What can we learn from all the talk about innovation?
In sum, it appears that new markets emerge when the rank ordering of current product features is upended or extended to focus on new priorities.
This may take the form of 'breaking the rules', i.e. relaxing an assumption that prevails in an industry -- like NetFlix did with late fees or iPod did with the assumption that it was about technology.
It may take the form of extending the feature set to add new items of value -- like Cirque Du Soleil did with multiple productions and artistic dominance. Kim and Mauborgne's Blue Ocean Strategy uses this approach in the creation of a visual strategy canvas.
Customers typically ask for more of the feature that you market -- sell based on capacity, they'll ask for more capacity. Sell on speed (not a very defensible proposition), they'll ask for faster. In short, current customers are conditioned to respond pretty much in the way they are sold. So asking them what they want is risky from an innovation point of view. This is pretty much the argument in Innovator's Dilemma, Clayton Christensen's work on understanding why firms fail.
This suggests that there are three questions one must ask when looking for new markets:
1. What are all the possible features we could include in a product?
2. What markets emerge if we change the rank ordering of currently available features?
- portability is more important than capacity
3. What markets emerge if we swap new features for current offerings?
- theme and story replace animals
The final point is understanding the attractiveness of a market; as is well proven that depends on where you sit. Innovative, particularly disruptive, markets have the benefit of being pure upside with no base to protect. So most innovation appears to come from small, upstart companies with 'nothing to lose.'
Tuesday, October 31, 2006
Monday, October 30, 2006
Cost Savings Are Customer Benefits
How do customers react to cost saving initiatives?
My library (yes library) has introduced radio frequency tagging, self-check out, automatic check-in, email and phone alerts, and a few other technical goodies. Now, the original purpose of this initiative was to improve the efficiency of material handling -- something like 7% annual growth with over 12 million items in circulation. Classic problem: How do we do more with same resources?
This may be another example of Jevon's Paradox --- increase in efficiency is supposed to reduce consumption but actually creates a greater demand. I know my use is up: I can log on to the library, search, and have the book, CD, or DVD shipped to my local branch. When its there, they leave me a voice mail.
According to Jim Cooper, Director of the county library system ....
Haven't tried the library's books to download; that's next.
My library (yes library) has introduced radio frequency tagging, self-check out, automatic check-in, email and phone alerts, and a few other technical goodies. Now, the original purpose of this initiative was to improve the efficiency of material handling -- something like 7% annual growth with over 12 million items in circulation. Classic problem: How do we do more with same resources?
This may be another example of Jevon's Paradox --- increase in efficiency is supposed to reduce consumption but actually creates a greater demand. I know my use is up: I can log on to the library, search, and have the book, CD, or DVD shipped to my local branch. When its there, they leave me a voice mail.
According to Jim Cooper, Director of the county library system ....
According to the Public Library Data Services Statistical Report, we are one of the busiest and most efficient library systems in North America; I think our staff does an outstanding job considering the volume of materials we move and the number of patrons who visit our facilities. Nevertheless, we constantly look for ways to improve our service.
Haven't tried the library's books to download; that's next.
Marketing Lights Up Las Vegas
What makes the meter go round?
In an excellent piece on the data centers required to power the search engines, George Gilder describes the future and the need for more infrastructure. Since search revenue is 95% advertising marketers might want to know what they're paying for. It turns out those 10c Google clicks add up to cost of the electricity necessary to light up Las Vegas.
So, what are the major players doing to keep costs down? They're building huge data centers along the Columbia River. Why? Two reasons - first, cheap hydroelectric power; second, close access to the Internet backbone.
Now it is a good thing that the Army Corps of Engineers and Mulholland didn't already divert the Columbia to Los Angeles. Yes, this was considered. For a good read on the history of water in the West I suggest Marc Reisner's "Cadillac Desert."
In an excellent piece on the data centers required to power the search engines, George Gilder describes the future and the need for more infrastructure. Since search revenue is 95% advertising marketers might want to know what they're paying for. It turns out those 10c Google clicks add up to cost of the electricity necessary to light up Las Vegas.
So, what are the major players doing to keep costs down? They're building huge data centers along the Columbia River. Why? Two reasons - first, cheap hydroelectric power; second, close access to the Internet backbone.
Now it is a good thing that the Army Corps of Engineers and Mulholland didn't already divert the Columbia to Los Angeles. Yes, this was considered. For a good read on the history of water in the West I suggest Marc Reisner's "Cadillac Desert."
Friday, October 27, 2006
Marketing Metrics - Finding common ground
What time is it?
Up until the late nineteenth century there was no way to agree on an answer. Every location had its own time- there were thousands of 'noons' in the US alone. The establishment of national railroads and the prime meridian finally gave a reason and a method for setting a common basis for understanding. Imagine a train timetable where you might arrive before you depart and providing you with no sense of duration. In 1884 a fixed location of 0 degrees longitude was established at the royal observatory in Greenwich allowing location and hence time to be defined precisely. From then on things made sense.
In thinking about marketing metrics it seems we're still in the late 19th century with hundreds to choose from and consultants promising to help find the right ones. Finance went through similar throes and ended up with one or two with cash flow being on everyone's short list.
It might be time to think about the ones that really matter at the end of the day.
Up until the late nineteenth century there was no way to agree on an answer. Every location had its own time- there were thousands of 'noons' in the US alone. The establishment of national railroads and the prime meridian finally gave a reason and a method for setting a common basis for understanding. Imagine a train timetable where you might arrive before you depart and providing you with no sense of duration. In 1884 a fixed location of 0 degrees longitude was established at the royal observatory in Greenwich allowing location and hence time to be defined precisely. From then on things made sense.
In thinking about marketing metrics it seems we're still in the late 19th century with hundreds to choose from and consultants promising to help find the right ones. Finance went through similar throes and ended up with one or two with cash flow being on everyone's short list.
It might be time to think about the ones that really matter at the end of the day.
Wednesday, October 25, 2006
Two Strong Images: Luxury and Mud
Is this a brand too far?
The Lincoln Mark LT pickup truck, yes pick up, strikes me as a contradiction. To me Lincoln symbolizes luxury and a ride to a New York airport. The idea that I'd want a pickup truck (and a very big one at that) from the company I associate with entirely different set of emotional cues is pushing it.
Must be going after the suburban Hummer crowd.
The Lincoln Mark LT pickup truck, yes pick up, strikes me as a contradiction. To me Lincoln symbolizes luxury and a ride to a New York airport. The idea that I'd want a pickup truck (and a very big one at that) from the company I associate with entirely different set of emotional cues is pushing it.
Must be going after the suburban Hummer crowd.
Electronic Lists: End of the Data Card
Did Oscar Goldman have it right, can we rebuild a business?
The list business is shifting as aggregators provide access to 100s if not 1,000s of different lists. If they have all those lists I no longer want individual PDFs for each individual list, I want information to help me choose the combination of lists that satisfies my needs.
I want to drop one list on top of another to see the overlap and net-down. I also want to use my housefile as well. Lists with high duplicate rates may outperform other sources due to intrinsic factors not readily seen in ‘factoids.’
We need to separate the cost model from the need model. I'll pay for performance not volume.
PS – We have the technology and it won’t cost $6m
The list business is shifting as aggregators provide access to 100s if not 1,000s of different lists. If they have all those lists I no longer want individual PDFs for each individual list, I want information to help me choose the combination of lists that satisfies my needs.
