DM STAT-1 Consulting

BACKGROUND

Introduce your business and what you do there.

I’m the president of a data analytics consulting firm.

OPPORTUNITY / CHALLENGE

What challenge were you trying to address with DM STAT-1 Consulting?

When I first started in this industry, my mentor, Bruce (Founder & Owner), took me under his wing. Currently, DM STAT-1 helps me with business consulting as needed.

SOLUTION

What was the scope of their involvement?

DM STAT-1 helps me service my clients whenever my expertise isn’t as detailed and specific as their projects demand. Basically, they’re my technical advisor on several of my projects.

Oftentimes, DM STAT-1 helps me identify the appropriate methodology needed to solve the problem at hand. For example, if I want to do a regression analysis for a specific project, Bruce can tell me if that’s the right method or if I need to implement another one, such as factor analysis.

On occasions, Bruce helps me execute the actual work using his expertise. Additionally, he always offers advice and tells me where to look, so I always continue to learn.

What is the team composition?

I only work with Bruce.

How did you come to work with DM STAT-1 Consulting?

I met Bruce while on my first job after I graduated. My publicist and I went to a seminary where he was speaking; that was when I initially met him. Over time, he became my mentor.

How much have you invested with them?

I don’t recall the exact amount I’ve spent on DM STAT-1 services. However, as a professional courtesy, Bruce has often given me advice without charging me. Other times, he just charges the client directly whenever he is in charge of the actual work.

What is the status of this engagement?

I met and started collaborating with Bruce in 1995, and we’re still working together.

RESULTS & FEEDBACK

What evidence can you share that demonstrates the impact of the engagement? 

For me, there’s nothing like having a person who has 30 years more experience than me helping me along the way. DM STAT-1 is the best in class, and Bruce is well-respected in the industry. He gives lectures, writes books, and knows everything related to data analytics. I’m very fortunate to have met him at the beginning of my career.

How did DM STAT-1 Consulting perform from a project management standpoint?

Bruce’s performance is exceptional. Whenever I have a question about anything, I can pick up the phone and call him. We communicate via phone calls, emails, and LinkedIn messages.

What did you find most impressive about them?

DM STAT-1’s ability to easily explain a complex situation in terms that a regular person can understand is impressive.

Are there any areas they could improve?

There’s nothing that DM STAT-1 could improve on. Bruce is very easy to work with, likable, and knowledgeable.

BACKGROUND

Can you please describe your company?

We consult with clients in the area of sales, marketing and service integration, strategy, design, program design/development and optimization.

What is your role and responsibilities?

I’m the President and CEO. We do not have any international operations. We have two virtual offices; one in Cincinnati and one in Atlanta.

OPPORTUNITY / CHALLENGE

What was your goal for the work with DM-STAT 1?

I engaged Bruce [Ratner] over the course of 20 to 25 years for different things, but he supports me primarily in the area of data mining and data modeling. So, designing and developing statistical models and segmentations that we use in conjunction with our marketing analytics work.

Was there a certain challenge that you were trying to address with his work?

Bruce consulted in the design and development of the team that I had when I was at Convergis. I had 12 statisticians and we used Bruce’s software, GenIQ, which is based on the genetic algorithm. He also helped us establish an approach to do data exploration and model development.

Since that time, he’s principally supported me in all of my modeling efforts with clients. The goal is to build models that continue to perform and validate over time. In some cases, we’ve had to take old models out that other people built and retrofit them and make them more productive and have higher levels of performance.

SOLUTION

How did you select him to work with in the first place?

He used to work for David Shepherd, and was a co-author with David on a number of different books. I had David consult with when I was at Sprint. I was introduced to Bruce through David and realized pretty quickly that Bruce was very smart. I started working with Bruce directly.

Can you provide a ballpark dollar figure for their work with your firm?

That has varied dramatically over the course of my work. I’d say in the early years probably in the hundreds of thousands and in recent years probably around $50,000 a year. My project work has tapered off quite a bit in the area of analytics. I’m sitting on a few things right now with him, though.

When was the last project he did for you or milestone completed?

It was probably last Spring of 2012. We updated a model for Merritt’s Motivation Company. We were updating some models that we had built for Han and Onscale, which are two of their clients and just trying to create more longitudinal value of the models that we were building for them. They were five-year forecast models. That’s the last project we did together.

Can you talk about one project in more detail?

I filed a patent when I was at Convergis in the area of term management, and Bruce was principally the statistician that I used to help me design and develop that methodology for life cycle management and term management. We were also trying to link it to lifetime values so we came up with an approach and a methodology that we applied across multiple clients. That was the basis for the patent that I filed.

RESULTS & FEEDBACK

Has DM STAT-1’s performance been successful?

Yes it has. That is why I’ve used him so much over many years.

What is unique or special about DM STAT-1?

He is extremely devoted to his practice. Bruce is an expert in his field and I would call him a world-class statistician among few peers. He’s very creative in terms of the way that he approaches projects. He’s always looking for new and additional ways to apply his knowledge to interesting problems and I think he would tell you that I bring him some of the most interesting and challenging types of work. I need that kind of creativity and out-of-the-box thinking. Bruce is a pleasure to work with. He’s efficient, and he’s fast. He does really good work.

Is there any area that you think DM STAT-1 could improve upon or maybe something you would do differently?

I would just recommend that they work on their written reports to clients. It would save me some time if they write reports a bit clearer and in a client-centric manner. Overall, the final work is great.

BACKGROUND

Can you provide a brief description of your company?

My company provides analytical services to a variety of different clients across an array of verticals. I am a statistician by training. I’m a data miner or data analyst.

