iSixSigma

challenging question for anyone

Six Sigma – iSixSigma Forums Old Forums General challenging question for anyone

Viewing 15 posts - 1 through 15 (of 15 total)
  • Author
    Posts
  • #50603

    Ward
    Participant

    I am learning the DMAIC process and have been stumped.  Is it possible to apply six sigma to the process of selling a photocopy machine? Eg: did I make a sale yes or no?  Were can I find an example of something similar with uses binary data?  What would CTQ, sipoc, look like and what variables should I use for the analyze portion? I have not seen anything like this in any case studies in class.  Thought you guys would like a challenge.

    0
    #174159

    c. wallas
    Participant

    I have never seen it used that way.  I would love to see some one  step up to the challenge.

    0
    #174163

    Gutierrez
    Participant

    Hi Pete,
    I have done a DMAIC project for our sales teams here and the challenge was how to metric their performance. I think it comes back to what you are trying to fix – more sales, more proposals, more time on the road, more sales value, etc? In the end we wanted to improve the ratio of proposals issues for the number of visits, so set a metric that meares this – that 1 in 3 sales calls results in a proposal issued. So the DPMO was based on the number of reps who did not meet the benchmark in a given month.
     

    0
    #174164

    Ward
    Participant

    I am trying to increase sales.  The measure I am using is was the sales call successful?  I am trying to inprove the ratio of visit made customers to successful sale.  any suggestions?  One of the other things I am looking at is the factors that contributed to an unsuccessful sale.  ex:  could not see customer, poor sales pitch etc..

    0
    #174165

    Alex Campos
    Participant

    Hello Pete
    Yes, the DMAIC can be used for this type of process. I would start with the keyword: Process. From what you are mentioning, you already stated a metric to improve. I assume that you already have identified your customers and outputs as well (as the last two SIPOC elements) and the CTQs for each, so now you must start by mapping and identifying the process that creates the metric selected.
    Sales is just another type of process and this sort of projects is very common in service industries. Usually you start with the marketing elements and map every step of the process (which may or may not have a lot of “hidden factories”) until the sales closure, execution and follow up.
    Once this is done you could also proceed with an Ishikawa, C&E matrix or some other root cause analysis to try to find out what is impacting your metric and then decide on the best course of action(s).
    From here on it depends a lot on your process itself. Maybe you have to run a few ANOVAs or even a few DOEs to find out the best formula. For sure you will have more than one factor influencing the sale (from the smile of the seller to the greeting, etc…) and you must also be very careful when selecting the customer “type” (a strong source of noise sometimes) in order to obtain meaningful data.
    This was all very simplistics and there are some more steps that must be added (Control phase) but I hope this helps to clarify a bit your doubts.

    0
    #174167

    Ward
    Participant

    Alex, thanks for the feedback.  This is all new to me.  I just started a class in 6sig and its a bit confusing.  Do you have any material I can reference?  These are the areas I am stuck on: 
    process Map – Supplier(self), inputs (information???), Process(sales pitch), output (sales contract), customers (peole I call on)
    definition of CTQ (I defined it as -delivering quality sale call which keep customer buying product) service- vs CTB (delivering compelling message and closing the customer) :  sounds right?
    measure performance- how many months should I look at.  Currently selected one month and looked at 50 call in that month and whether they were successful or not.  I took the average of successful calls.  If not successful I wrote down the reasons why.  then I put reason a pareto chart.  Broke down each reason into fishbone diagram.
    That’s as far as I got. …
    should I add more than one month so I can set up a distrbution and stdv or is there another way I can look at those 50 call?
    thanks for your feedback!
     
     
     
     

    0
    #174179

    Ken Feldman
    Participant

    In the end you want to end up with a nice binary logistic regression.  The Y will be Yes or No and the Xs will be the variables that you think affect whether you get the sale or not.  Such items as; time spent with customer, are they a previous buyer or not, number of competitor in the marketplace, number of sales calls or details made to the customer, gender of the sales rep, size of business, etc.  This way, you put all the data together and come up with a prediction equation that will let you predict the probability of getting the sale or not getting the sale based on the variables.