I want to drop one list on top of another to see the overlap and net-down. I also want to use my housefile as well. Lists with high duplicate rates may outperform other sources due to intrinsic factors not readily seen in ‘factoids.’
We need to separate the cost model from the need model. I'll pay for performance not volume.
PS – We have the technology and it won’t cost $6m
Monday, October 23, 2006
Marketing's Three Rs Require Data
Why is information so important?
Originally, the three R's were Reading, 'riting and 'rithmetic --- although I'd have flunked with that spelling. Mila D'Antonio's blog reported the new mantra from the DMA:
All three are data centric concepts - one certainly can't determine results or relevance with out some objective measure and feedback. Responsibility is also data driven -- if we fail to live up to our side of the trust bargain customers will speak loudly with their voice, feet, and checkbook -- all things we can measure.
PS: Other uses of the R.
Reading, Writing, RFID in logistics.
Reduce, Replace, Refine in animal testing.
Reduce, Reuse, Recycle in the environment.
Originally, the three R's were Reading, 'riting and 'rithmetic --- although I'd have flunked with that spelling. Mila D'Antonio's blog reported the new mantra from the DMA:
....John Greco, president and CEO, of the Direct Marketing Association, called for widespread use of something he's coined the Three R's: Results, Responsibility, and Relevance. He said they represent a delicate balance in the power of direct marketing. 'Our goal is to build a bridge of trust with consumers,' he told the crowd.
All three are data centric concepts - one certainly can't determine results or relevance with out some objective measure and feedback. Responsibility is also data driven -- if we fail to live up to our side of the trust bargain customers will speak loudly with their voice, feet, and checkbook -- all things we can measure.
PS: Other uses of the R.
Reading, Writing, RFID in logistics.
Reduce, Replace, Refine in animal testing.
Reduce, Reuse, Recycle in the environment.
Marketing Database: Secret Sauce to Success
How do you collect Marketing's requirements?
The industry is filled with anecdotal stories (some are very true) about the difficulty in gathering marketing's requirements when building a marketing database.
Because marketing constantly tries new ideas, tests results, and compares one group to another traditional requirements around 'what reports would you like to see' simply fails to capture the essence of the job. Instead, successful implementations base the design of the marketing database on the concepts of events and context.
Events are those marketing and customer activities that are relevant to executing and understanding the process. For example, this would include 'Contact', 'Request', 'Registration' etc. across channels. Start by listing all the specific events from the beginning of the customer journey to end.
Context is the information that helps marketing understand the 'who, what, when, and why'. For each event list all of the possible questions that need to be answered to fully comprehend a specific event.
A successful design allows a marketer to easily understand and talk about the events and context when segmenting, profiling or selecting people for campaigns.
Don't stop the exercise because 'that data isn't available' --- this is a logical problem first and foremost. A good design allows data to be added over time as required without restructuring the system.
The industry is filled with anecdotal stories (some are very true) about the difficulty in gathering marketing's requirements when building a marketing database.
Because marketing constantly tries new ideas, tests results, and compares one group to another traditional requirements around 'what reports would you like to see' simply fails to capture the essence of the job. Instead, successful implementations base the design of the marketing database on the concepts of events and context.
Events are those marketing and customer activities that are relevant to executing and understanding the process. For example, this would include 'Contact', 'Request', 'Registration' etc. across channels. Start by listing all the specific events from the beginning of the customer journey to end.
Context is the information that helps marketing understand the 'who, what, when, and why'. For each event list all of the possible questions that need to be answered to fully comprehend a specific event.
A successful design allows a marketer to easily understand and talk about the events and context when segmenting, profiling or selecting people for campaigns.
Don't stop the exercise because 'that data isn't available' --- this is a logical problem first and foremost. A good design allows data to be added over time as required without restructuring the system.
Friday, October 20, 2006
The Math of ROI and the Budget Deficit
How can one question have two answers?
In this election season lots of candidates are talking about the deficit. And you can find two numbers circulating: $260 billion and $560 billion. The difference is obviously 'real money' according to the quote attributed to Senator Dirksen.
The argument goes something like this:
Deficit equals revenue less expenditures. So far, so good.
However revenue consists of tax collections plus borrowed funds. Since one would expect to repay a loan, or we're just robbing Peter to pay Paul, it should be accounted for in the figures.
When considering all sources of income, the deficit is $260 billion.
When considering borrowing as only a temporary source, it is $560 billion.
A similar scenario occurs in evaluating marketing programs. ROI is often portrayed as (Profit - Costs)/Costs. But profit should only be the incremental impact of the campaign since in most cases some customers would have bought anyway.
Focusing on the difference between actual and estimated sales has a long history in the syndicated market research world, it is less common in the direct marketing world. Something to work on.
In this election season lots of candidates are talking about the deficit. And you can find two numbers circulating: $260 billion and $560 billion. The difference is obviously 'real money' according to the quote attributed to Senator Dirksen.
The argument goes something like this:
Deficit equals revenue less expenditures. So far, so good.
However revenue consists of tax collections plus borrowed funds. Since one would expect to repay a loan, or we're just robbing Peter to pay Paul, it should be accounted for in the figures.
When considering all sources of income, the deficit is $260 billion.
When considering borrowing as only a temporary source, it is $560 billion.
A similar scenario occurs in evaluating marketing programs. ROI is often portrayed as (Profit - Costs)/Costs. But profit should only be the incremental impact of the campaign since in most cases some customers would have bought anyway.
Focusing on the difference between actual and estimated sales has a long history in the syndicated market research world, it is less common in the direct marketing world. Something to work on.
DMA Buzz
What got the excitement meter up at this year's show?
It seems that the consensus circulating a few days after the show is that there really wasn't any buzz on the exhibit floor this year. Could mean one of three things:
1. We're in the trough between two hype cycles,
2. We're so jaded by advertising that we've tuned it out, or
3. We're just getting on with doing the work.
It was unfortunate that the 'Interactive' hall was separated from the 'Multi-channel' hall. It made for some long walks.
It seems that the consensus circulating a few days after the show is that there really wasn't any buzz on the exhibit floor this year. Could mean one of three things:
1. We're in the trough between two hype cycles,
2. We're so jaded by advertising that we've tuned it out, or
3. We're just getting on with doing the work.
It was unfortunate that the 'Interactive' hall was separated from the 'Multi-channel' hall. It made for some long walks.
Thursday, October 19, 2006
Good Reading Material
Where can you find interesting reviews of marketing technology?
David Raab, now president of Client X Client, publishes his thoughts on products as well as an interesting view of industry white papers on his new blog.
Also, Kevin Hillstrom publishes thoughts on database marketing and multi-channel marketing on his blog.
Both are worth a read. Links to the left.
David Raab, now president of Client X Client, publishes his thoughts on products as well as an interesting view of industry white papers on his new blog.
Also, Kevin Hillstrom publishes thoughts on database marketing and multi-channel marketing on his blog.
Both are worth a read. Links to the left.
Totems, Parables and Stories
How do we learn?
One of the most effective ways of communicating is through stories. Most cultures, including organizations, teach through storytelling - be it verbal or visual, as in the case of totems.
