 decades and I have seen the same behavior many times as a result of new paradigms and procedures. Initially after the introduction of the new process culture, things improve. However, as time passes the improvement degrades and eventually the same level or lower quality results. And when I refer to quality I mean that the resulting product meets the intended goals.
I have a hypothesis concerning this phenomena. When people learn they go through the phases: conscious incompetence, conscious competence and unconscious competence. Upon introducing the new material people are put into the conscious incompetence phase. They have to think about what they're doing and learn the new material. As time passes, they become more familiar with the new stuff but still have to think about it to do it. This is the conscious competence phase. However, in the final stage people can perform the new dance in their sleep and often do. Basically thinking virtually ceases and the reasons for the procedures are forgotten and quality begins to decline and the rates of defects increases.
The problem is that the workers have fallen asleep on the job. I have observed this many times in new business paradigms, quality improvement processes, new products, etc... any time that there is a change that "wakes-up" the employees.
Perhaps a solution would be to have an automated system that measures some facet of the product and varies the procedures periodically to keep the workers awake. The automated system would do a statistical analysis of product variations and ideally select the process modifications to optimize the process.
I have an image of a robot controlled by this system running around on tank like tracks with a swinging appendage on the front that terminates in a big combat boot. The robot would run around and stimulate the workers at the "seat-of-the-problem"...
Just a thought.
Hole in the Bail-Out Bucket
I attended a presentation a few weeks ago given by a venture capitalist. It seems that the money is trapped behind a wall of undefined requirements and procedures. In short, although the bail-out money is there, few people understand what the requirements are and how to apply for it.
In addition, VC's are moving to the less risky funding ( C-D levels) and away from the initial funding ( A-B ). So the higher risk start ups are suffering. And the VC's are demanding a bigger piece of the equity pie (80-90%) so the start up crew and the CEO are basically working for wages -- high risk and low pay with little or no equity at the end. This is going to kill the golden goose for sure!
Currently, I am trying to either increase my line of credit or apply for an SBA loan. I am figuring out the ropes with the help of my wife. She has a knack for figuring out bureaucratic document mazes. She is a freelance medical writers that deals with FDA requirements which I guess gives her the experience she needs.
Anyhow, we will try to figure it out. If we do, we will post our findings somewhere.
On another note, although credit is suppose to be loosening up, in the last few weeks my credit card interest tripled. And I have read of credit card companies cutting off or reducing the credit limits for business customers.
On the plus side, at least we're not getting 5 preapproved credit card applications a day any more.
In addition, VC's are moving to the less risky funding ( C-D levels) and away from the initial funding ( A-B ). So the higher risk start ups are suffering. And the VC's are demanding a bigger piece of the equity pie (80-90%) so the start up crew and the CEO are basically working for wages -- high risk and low pay with little or no equity at the end. This is going to kill the golden goose for sure!
Currently, I am trying to either increase my line of credit or apply for an SBA loan. I am figuring out the ropes with the help of my wife. She has a knack for figuring out bureaucratic document mazes. She is a freelance medical writers that deals with FDA requirements which I guess gives her the experience she needs.
Anyhow, we will try to figure it out. If we do, we will post our findings somewhere.
On another note, although credit is suppose to be loosening up, in the last few weeks my credit card interest tripled. And I have read of credit card companies cutting off or reducing the credit limits for business customers.
On the plus side, at least we're not getting 5 preapproved credit card applications a day any more.
Subscribe to:
Posts (Atom)