    0
    #174180

    Ward
    Participant

    thanks darth

    0
    #174181

    GPaisley
    Participant

    Part of the challenge you will have (and all the other suggestions you have received thus far are all spot on) will be with your measurement system.  When you look at the Xs (e.g. did the presentation go well, did the person smile at the customer, etc….), it can be hard to be consistent in the definitions of these.  I might think that the presentation goes well if I simply get through all the slides, whereas someone else may think it went well if he only got to the third slide but had a deep and meaningful conversation arout your product/service with the customer.  Even if you have a small sales force, ensuring consistency of scoring these attributes over time can be difficult and will require some careful definitions and training to be sure everyone recording data can be consistent.  For this reason, you may want to look carefully at the Xs you identify to see how measurable they are.  There may be some hidden Xs (like someone’s personality or even the smell of their cologne/perfume) that may be a factor but that you may either be unaware of or find difficult to measure.
    This is a good application of DMAIC, especially if your product/service is fairly straghtforward and not something hugely complex like ERP software systems or something like that.
    As I say that, maybe another category of Xs to look at is on the side of the customer.  For example:  did you not make the sale because the customer is distracted with somethign going on in their business or industry, did they just have layoffs or have they just made a large expenditure on some other item.  Do you know if they have spoken with your competitors or have they considered any substitutes for your product, etc.  Obviously, these are things you cannot control, but if you can see them coming and know that the presence or absence of any of these factors affects your ability to make a sale, then you would want to consider it in your data.
    Best regards,
    GP

    0
    #174188

    Erik L
    Participant

    pete,
    If you’d like a very good general reference text, I highly encourage that you look into Applied Linear Statistical Models, by Neter, Kutner, et al.  This is a very complete coverage with many examples of data sets along the lines that you are interested in pursuing. 
    Another item to throw into the mix.  Is going to be assessing the strength of any relationshilps observed from ‘passive’ tools/techniques-such as ANOVA and regression.  The most powerful tool that you have in your hip pocket to strengthen the argument for causality would be randomization.  I’ll give you a for instance.  Let’s say that you are looking at whether time in sales school influences revenue generated by the sales force.  One could then apply a 4-week program for West Coast sales force, 8-week for Central sales force, and 12-week for East coast sales force.  If a relationship develops, can we really peg the length of the sales training as the ’cause’?  Perhaps-to no.  However, if you randomly assigned new sales force recruite from across the country to one of the training variants and then you saw a difference in revenue generated you would have a much more concrete basis to work from.  Remember the basic experimental rule….”block known sources of difference and randomize against those that you don’t know…”  Hope that helps.
    Regards,
    Erik
     

    0
    #174195

    Alex Campos
    Participant

    I personally like “Lean Six Sigma for Service” by Michael George. It has a lot of examples and it’s a very easy lecture, specially for basic training levels.
    As for your questions, these are my comments:
    * When process mapping you have to be careful to differentiate between your company as a supplier to someone else and your own suppliers. I would recommend to place yourself in the Process area and to think about who your suppliers are in the Supplier area. Also be careful not to mix what you expect out of the process with whatever the customer expects as an Output (maybe you expect a contract, be he might be expecting a quick solution).
    * Your CTQ definition must be expressed based on your VOC analysis. In this case it sounds as if you are your own process owner, therefore you must decide what is critical for to obtain as results of the project (more customers? more money? less cost? quicker sales?).
    * I usually take 1 year back to measure performance, but it also depends on how easily available the information is and how stable your process has been. I would recomment that you at least take a couple of months more in consideration (specially if your sales are affected by cyclic effects).
    * Before you get into distributions I would warn you about running statistically fancy analysis without knowing what to use them for. Perhaps from your Pareto you can already see what the problem is and then you can start generating several options as improvent actions.
    One more word of advice: You should always go to the Gemba, this means observing and talking to the people doing the work, your customers, your sellers, etc and try to obtain their own opinion. This may give you a lot of insight about what really is going on, much better than looking at a bunch of data (take a look at the post regarding reliability of the information and people perceptions).

    0
    #174242

    Bryan
    Participant

    Pete – Take a read through of this link: https://www.isixsigma.com/library/content/c041011a.asp
     

    0
    #174456

    rmukeshgupta
    Member

    I am also trying to put the Six sigma to work on a sales process. Any inputs from the experts will be greatly appreciated
     

    0
    #174496

    Stanly
    Member

    I have all that stuff but I’m not going to share it with anybody  (based on bad experience in the past)

    0
    #174509

    Mikel
    Member

    Stanly,Change your name back to Dog Sxxt – you are getting dumber and
    embarrassing me with your stupidity.

    0
Viewing 15 posts - 1 through 15 (of 15 total)

The forum ‘General’ is closed to new topics and replies.