Even when presenting the 'hard facts' we often turn them into vignettes. You know this is coming when listening to a presentation and the speaker follows up the numbers with "For example ...." I was once asked how large the forecast error was for a retail chain. The answer, at least in terms of metrics, was right there in terms of RMSE - but that didn't equate. The message that did work was: we're off by about one full shopping cart every six months. Technically the same metric, only the context changed.
It seems the secret to communicating metrics is to first convert them into everyday terms the audience understands.
One of the most effective ways of communicating is through stories. Most cultures, including organizations, teach through storytelling - be it verbal or visual, as in the case of totems.
Even when presenting the 'hard facts' we often turn them into vignettes. You know this is coming when listening to a presentation and the speaker follows up the numbers with "For example ...." I was once asked how large the forecast error was for a retail chain. The answer, at least in terms of metrics, was right there in terms of RMSE - but that didn't equate. The message that did work was: we're off by about one full shopping cart every six months. Technically the same metric, only the context changed.
It seems the secret to communicating metrics is to first convert them into everyday terms the audience understands.
Wednesday, October 18, 2006
Sampling: When less is good enough
Why do we use a subset?
We sample for one of three reasons: to prove, project, or probe.
Prove: provide credence for a given point of view or confirm a suspicion. Surveys are often done by companies or politicians to justify a position. Not bad in and of itself, but taken for what it is. They may include two steps: First, is something important and second are you doing something about it. Often known as the "gap" approach. In customer analysis it is often common to find a sample used as a matter of convenience for understanding behavior. Many companies have a sandbox for asking questions of the form: Do customers {fill in hunch here}?
Project: when you need the "true number" and asking everybody costs too much in terms of time or money. In this case care is taken in the sampling and weighting of that sample to ensure that the results mirror the larger population from which it is taken. Market share from panel data, media ratings, etc. are good examples. Daily political surveys may or may not be good.
Probe: when you need to understand how different groups behave or respond to marketing ideas. The compare and contrast approach rests on relative differences more than absolute numbers. The champion-challenger approach to message testing is a classic example.
Sampling is a powerful tool, but it is important to use the methods appropriate with the objective. For instance A/B testing does not need static samples and complex projection schemes.
We sample for one of three reasons: to prove, project, or probe.
Prove: provide credence for a given point of view or confirm a suspicion. Surveys are often done by companies or politicians to justify a position. Not bad in and of itself, but taken for what it is. They may include two steps: First, is something important and second are you doing something about it. Often known as the "gap" approach. In customer analysis it is often common to find a sample used as a matter of convenience for understanding behavior. Many companies have a sandbox for asking questions of the form: Do customers {fill in hunch here}?
Project: when you need the "true number" and asking everybody costs too much in terms of time or money. In this case care is taken in the sampling and weighting of that sample to ensure that the results mirror the larger population from which it is taken. Market share from panel data, media ratings, etc. are good examples. Daily political surveys may or may not be good.
Probe: when you need to understand how different groups behave or respond to marketing ideas. The compare and contrast approach rests on relative differences more than absolute numbers. The champion-challenger approach to message testing is a classic example.
Sampling is a powerful tool, but it is important to use the methods appropriate with the objective. For instance A/B testing does not need static samples and complex projection schemes.
Tuesday, October 17, 2006
This is why we test
In a synopsis of the DMA06 activity and other news, results of a new study on targeting was highlighted. The point is that until we try, we won't know.
Study Finds Customized Segmentation Beats Automated Targeting
Behaviorally targeted ads are more likely to generate higher conversion rates in customized market segments, compared to automated pre-packaged rules-based targeting, according to new study findings released by BL Labs of San Jose, CA. The study concludes that online advertising networks need to focus on custom segmentation and data analysis to optimize the nuances of behavioral targeting. Although online ads placed with contextually relevant content are likely to generate higher click through rates, ads shown with unrelated content typically pulled higher conversion rates. For example, the study found that advertising targeting travelers achieved the highest click through and conversion rates when placed on food Web sites.
The Envelope Please
Is the CMO is the new CIO
Ten to fifteen years ago the CIO had longevity similar to today’s CMO – about two years.
At that time there was a joke circulating around about what it took to succeed as a CIO.
A new CIO asks his predecessor for insight on how to be successful. The former CIO replies "when times get tough, open one of the 3 envelopes I left for you in the desk". Needless to say, times get rough so the CIO opens the first envelope and it says 'decentralize everything'. Things improve then degrade -- second envelope says 'centralize it all'. Things get better then fall apart -- third envelope says 'fill out three envelopes.'
Are branding and direct today's equivalents?
Like the IT dilemma, it is by no means an either or scenario -- but both. This point was repeatedly made at this years DMA.
Ten to fifteen years ago the CIO had longevity similar to today’s CMO – about two years.
At that time there was a joke circulating around about what it took to succeed as a CIO.
A new CIO asks his predecessor for insight on how to be successful. The former CIO replies "when times get tough, open one of the 3 envelopes I left for you in the desk". Needless to say, times get rough so the CIO opens the first envelope and it says 'decentralize everything'. Things improve then degrade -- second envelope says 'centralize it all'. Things get better then fall apart -- third envelope says 'fill out three envelopes.'
Are branding and direct today's equivalents?
Like the IT dilemma, it is by no means an either or scenario -- but both. This point was repeatedly made at this years DMA.
Monday, October 16, 2006
One to Many
Do you eat your own dog food?
It seems odd to me that one of the leading proponents of 1 to 1 marketing doesn't actually follow its own recommendations. I now get 3 copies of the print version of their magazine, even after emailing them with the necessary changes.
As they say there should be a way to manage one's own preferences.
It seems odd to me that one of the leading proponents of 1 to 1 marketing doesn't actually follow its own recommendations. I now get 3 copies of the print version of their magazine, even after emailing them with the necessary changes.
As they say there should be a way to manage one's own preferences.
Friday, October 13, 2006
Enterprise Platforms
Does one size fit all?
In CIO and CFO circles there are valid arguments to 'standardize on one enterprise platform.' But in the realm of Business Intelligence there are some subtle and not so subtle differences between various segments.
Consider the standard types of "Business Intelligence"
1. Management Intelligence -- are we running the business according to plan?
2. Operational Intelligence -- are the processes working correctly?
3. Customer Intelligence -- how does marketing impact customer behavior?
4. Market Intelligence -- where are we headed?
5. Competitive Intelligence -- how do we fare with respect to the other guys?
I'm still hard pressed to think that one platform can support all of those needs. Technology is versatile, not omnipotent.
In CIO and CFO circles there are valid arguments to 'standardize on one enterprise platform.' But in the realm of Business Intelligence there are some subtle and not so subtle differences between various segments.
Consider the standard types of "Business Intelligence"
1. Management Intelligence -- are we running the business according to plan?
2. Operational Intelligence -- are the processes working correctly?
3. Customer Intelligence -- how does marketing impact customer behavior?
4. Market Intelligence -- where are we headed?
5. Competitive Intelligence -- how do we fare with respect to the other guys?
I'm still hard pressed to think that one platform can support all of those needs. Technology is versatile, not omnipotent.
Data Driven Marketing
What is coming down the road?
A little while ago I wrote an article on things I'd like to see more visible in marketing analytics.
In summary:
1. Integrate survey and other syndicated data with house and prospect files. This type of augmentation takes the concept of overlay into lightly-chartered territory.