What is your role and responsibilities?

I lead the analytical team. I’m responsible for methodology, strategy and executing client deliverables from a technical perspective

OPPORTUNITY / CHALLENGE

What business challenge was your company trying to address?

I have used DM-Stat 1 on several occasions when I get clients whose requests were slightly above my level of knowledge. Their team acts as a hired gun to tackle really advanced projects that couldn’t be covered by either myself or my team.

SOLUTION

Can you describe the project’s scope in more detail?

This was a very challenging project. I had a client who wanted me to develop a campaign response model for his company so that he could target the appropriate people to mail. marketers do that all the time. The problem was that this client had a very small amount of data.

DM-Stat 1 suggested that instead of building a traditional model for the client, we should turn away from statistical methods and use machine learning to tackle the problem. Machine learning does not require the sample size necessary for building a traditional response model via statistics. We did that and got it to the client. The client put it in the mail and got pretty decent results. Now that actually brought about two good things. Number one, it saved me the client. Secondly, it saved the client from having to buy additional data and additional lists. The project was successful by every measure.

DM-Stat 1 came up with a novel approach, something that I hadn’t heard of before although I’m a trained statistician. The method is used a lot more now because more people know about machine learning or artificial intelligence. At that time, it was just beginning to come to life, at least in the database marketing arena.

What technologies were used?

The technology is basic hardware. The software was a proprietary tool that Dr. Bruce Ratner[owner of DM-Stat 1] developed. It used genetic algorithms and is called Gen-IQ.

What was your process for selecting DM-Stat 1 to work with?

I met Dr. Ratner at a conference where he was giving a speech several years ago. His methods are somewhat unorthodox but very interesting. I selected him as a partner because I was interested by his work and knew him to be an expert in data analytics.

Can you please provide a sense of the size of the initiative?

It was a significant engagement that took two to three weeks of full-time work. It wasn’t a very large project but important to us.

When was the project completed?

I believe it was 2009.

RESULTS & FEEDBACK

Can you describe the results of the project?

The client did not have to rent a list and also attributes from data vendors, saving them a lot of money.

How did DM-Stat 1 perform during the project?

Excellent. I gave him the specs and left him alone. I checked up just to see how things were going and I was happy, and so was the client.

What’s unique or special about DM STAT-1 compared to some other companies you may have worked with?

The novel approaches that they take stand out to me. Their team is easy to communicate with as well. Many experts in the data analytics field are rigid without personality. Their team doesn’t have that problem. Their principles are very personable and don’t make you feel stupid for asking questions. I like their company a lot so I’ve done multiple projects with them.

Is there any area that you think DM-Stat 1 could improve upon or something you would do differently?

No, I can’t think of anything.

BACKGROUND

Can you describe your company?

We’re an infant formula producer.

What is your role and responsibilities?

I do statistical modeling and data analysis.

OPPORTUNITY / CHALLENGE

What was your business goal when working with DM-STAT-1?

Bruce [Ratner, founder and president of DM-STAT-1] has one unique tool called the GenIQ. That tool can help you with non-linear regression. They have some unique tools that can help figure out how to classify the right targets. We worked with them for that specific tool to help us to improve our model performers.

SOLUTION

What was your process for selecting DM STAT-1?

Bruce wrote a book. It is a great guide for data analysis. I had heard of him because I studied his book in graduate school, and I had talked with him before about this topic.

Several years ago, I brought up DM-STAT-1 to my boss. He agreed with me that we could use their help for modeling. We started with DM-STAT-1 on a small project for about $10,000, and then gradually gave them larger projects of more importance.

What technologies were used?

I’m using SAS, some customized tools, and SPSS. SAS is the main one we use.

What does the SAS platform offer your company?

I think SAS is very popular everywhere. Students all learn SAS in college or graduate school. It is fully functional, and many people know how to use it. It’s powerful and very convenient to use.

Can you describe the scope of the work that DM-STAT-1 performed?

We use their unique tool GenIQ to do the classification.

Can you give us a sense of the size of that project?

The price range is probably around $10,000 to $20,000 for each project, we have had them do several projects over the years.

When was the project completed?

The last project was completed in December of 2012.

RESULTS & FEEDBACK

Was the project successful?

Most of the projects they did for us were deemed successful. If you have a good data set that is clean and accurate, he can handle it quite well. DM-STAT-1 will find the insights that you need. Just make sure that your data sets are easy to use because their strength is modeling and not cleaning up data groups.

DM-STAT-1 has a lot of knowledge of data analysis, modeling, and regression. Their team will provide solutions that are very innovative. We were often surprised by the new solutions that they provided. Some projects didn’t go as well because the data we provided wasn’t accurate enough. That was more on our side.

I suggest that companies look at DM-STAT-1 for very complex projects where traditional data mining methods won’t work. Their team likes the difficult challenges. Other firms can handle the easier, average data challenges.

Is there anything unique or special about DM-STAT-1 compared to other IT firms you have encountered?

Their GenIQ tool is very unique and respected throughout the industry. That is what made their owner famous, and why most people read his books and hire him.

He is flexible when it comes to the contract terms or scheduling the project. He is also easy to work with during the project, and can adjust to our schedule.

Is there anything that you think DM-STAT-1 can improve upon or you would do differently?

Their team has great technical minds, and knows how to finish the project cleanly. However, it may take them some time to translate business objectives into data analysis and modeling. Sometimes, it requires extra explanation because their project managers are more technical than business minded. This isn’t a big problem, but is necessary to think about before the project begins.