Monday, January 4, 2016

Using a Fish to Generate a Rainbow at the end of a (Brain)Storm

Take the noise out of your brainstorming sessions


Have you ever finished a brainstorming session and thought, wow, that was fun, but what just happened? Chances are that brainstorming session wasn’t very structured or didn’t yield the results you were looking for.

One way of structuring these sessions and coming up with a clear course for action is to use a tool called the “fishbone diagram” or the “cause and effect diagram.”

Rather than thinking of ways to solving a problem, we put the problem as the primary focus and think of the various “micro-problems” or causes that contribute to the greater problem. This will give you the long view and help you identify the most impactful ways of addressing the situation.

The structure of this tool is that of a fishbone, hence it’s title. Think of the “macro-problem” as the head of the fish.  Then find the center of the left side of the box and draw a line outwards, like so:
That’s the spine. And from there you have the smaller bones. You’ll need about 6-8 of those. Draw 3-4 up top, and 3-4 at the bottom.

Each bone is its own category, and the categories are:

  1. People
  2. Method
  3. Machine
  4. Measure
  5. Mother nature
  6. Material
  7. Management (optional)
  8. Money (optional)



These categories serve to force us to take a look at all components feeding the problem.  From here we identify the causes that impact the macro-problem and sort each cause by category.

Let’s look at this from a marketing context.

Situation:


We are generating a lot of leads in Q4 but they are slower to convert compared to Q3.

  1. Head – leads slow to convert during Q4
  2. People – sales people not closing, education
  3. Method – lead acquisition channels could be improved, nurture campaign not educating on benefits of solution or product differentiators, are the form fields the best to help sales close?, are there too many form fields?
  4. Machine – automated campaign – emails trigger during holidays
  5. Measure – Using Q3 data to benchmark Q4 data – might not be comparable
  6. Mother nature – inclimate weather – difficult traveling to work
  7. Material – campaign asset not appealing

Breaking it up in these buckets, often called the 5 M’s and 1 P, will allow us to look at the problem from all angles. The list generated should highlight process elements vs. people issues. If the people bucket is a bit full, take a look again and see how many of those are really process issues or break down the people issues using a 5Y analysis, which serves to chisel the problem to its root cause.

Once you have completed your diagram you can take a step back and analyze which is the best direction to go. Multi-voting with the team is a great way of narrowing the list and identifying the next best course of action.

But the idea here is to generate a good working list of micro-problems by taking a long view of the process and isolating the contributing factors. A good brainstorming session shouldn’t hit you like a whirlwind, but instead, help you generate, at minimum, a strong hypothesis you can bring back to your marketing lab for further analysis.

Monday, December 28, 2015

The "Voice of the Customer" is a Key Organizational Driver

To better serve customers and markets, organizations must first understand them


One of the key drivers for any marketing initiative is to bring to market products or services with the needs of the customer in mind to build awareness, loyalty, and relationships. This entails a strong understanding of the requirements and desires of customers and/or market. Capturing the "Voice of the Customer" (VOC) isn't accomplished through guesswork or management having a hunch on something. It involves an integration and analysis of survey data, focus group findings, web insight (including research, social media and analytics), and any other insights that can be obtained. By looking closely at data it and combining it with other analysis, an organization can sharpen their processes and better serve their customers and their market.

Looking at isolated data points will not provide the complete picture needed to make key decisions


A key component to obtaining the VOC is having a regular process in place by which to capture and aggregate data. Most marketing tools have built in dashboards that can be accessed at any time. These tool-specific insights do a nice job of focusing on the strengths and weaknesses of the marketing initiative the tool is supporting. But those insights tell a fraction of the story. The challenge is to identify the key analytics the tool is presenting, and combine it with key analytics from some of the other tools that are available. Setting up dashboards that aggregate this data in Sharepoint or Excel and keeping these updated can be a great place to begin. There are business intelligence tools available as well that can help aggregate data. But data derived from any database is only as good as the data that went into it. It's important to be as thorough as possible when setting up dashboards and fields so you can splice as needed.

Survey's can help capture the "Voice of the Customer"


A regular survey program or process can also help determine the needs of the customer or market. But relying on surveys alone to glean market or customer insights can be dangerous because response rate might be low. But a carefully worded and analyzed survey can provide the insight needed, and this insight can be combined with other data captured can be extremely useful.  

In conclusion


The "Voice of the Customer" can determine messaging, process improvements, and new product or service releases and updates. It's important to have enough data on hand for analysis, and not to rely on one source for capturing the customer's voice. Developing processes to capture data and actionable dashboards to aggregate data can help understand the VOC.


Monday, December 21, 2015

What is the Probability that X Many People will do Y?

How to Apply the Poisson Formula


Let's say there is an event you've attended for the past few years. You've had great success at this event, and you want to make sure you have enough giveaways. In the past few years you've noticed that on average 25 people visit your booth during peak time between 5 and 6 pm. These giveaways are expensive. You don't want to order too many, so you ask yourself, what is the probability of exactly 25 people showing up to the booth?

To do this you need to use what is called the Poisson Distribution. The formula looks like this:


It's fairly easy to calculate on a calculator, and you can view a great video on how to interpret this formula by clicking here.

The formula probably looks more complicated than it actually is. But in Excel or in Google Sheets, this task is very easy complete.

To do so, use =POISSON(X,MEAN,FALSE)

X = 25, which is the probability you are trying to work out
MEAN = the mean, in this case that is 25
FALSE = this is not a cumulative number - we are looking for an exact number

To solve, it is =POISSON(25,25,False), which gives us a probability of 0.08.

The chart below shows us how the probability pans out when calculated from 0 to 50 people.


Another question you might have is, "What is the probability of more than 30 people stopping by during this time frame?"

To calculate that using Excel or Sheets you use the formula:

=1-POISSON(29,25,TRUE)

This looks a bit different than the previous formula. First, there is a one in front of the formula. This is because we need to subtract the answer from the whole to calculate the answer. The other major difference is the word "TRUE" is at the end of the formula. This tells Excel or Sheets that this is "Cumulative", meaning, that we aren't looking for a specific number to calculate the probability of.

The other thing you might notice is that the X = 29. This is because we are looking for 30 and everything above it.

So when we plug this formula in, we get 0.18, meaning that there is a probability of 18% that 30 or more people will visit the booth.

The key to using this formula is to have an average number first, then you can calculate the probability of X number of people or clicks doing Y. This is a great planning tool to help you anticipate results or properly plan for results of an action taking place.