2. Merge marketing mix modeling with database marketing to get an end-to-end allocation of resources and people.
3. Shorten the path between analysis / reporting to campaign execution by developing truly integrated tools.
4. Continue focusing and measuring the influence of marketing spend on specific individual customer behaviors.
A little while ago I wrote an article on things I'd like to see more visible in marketing analytics.
In summary:
1. Integrate survey and other syndicated data with house and prospect files. This type of augmentation takes the concept of overlay into lightly-chartered territory.
2. Merge marketing mix modeling with database marketing to get an end-to-end allocation of resources and people.
3. Shorten the path between analysis / reporting to campaign execution by developing truly integrated tools.
4. Continue focusing and measuring the influence of marketing spend on specific individual customer behaviors.
Thursday, October 12, 2006
Free Market Analytics
Recent announcement: Netflix Offers $1 Million Prize For Improved System
The objective is to improve ratings by a minimum of 10%. To help you Netflix will provide a sample (100 million records, so don't be looking at Excel). The objective is to close the gap between 'predictied rating' and 'actual rating.'
At last count there were 9,000 teams and over 150 entries to date. Nobody's close yet.
The objective is to improve ratings by a minimum of 10%. To help you Netflix will provide a sample (100 million records, so don't be looking at Excel). The objective is to close the gap between 'predictied rating' and 'actual rating.'
At last count there were 9,000 teams and over 150 entries to date. Nobody's close yet.
Lessons from Big Oil
When is your brand's Hubbert's Peak?
In any market there comes a time when production peaks and then begins to decline. In the global petroleum business Peak Oil is the date when production starts to decrease. Early in the life cycle production increases as infrastructure is added. Later it falls as resources are depleted.
Some time soon, the oil category will peak.
Looks similar to any product lifecycle chart I've seen.
At some point, product sales slip into decline.
Now distribution represents infrastructure.
And demand represents resources.
The difference between brands is often how long the 'maturity' phase lasts. In the case of oil the peak is a singular moment in time. For a lot of products it is often a flat line for some period.
The point remains, at some point we need to plan for product obsolescence: How do we manage relationships with customers? When the firm IS the brand, this gets a bit tricky with the short-term focus of financial markets.
In any market there comes a time when production peaks and then begins to decline. In the global petroleum business Peak Oil is the date when production starts to decrease. Early in the life cycle production increases as infrastructure is added. Later it falls as resources are depleted.
Some time soon, the oil category will peak.
Looks similar to any product lifecycle chart I've seen.
At some point, product sales slip into decline.
Now distribution represents infrastructure.
And demand represents resources.
The difference between brands is often how long the 'maturity' phase lasts. In the case of oil the peak is a singular moment in time. For a lot of products it is often a flat line for some period.
The point remains, at some point we need to plan for product obsolescence: How do we manage relationships with customers? When the firm IS the brand, this gets a bit tricky with the short-term focus of financial markets.
Fluid Segments
Why do we need leaky buckets?
Most approaches to marketing base execution on segmenting customers or prospects. So far so good.
Using a top-down, hierarchical system similar to the Dewey Decimal system for books requires that each and every customer be in one and only one slot. This requirement rests on the ability of some central authority to make that decision. It can result in some quirks: Books on office productivity, e.g. "Word" are in both the 000s (Generalities) and five stacks away in the 600s (Technology).
However, segmentation is fluid with respect to both time and objective. Customers change and our reason to communicate changes.
A more natural tool for segmentation may be the concept of 'tagging'. Joshua Schacter, founder of Del.icio.us and MIT Technology Review's "Innovator of the Year" has shown that using user-defined categorization (segmentation) ends up being a pretty good way of organizing the web. At any point in time a Web page may have any number of descriptors as a way of improving memory and recall.
Since a customer can be both "High Value" AND "Vulnerable" at the same time; additional flexibility in segmentation schemes is required and tagging makes an interesting choice.
Most approaches to marketing base execution on segmenting customers or prospects. So far so good.
Using a top-down, hierarchical system similar to the Dewey Decimal system for books requires that each and every customer be in one and only one slot. This requirement rests on the ability of some central authority to make that decision. It can result in some quirks: Books on office productivity, e.g. "Word" are in both the 000s (Generalities) and five stacks away in the 600s (Technology).
However, segmentation is fluid with respect to both time and objective. Customers change and our reason to communicate changes.
A more natural tool for segmentation may be the concept of 'tagging'. Joshua Schacter, founder of Del.icio.us and MIT Technology Review's "Innovator of the Year" has shown that using user-defined categorization (segmentation) ends up being a pretty good way of organizing the web. At any point in time a Web page may have any number of descriptors as a way of improving memory and recall.
Since a customer can be both "High Value" AND "Vulnerable" at the same time; additional flexibility in segmentation schemes is required and tagging makes an interesting choice.
Wednesday, October 11, 2006
Remarkable Analytics
Why do we need to analyze?
Seth Godin summarizes it well in The Purple Cow: "Marketing isn't guaranteed."
In the face of that fact there are three paths we could take regarding measuring marketing's impact. First, we could stick our heads in the ground and not worry about it. Second, we could throw spaghetti on the wall and see what sticks-- or is that sales? Third, we can try to figure out what actually works.
Some simple logic suggests option three is the only viable route. It's budget season so I'm willing to bet most firms are going for a growth rate that exceeds the category average because 'we're better'. Now what is in short supply when each individual firm estimates growth greater than the overall market? Customers. They are the scarce resource So, achieving our objectives must come at the expense of the other guys. That means our analytics have to be sharper, quicker, and more in tune with customer needs. Or in the vernacular of the Cow -- "remarkable."
Seth Godin summarizes it well in The Purple Cow: "Marketing isn't guaranteed."
In the face of that fact there are three paths we could take regarding measuring marketing's impact. First, we could stick our heads in the ground and not worry about it. Second, we could throw spaghetti on the wall and see what sticks-- or is that sales? Third, we can try to figure out what actually works.
Some simple logic suggests option three is the only viable route. It's budget season so I'm willing to bet most firms are going for a growth rate that exceeds the category average because 'we're better'. Now what is in short supply when each individual firm estimates growth greater than the overall market? Customers. They are the scarce resource So, achieving our objectives must come at the expense of the other guys. That means our analytics have to be sharper, quicker, and more in tune with customer needs. Or in the vernacular of the Cow -- "remarkable."
Combinatorial Complexity
Are there too many good ideas to test?
We often have lots of ideas that we know we should test, but sequentially going through the list would take forever and a day. The good news is that this problem is readily solvable using something called experimental design.
On a replay of a Direct webinar on multi-variate email testing Katie Cole of MerkleQuiris addressed this issue and stated something like “our account managers love to say fractional factorial in a sales meeting.”
Obviously alliteration; but what does it mean?
When designing a campaign we can use any number of factors – creative, list, segment, offer, etc. each with any number of variations. If we want to test this whole lot we need to consider ‘multivariate testing.’ Note that the old A/B test is the simplest case: One factor with two variations. When lots of factors are involved the number of possible combinations is simply the product of the number of variations for each factor.
For example: If an email test has 3 subject lines, 5 lists, and 4 segments and we evaluate all 60 possible (3 * 5 * 4) versions we’d be doing a ‘full factorial’ test.
At some point there are too many combinations to test. Imagine adding text vs. html, three wire-frame layouts, ads on left or right side, three offers located in four locations. Now we have over 14,000 combinations! This is where ‘fractional factorial’ comes in. It is a means of systematically reducing the number of combinations to a manageable number but still retaining the ability to answer some, but not all possible questions.
The smallest magic subset of combinations allows us to test the fundamental questions about each of the major factors:
- Does subject line matter?
- Does list impact response?
- Do segments behave differently?
- etc.
As more combinations are added to the mix we can ask questions with nuance.
- Does the choice of list impact how the subject line works?
- Does format affect the click-through rate of ad location?
The only way to answer all possible questions is to test all possible combinations – full factorial. However, in the majority of cases we’re not looking to understand the subtleties of how each factor interacts with each other. We want to know the major levers to customer behavior.
Like triage, ‘fractional factorial’ focuses on the important cases and allows us to assess the big questions of campaign design. Once we’ve identified the factors that drive behavior we’ll work on fine tuning the variations.
We often have lots of ideas that we know we should test, but sequentially going through the list would take forever and a day. The good news is that this problem is readily solvable using something called experimental design.
On a replay of a Direct webinar on multi-variate email testing Katie Cole of MerkleQuiris addressed this issue and stated something like “our account managers love to say fractional factorial in a sales meeting.”
Obviously alliteration; but what does it mean?
When designing a campaign we can use any number of factors – creative, list, segment, offer, etc. each with any number of variations. If we want to test this whole lot we need to consider ‘multivariate testing.’ Note that the old A/B test is the simplest case: One factor with two variations. When lots of factors are involved the number of possible combinations is simply the product of the number of variations for each factor.
For example: If an email test has 3 subject lines, 5 lists, and 4 segments and we evaluate all 60 possible (3 * 5 * 4) versions we’d be doing a ‘full factorial’ test.
At some point there are too many combinations to test. Imagine adding text vs. html, three wire-frame layouts, ads on left or right side, three offers located in four locations. Now we have over 14,000 combinations! This is where ‘fractional factorial’ comes in. It is a means of systematically reducing the number of combinations to a manageable number but still retaining the ability to answer some, but not all possible questions.
The smallest magic subset of combinations allows us to test the fundamental questions about each of the major factors:
- Does subject line matter?
- Does list impact response?
- Do segments behave differently?
- etc.
As more combinations are added to the mix we can ask questions with nuance.
- Does the choice of list impact how the subject line works?
- Does format affect the click-through rate of ad location?
The only way to answer all possible questions is to test all possible combinations – full factorial. However, in the majority of cases we’re not looking to understand the subtleties of how each factor interacts with each other. We want to know the major levers to customer behavior.
Like triage, ‘fractional factorial’ focuses on the important cases and allows us to assess the big questions of campaign design. Once we’ve identified the factors that drive behavior we’ll work on fine tuning the variations.
Tuesday, October 10, 2006
Cross Category Analysis
How do I identify products for acquisition or cross-sell strategies?
I want to see the impact of each and every product (SKU) in my portfolio in terms of its ability to generate incremental penetration. I don't want to look within a category, but across them all.
Reason: Products that attract customers should be marketed differently than those that reinforce the decision.
I want to see the impact of each and every product (SKU) in my portfolio in terms of its ability to generate incremental penetration. I don't want to look within a category, but across them all.
Reason: Products that attract customers should be marketed differently than those that reinforce the decision.
Role of Fly Fishing in Marketing
What does feather tossing have to do with marketing?
In a word, image.
Since the mid-nineteenth century fly fishing has been at the center of image marketing.
First there was Cornelia "Fly Rod" Crosby the women who marketed Maine as the 'playground of the nation' where visitors could experience the 'wilderness with all the comforts.' Now the companies behind this campaign were the railroads since that was the only way someone could get to the backwoods. They sponsored casting clinics in Central Park to entice people to travel to Maine.
Hiram Leonard, Eustis Edwards, and a few others became 'signature brands' as makers of bamboo fly rods. As they became popular, it wasn't unusual to partner or be acquired by more 'business-focused' firms who needed to figure out how to go from a couple of dozen rods per year to 1,000+. The impact of technology, e.g. beveling machines of the late 1870s, on a craft business makes for interesting reading in "Casting a Spell" by George Black.
Orvis, which started issuing catalogs prior to the Civil War, focused on the total image starting in the mid-1960s. Leigh Perkins, who had bought Orvis in 1965 described the company as 'what we were creating and selling was, for a lack of better word, a lifestyle, a kind of Americanized version of elegant, English country living.' For several years Orvis swapped house files with the likes of Abercrombie & Fitch and L.L. Bean as part of its direct marketing program.
Robert Redford's "A River Runs Through It" based on Norman Maclean's novella is credited with spawning (or ruining) even more interest in fly-fishing. Since the movie's debut in 1992 western trout streams have seen a lot of 'image pressure' - fishermen (and women) who look good but don't catch much.
In a word, image.
Since the mid-nineteenth century fly fishing has been at the center of image marketing.
First there was Cornelia "Fly Rod" Crosby the women who marketed Maine as the 'playground of the nation' where visitors could experience the 'wilderness with all the comforts.' Now the companies behind this campaign were the railroads since that was the only way someone could get to the backwoods. They sponsored casting clinics in Central Park to entice people to travel to Maine.
Hiram Leonard, Eustis Edwards, and a few others became 'signature brands' as makers of bamboo fly rods. As they became popular, it wasn't unusual to partner or be acquired by more 'business-focused' firms who needed to figure out how to go from a couple of dozen rods per year to 1,000+. The impact of technology, e.g. beveling machines of the late 1870s, on a craft business makes for interesting reading in "Casting a Spell" by George Black.
Orvis, which started issuing catalogs prior to the Civil War, focused on the total image starting in the mid-1960s. Leigh Perkins, who had bought Orvis in 1965 described the company as 'what we were creating and selling was, for a lack of better word, a lifestyle, a kind of Americanized version of elegant, English country living.' For several years Orvis swapped house files with the likes of Abercrombie & Fitch and L.L. Bean as part of its direct marketing program.
Robert Redford's "A River Runs Through It" based on Norman Maclean's novella is credited with spawning (or ruining) even more interest in fly-fishing. Since the movie's debut in 1992 western trout streams have seen a lot of 'image pressure' - fishermen (and women) who look good but don't catch much.
Monday, October 09, 2006
Making it Easy to Sell Makes it Hard to Buy
At what point does choice become too much?
We’ve all been taught that being a leader in a category is critical to success. Just search on the phrase “industry leader” on any job-posting site to see how far this has been taken.
The argument for categorization is usually one of two: People remember leaders and/or people buy from leaders. Since we seem to have an innate ability to remember ‘the first’ in an infinite number of categories it makes sense to be #1 in something. Recalling brand #2 is difficult; Avis not withstanding.
But at some point does this fragmentation work against us? Is there a point at which the resulting ‘hyper-choice’ turns us off?
Personally – yes. I have not bought, when I thought I wanted to, because of too many options and no simple way of paring down the choices to a relevant set. It also appears to be real in the grocery store as well. Fewer options on the shelf result in higher sales.
Relevance applies to the set of choices as much as it does to any other “P”. While narrowing options is much easier to do on-line via search and tagging than it is on a shelf, it is still too feature driven (4 mega-pixel vs. 5 mega-pixel as if I care or even want to know). It is still rare to find tools that are situation or need driven. For a cool one, see www.noodletools.com.
Identifying segments based on need will require profiling, quite possibly in the Quantico sense.
We’ve all been taught that being a leader in a category is critical to success. Just search on the phrase “industry leader” on any job-posting site to see how far this has been taken.
The argument for categorization is usually one of two: People remember leaders and/or people buy from leaders. Since we seem to have an innate ability to remember ‘the first’ in an infinite number of categories it makes sense to be #1 in something. Recalling brand #2 is difficult; Avis not withstanding.
But at some point does this fragmentation work against us? Is there a point at which the resulting ‘hyper-choice’ turns us off?
Personally – yes. I have not bought, when I thought I wanted to, because of too many options and no simple way of paring down the choices to a relevant set. It also appears to be real in the grocery store as well. Fewer options on the shelf result in higher sales.
Relevance applies to the set of choices as much as it does to any other “P”. While narrowing options is much easier to do on-line via search and tagging than it is on a shelf, it is still too feature driven (4 mega-pixel vs. 5 mega-pixel as if I care or even want to know). It is still rare to find tools that are situation or need driven. For a cool one, see www.noodletools.com.
Identifying segments based on need will require profiling, quite possibly in the Quantico sense.
Significance vs. Substance
Does your analysis have 'oomph'?
Apparently this is technical term in economic circles -- it summarizes the idea that 'statistical significance' is not what analysis is about. Rather the focus should be on How Much of a difference something makes not Whether it makes a difference.
Deirdre McCloskey makes this point in Secret Sins of Economics.
In short, statistical significance is neither necessary nor sufficient for a result to be scientifically significant. Most of the time it is irrelevant. A researcher is simply committing a scientific error to use it as ….an all-purpose way of judging whether a number is large enough to matter. Mattering is a human matter; the numbers figure, but after collecting them the mattering has to be decided finally by us; mattering does not inhere in a number.
The point is that we should ask: How much impact does {insert variable here} have on customer behavior?
For marketing the question isn't whether something is significant but rather whether it has substance - or oomph, a term that fits well in the creative vernacular.
Apparently this is technical term in economic circles -- it summarizes the idea that 'statistical significance' is not what analysis is about. Rather the focus should be on How Much of a difference something makes not Whether it makes a difference.
Deirdre McCloskey makes this point in Secret Sins of Economics.
In short, statistical significance is neither necessary nor sufficient for a result to be scientifically significant. Most of the time it is irrelevant. A researcher is simply committing a scientific error to use it as ….an all-purpose way of judging whether a number is large enough to matter. Mattering is a human matter; the numbers figure, but after collecting them the mattering has to be decided finally by us; mattering does not inhere in a number.
The point is that we should ask: How much impact does {insert variable here} have on customer behavior?
For marketing the question isn't whether something is significant but rather whether it has substance - or oomph, a term that fits well in the creative vernacular.
Friday, October 06, 2006
Marketers are Unreasonable
The reasonable man adapts himself to the world: the unreasonable one persists in trying to adapt the world to himself. Therefore all progress depends on the unreasonable man.
GEORGE BERNARD SHAW, Maxims for Revolutionists
Marketers are the one function in the organization charged with changing the status quo. They alone are responsible for new products, new markets, new segments, and new revenue.
They define a vision of what should be and then they test, tinker, and tweak how things work in order to alter customer behavior to get there.
So a message to IT: Please forgive their 'unreasonableness' for they are paid to be so.
Overlapping Networks
Is MySpace where the LinkedIn crowd hang out after work?
It has certainly has been stated that LinkedIn is for adults. But like any two products in a category, they each have there own persona.
- LinkedIn makes its money capitalizing on the age-old adage: "It's not what you know, but who you know."
- MySpace makes its money reaching a key demographic: "It's not what you know, but who you are."
Originally started with executives in mind, LinkedIn has moved 'down-market' as it grows to an estimated 10 million users by year end. Originally started for adults (but not adult content), MySpace has proven attractive to younger audiences with its 100 millionth account created this summer.
The odds of overlap are pretty small in absolute numbers (no more than 10% of MySpace could have a LinkedIn profile.) Of my extended network of 1mm+ LinkedIn users the phrase "MySpace" returned only a couple of hundred people. Now I'll admit many may not cross-post in order to separate their after-hours life from their day job.
[UPDATE] - Age profile of MySpace getting older. Half of the users are over 35, up from nearly 40% last year. So, is the brand diluting?
Thursday, October 05, 2006
Pulling the nth Degree
Is nth-selection random?
Recently asked the above, and like all questions these days it seems the answer is 'it depends'.
Picking every 4th, 10th, 100th etc. name as a way of sampling is often described as 'random'. Two practical scenarios make this unlikely.
Say I need 10,000 names for a test from a house-file of 10,000,000. If I pick every 10th or even 100th name I will only go through a small portion of the whole list. So, if that list is sequenced by anything meaningful like age of account, last name, value, etc. I end up with a biased list. And most database administers will order a file by some attribute to speed load, linking, and indexing.
When doing repeated pulls over time nth selections may result in the same list being pulled. Changing the starting point or randomly sorting before the pull minimizes this risk.
Random assumes that every record has an equal chance of being selected -- while nth is a means of sampling, it is by no means random. Once the first record from a list is flagged the chance of pulling any other record is defined.
Whether this matters a hill of beans or not depends a lot on why one samples in the first place. (More on that soon.)
Recently asked the above, and like all questions these days it seems the answer is 'it depends'.
Picking every 4th, 10th, 100th etc. name as a way of sampling is often described as 'random'. Two practical scenarios make this unlikely.
Say I need 10,000 names for a test from a house-file of 10,000,000. If I pick every 10th or even 100th name I will only go through a small portion of the whole list. So, if that list is sequenced by anything meaningful like age of account, last name, value, etc. I end up with a biased list. And most database administers will order a file by some attribute to speed load, linking, and indexing.
When doing repeated pulls over time nth selections may result in the same list being pulled. Changing the starting point or randomly sorting before the pull minimizes this risk.
Random assumes that every record has an equal chance of being selected -- while nth is a means of sampling, it is by no means random. Once the first record from a list is flagged the chance of pulling any other record is defined.
Whether this matters a hill of beans or not depends a lot on why one samples in the first place. (More on that soon.)
Marketing Hiring Practices
Have we closed the skills gap?
The need for analytic and technical skills has been argued for a while now.
Time - late 1998: In a CIO Magazine article entitled "Improving the Odds"
.... marketing takes a dominant role in shaping organizations' interactions with consumers. Marketing thus becomes the company's darling and the information systems (IS) department's new best friend. ... To rise to the challenge of leading the rest of the company into the new customer-centered paradigm, marketing must get itself in shape.
Time - early 2005: CMO Magazine (now disappeared) but cache is available. In an Editorial entitled "Eat Your Own Dog Food"..
What can Brown do for you? I don't know this Brown dude, but if he's good with either numbers or technology, hire him. CMOs need to expand the skill sets of their personnel if they want to increase the profile and the impact of marketing.
Time - this week: In Intelligent Enterprise on the topic of analytics (web in particular) :
The level of sophistication in using analytics tools runs the gamut.... A big part of the difference has to do with the way people think and their corporate culture. .... Left Brains Needed
We've made progress, but apparently not enough.
The need for analytic and technical skills has been argued for a while now.
Time - late 1998: In a CIO Magazine article entitled "Improving the Odds"
.... marketing takes a dominant role in shaping organizations' interactions with consumers. Marketing thus becomes the company's darling and the information systems (IS) department's new best friend. ... To rise to the challenge of leading the rest of the company into the new customer-centered paradigm, marketing must get itself in shape.
Time - early 2005: CMO Magazine (now disappeared) but cache is available. In an Editorial entitled "Eat Your Own Dog Food"..
What can Brown do for you? I don't know this Brown dude, but if he's good with either numbers or technology, hire him. CMOs need to expand the skill sets of their personnel if they want to increase the profile and the impact of marketing.
Time - this week: In Intelligent Enterprise on the topic of analytics (web in particular) :
The level of sophistication in using analytics tools runs the gamut.... A big part of the difference has to do with the way people think and their corporate culture. .... Left Brains Needed
We've made progress, but apparently not enough.
Wednesday, October 04, 2006
Process is the New Black
What's in fashion these days?
On the runway some color, fabric or style becomes the 'new black' -- that indispensable item that is the basis of any wardrobe.
In the realm of analytics there is a new kid in town: "Process"
In "Competing on Analytics" Tom Davenport argues "business processes are among the last remaining points of differentiation" and that success comes from using analytics to "wring every last drop of value..."
In "An Analytics Manifesto" Neil Raden lays out arguments for an architecture that "enables vastly smarter business processes."
Both are saying that "analysis" should be pervasive and the foundation of any business wardrobe. I agree, particularly since analysis is all about understanding what's going on and then doing something about it. Davenport often uses a core, high-end predictive modeling group argument; Raden argues for an embedded approach. In any complex business, and aren't they all, both strategies are appropriate. The deciding factor of which approach is right depends on the risks of being wrong. Ok, this is beginning to sound like a Head vs. Long Tail of analytics argument.
It's also noteworthy to look at who sponsored the research: SAS in the case of "Competing" and SAP in the case of "Manifesto". Now, neither company is known for the contra position, i.e. nobody would implement SAS for resource planning and they wouldn't use SAP for advanced statistics.
The positioning aside, both articles point to the conclusion that this is different than how 'business intelligence' tools approach the problem. SAP in particular is positioning process-embedded analytics for Everyman. Their argument is based in part on the fact that there are 100 times more Excel users than all Business Intelligence tools combined.
If I were a cynic I'd argue: The BI vendors had 15-25 years to get it right, and didn't. So let's try another route. Given that SAP BW isn't highly regarded as user friendly, that's not much of a stretch.
Back on the serious side: providing people with the events, context, alternatives and potential impact necessarily for their roles is a very good thing.
On the runway some color, fabric or style becomes the 'new black' -- that indispensable item that is the basis of any wardrobe.
In the realm of analytics there is a new kid in town: "Process"
In "Competing on Analytics" Tom Davenport argues "business processes are among the last remaining points of differentiation" and that success comes from using analytics to "wring every last drop of value..."
In "An Analytics Manifesto" Neil Raden lays out arguments for an architecture that "enables vastly smarter business processes."
Both are saying that "analysis" should be pervasive and the foundation of any business wardrobe. I agree, particularly since analysis is all about understanding what's going on and then doing something about it. Davenport often uses a core, high-end predictive modeling group argument; Raden argues for an embedded approach. In any complex business, and aren't they all, both strategies are appropriate. The deciding factor of which approach is right depends on the risks of being wrong. Ok, this is beginning to sound like a Head vs. Long Tail of analytics argument.
It's also noteworthy to look at who sponsored the research: SAS in the case of "Competing" and SAP in the case of "Manifesto". Now, neither company is known for the contra position, i.e. nobody would implement SAS for resource planning and they wouldn't use SAP for advanced statistics.
The positioning aside, both articles point to the conclusion that this is different than how 'business intelligence' tools approach the problem. SAP in particular is positioning process-embedded analytics for Everyman. Their argument is based in part on the fact that there are 100 times more Excel users than all Business Intelligence tools combined.
If I were a cynic I'd argue: The BI vendors had 15-25 years to get it right, and didn't. So let's try another route. Given that SAP BW isn't highly regarded as user friendly, that's not much of a stretch.
Back on the serious side: providing people with the events, context, alternatives and potential impact necessarily for their roles is a very good thing.
Do not test creative, assess the results
Putting emphasis on what is important.
There is a world of difference between 'the creative isn't good' and 'we didn't meet objectives'. The purpose of testing is to figure out what works in terms of shaping and capturing the desired customer behavior. Thus, all tests should explicitly, and before the fact, define what the goal is and how it will be measured.
Kind of like the scientific method.
There is a world of difference between 'the creative isn't good' and 'we didn't meet objectives'. The purpose of testing is to figure out what works in terms of shaping and capturing the desired customer behavior. Thus, all tests should explicitly, and before the fact, define what the goal is and how it will be measured.
Kind of like the scientific method.
Moneyball: The ultimate in being data-driven
Why is this book important to marketing?
Usual reasons include:
Illustrative use of information to run a business; it does.
A good read anyway; it is.
Key point: Measure that which causes success. It isn't always the stats (dare I say Key Performance Indicators) of conventional wisdom that relate to success. For instance while 'batting average' may raise the payroll it has little to do with understanding whether a team produces more runs than its opponent. Unheralded metrics like 'base-on-balls' do better.
It is a story about using the specific metrics that relate to game-day performance, as ultimately measured by W-L, to guide investment in players. Its even spawned its own discipline "sabermetrics" loosely defined as the objective search for what works. So when building a Marketing Box Score, make sure to include those metrics that actually relate to the objective --- financial performance.
Usual reasons include:Illustrative use of information to run a business; it does.
A good read anyway; it is.
Key point: Measure that which causes success. It isn't always the stats (dare I say Key Performance Indicators) of conventional wisdom that relate to success. For instance while 'batting average' may raise the payroll it has little to do with understanding whether a team produces more runs than its opponent. Unheralded metrics like 'base-on-balls' do better.
It is a story about using the specific metrics that relate to game-day performance, as ultimately measured by W-L, to guide investment in players. Its even spawned its own discipline "sabermetrics" loosely defined as the objective search for what works. So when building a Marketing Box Score, make sure to include those metrics that actually relate to the objective --- financial performance.
Tuesday, October 03, 2006
Customer Capacity and Headroom
How much can a customer consume?
Recent article I wrote on using capacity as a means of targeting can be found on Chief Marketer.
The main point: Customers have value to the firm as well as an overall capacity to consume. The difference between the two, e.g. headroom, makes Life Time Value a two-sided question. How much do I satisfy their needs? How large is their total need? This can be summarized in the following matrix:
Based on the value and headroom, different segments emerge - each with their own strategy.
Even 'high-value' customers can be segmented into those who are sated and those with lots more room.
Recent article I wrote on using capacity as a means of targeting can be found on Chief Marketer.
The main point: Customers have value to the firm as well as an overall capacity to consume. The difference between the two, e.g. headroom, makes Life Time Value a two-sided question. How much do I satisfy their needs? How large is their total need? This can be summarized in the following matrix:
Based on the value and headroom, different segments emerge - each with their own strategy.Even 'high-value' customers can be segmented into those who are sated and those with lots more room.
Yes, No,or Something In Between
When is an average wrong?
Most events marketing is concerned about are binary -- that is they happen or they don't. Buy, cancel, respond, churn, request, register, and click are all examples of a behavior that either occurs or it doesn't. Given that it is probably impossible and certainly impractical to categorize people into one of the two camps for future campaigns we've come up with some ways of estimating the probability that a customer or prospect will do something. Now we have values that look continuous from 0 to 1.
However, there are risks in interpretation when doing numeric gymnastics.
When reporting this likelihood with an assigned value to it, e.g. estimated sales revenue, strange numbers can occur. For instance, consider an email campaign where responders will spend $100 and non-responders will spend $0. The average may be $50 but the expected value isn't that at all if conversion rates are anywhere close to typical. The likely contribution for a 1,000 mailing with a 2% conversion rate is $2,000 (1,000 * 2% * $100) not $50,000 (1,000 * $50).
And yes, these simple errors are made. And they do impact decision making.
A widely publicized example comes from government circles. In 2000, the USGS published a tome about the status of known oil fields and the likelihood of finding more. Using the math similar to the email campaign it was concluded that there is a lot of oil to be found in Greenland.
Most events marketing is concerned about are binary -- that is they happen or they don't. Buy, cancel, respond, churn, request, register, and click are all examples of a behavior that either occurs or it doesn't. Given that it is probably impossible and certainly impractical to categorize people into one of the two camps for future campaigns we've come up with some ways of estimating the probability that a customer or prospect will do something. Now we have values that look continuous from 0 to 1.
However, there are risks in interpretation when doing numeric gymnastics.
When reporting this likelihood with an assigned value to it, e.g. estimated sales revenue, strange numbers can occur. For instance, consider an email campaign where responders will spend $100 and non-responders will spend $0. The average may be $50 but the expected value isn't that at all if conversion rates are anywhere close to typical. The likely contribution for a 1,000 mailing with a 2% conversion rate is $2,000 (1,000 * 2% * $100) not $50,000 (1,000 * $50).
And yes, these simple errors are made. And they do impact decision making.
A widely publicized example comes from government circles. In 2000, the USGS published a tome about the status of known oil fields and the likelihood of finding more. Using the math similar to the email campaign it was concluded that there is a lot of oil to be found in Greenland.
- 5% chance of finding 112 billion barrels
- 95% chance of finding 1 barrel
- Average reserve is reported as over 47 billion barrels
With that figure, which is twice the size of the estimated reserves of the US before the recent find in the Gulf, a lot of attention has been paid to an area that has produced little to date.
Using the email logic, the expected value should probably be somewhere in the 5 billion range, or the size of the United Kingdom. Or less than a year's supply.
Monday, October 02, 2006
Search, Shopping and Trademarks
How do people approach the search problem?
Hitwise just released the dog-days of summer shopping activity. But are all the product searches for the actual term used?
Searching is all about figuring out how best to reduce the clutter to find what's relevant. And at present we are more or less limited to using the text box without any help. Sometimes broad categories work, like those in the 'generic column'. Other times specific terms help constrain the results better. There are times when a brand name is the best tool for finding associated content. This is akin to using collaborative filtering and entering a book you already own for the sole purpose of finding other suggestions.
So how many people were searching in the general sense as opposed to specifically looking for barbie or ipod? Can this distinction be used to drive traffic?
Now what happens when somebody pays money for a key word that is a trademark owned by another. Is that legal? Appears to be. Google had a case against them dismissed as to whether competitor trademarks could be used in keyword searches. The argument (and I'm no lawyer) is that infringements occur when the trademark is 'use[d] in commerce' meaning some public display, e.g. on a product, in advertising etc., that a consumer can see. Apparently the internal machinations of ad-serving and page ranking don't qualify as 'use in commerce', although they do generate a lot of revenue. For a discussion of the use of a trademark in search for see Eric Goldman's blog and article on the topic.
BTW: I wonder what the age profile is of a 'generic' vs. 'product' searcher looks like? Is there a skew in how people search when shopping?
And yes there 100+ 'lingerie barbie' results on eBay.
Hitwise just released the dog-days of summer shopping activity. But are all the product searches for the actual term used?
Searching is all about figuring out how best to reduce the clutter to find what's relevant. And at present we are more or less limited to using the text box without any help. Sometimes broad categories work, like those in the 'generic column'. Other times specific terms help constrain the results better. There are times when a brand name is the best tool for finding associated content. This is akin to using collaborative filtering and entering a book you already own for the sole purpose of finding other suggestions.
So how many people were searching in the general sense as opposed to specifically looking for barbie or ipod? Can this distinction be used to drive traffic?
Now what happens when somebody pays money for a key word that is a trademark owned by another. Is that legal? Appears to be. Google had a case against them dismissed as to whether competitor trademarks could be used in keyword searches. The argument (and I'm no lawyer) is that infringements occur when the trademark is 'use[d] in commerce' meaning some public display, e.g. on a product, in advertising etc., that a consumer can see. Apparently the internal machinations of ad-serving and page ranking don't qualify as 'use in commerce', although they do generate a lot of revenue. For a discussion of the use of a trademark in search for see Eric Goldman's blog and article on the topic.
BTW: I wonder what the age profile is of a 'generic' vs. 'product' searcher looks like? Is there a skew in how people search when shopping?
And yes there 100+ 'lingerie barbie' results on eBay.
Pacifier for 140 mph
When is enough too much?
NASCAR is moving the way Harley-Davidson did with licensing and brand extensions that make little sense relative to the brand itself.
Last year Harley-Davidson a 'cake-decorating kit' was voted as being the worst extension in a Tipping Sprung, Brandweek survey. To give them credit the company won the year before with a foot-wear extension. I can imagine footwear that is 'rebellious and edgy'; I have a hard time with a sheet-cake being described that way.
Now NASCAR offers baby pacifers with the number of a favorite driver -- doesn't have the lead brand that's emblazoned on uniforms and cars, which is often an adult-focused co-brand like 'beer'.
On a morning news show today an executive said something along the lines of ‘If we get them young, we’ll keep them old.’ Sounds like the old tobacco and liquor marketing. While NASCAR generates $2 billion in merchandise sales and boasts 75 million fans; this kind of extension is just a bit too much.
Besides, I still prefer racing that makes right turns.
NASCAR is moving the way Harley-Davidson did with licensing and brand extensions that make little sense relative to the brand itself.
Last year Harley-Davidson a 'cake-decorating kit' was voted as being the worst extension in a Tipping Sprung, Brandweek survey. To give them credit the company won the year before with a foot-wear extension. I can imagine footwear that is 'rebellious and edgy'; I have a hard time with a sheet-cake being described that way.
Now NASCAR offers baby pacifers with the number of a favorite driver -- doesn't have the lead brand that's emblazoned on uniforms and cars, which is often an adult-focused co-brand like 'beer'.
On a morning news show today an executive said something along the lines of ‘If we get them young, we’ll keep them old.’ Sounds like the old tobacco and liquor marketing. While NASCAR generates $2 billion in merchandise sales and boasts 75 million fans; this kind of extension is just a bit too much.
Besides, I still prefer racing that makes right turns.
Subscribe to:
Posts (Atom)
Posts
