What to do when no one wants to use your self service data stack?!? You turn to AI.
Monitoring your Sales team performance is very important for the growth of your company whether it is an infant startup or a tenured enterprise. If your company is using a CRM solution, such as Salesforce.com or SugarCRM, you already have a wealth of data to leverage. The trick is to get the right level of actionable data in front of the stakeholders who need it to make a decision.
A Sample Sales Team Dashboard
The dashboard’s purpose below is to communicate the health of a sales team for a company based in Canada. The stakeholder required a view of sales by region and product category. In addition, they wanted visibility into product returns and profitability. We used Tableau Public to create the dashboard below with a Canada Superstore data set available for free. So let’s dive into the dashboard and some of the insights it gives us.
Sales Team Performance
The first three charts on the dashboard illustrate which regions and product categories are driving the most gross revenue. The first heat map shows Product Category against Region with the darker green color indicating more revenue. The conclusion is the West drives the most revenue of the regions, with Nunavut bringing up the rear. However, technology sales are highest in the West, whereas other regions have a more equal distribution.
The second heat map shows Customer Segment against Product Category. Corporate customers tend to drive the most revenue, with technology and furniture being the highest revenue generator. The lowest generator is Consumer Office Supplies.
Data is the future. The future will continue to see an explosion of data collection and an increasing need to digest it. This is what the industry refers to as “big data.” The ability of one company to collect, analyze and take action on large amounts of data can be a serious game changer in the marketplace.
Any stakeholder who seeks to be successful in their role will leverage data. Given the imperfections of our world, the stakeholder may have access to a limited data set. While the stakeholder recognizes their need for clean, accessible data, the IT and BI teams may be months away from delivering.
The stakeholder has has two choices: 1) throw up their arms, complain about the data and cause a ruckus, or 2) work with the data they have and make the best of the situation.
Throwing Up Their Arms
“The data is wrong!” yells the marketing analyst sitting in a meeting with IT and BI teams. The IT and BI managers shrug their shoulders and reply, “then tell us what is right.” The marketing analyst bangs her fist on the conference table in frustration.
Bottom line, stakeholders who don’t embrace even the worst data, does not understand how to measure their business. I’ve seen exchanges between BI and stakeholders where data has been subjected to strict QA by the stakeholder, but the stakeholder has never referred to the data as wrong.
Work With the Data
Every stakeholder interested in a data set needs to have the long term picture of the business in mind and understand the KPIs and other metrics involved to manage their part of the business. All data used in analysis are typically seen through the lens of the business KPI which provides the context. Chances are a stakeholder would never accept a data set that is so far from the truth to be useless.
Based on my career, the best course of action is to work with the data you have. Granted you might not be able to answer more complex business questions, but you will start a journey along a road that will get you there. Take the data you have and create three lists:
- parts of the data set that works for your requirements
- parts of the data set that should be modified
- parts of the data set that are important, but not pertinent to the requirements
Your goal is to understand the ins and outs of the data you have and create a constructive list of actions that evolve your knowledge and the data set into a market changing analysis. Providing documentation on to help the IT and BI teams evolve your data and turn into your pot of gold is the best course of action.
Data is Not Static and Neither is Your Knowledge
Keep in mind that as you interact with data, ask questions, build more detailed documentation and draw correlations or disassociations, your data will have to change to follow your in-depth understanding. This is why maintaining a positive relationship with the team that you rely on is so important.
Iterations of data sets can be subtle and they can also be large. Just remember, that the data you had for version 1 is NOT wrong compared to version 2. When reflecting back on version 1, understand where you came from and that you are looking at a less evolved set of data. Then you can laugh when you look at version 3 and wonder how you managed the business with version 1.
Working with data is an awesome thing. It should be a fun, productive journey for both the analyst, IT team, BI team, and all stakeholders involved. When you here the word “wrong” come up, defend the evolution of data and point out that perfect data sets don’t come out of thin air.
Just like water entering a sink and then finding its way down the drain, your *aaS (SaaS, IaaS, PaaS) business will likely always have customers leaving and sales coming in.
While it doesn’t really matter too much (at least in the context of this post) whether water coming out of the faucet at high rate or low rate, what does matter is the size of the leak at the bottom… the water going out.
First of all, is the drain wide open? If it is, your company is doing more to drive customers away than you are to acquire them. Figure out an effective customer insights strategy and close that drain! Setup effective surveys, reporting, and a churn action committee to address the reasons why your customers are leaving.
The drain will leak. No matter what you do, you will always have a customer cancel due to death or other reasons outside your control. Do you know how much your sink leaks? Again an effective system of capturing data and listening to customers is important.
But, you can go one step further by simply investing in a dedicated analytics team. Let them model churn and build predict future customers who are likely to churn. During this process, they will identify the key drivers of churn and the company can then build a proactive strategy to engaging and saving the customers that prevent your growth.
Before you know it, that sink is a little small for the water you’ve collected, so you have to upgrade to a bigger sink. Just make sure you don’t buy a bigger drain with it.
Understanding customer churn is essential for every SaaS, IaaS, or service company doing business in any marketplace today. Too often companies will be so focused on what is coming in through the marketing facet that they forget to check the flow of water going down the drain. Your sink full of customers might be draining slowly and along with it, your company’s future.
Even the smallest companies will benefit from tracking customer churn every month. Reporting around churn will give you a lagging indicator which can tell you a lot about your company, your product, and your policies.
1) Do you have access and/or the ability to gather information on why your customers leave?
If not, consider an exit survey and follow-up courtesy call to the customer. Also consider setting up a data warehouse to house the information you gather so it may be linked back to customer records, usage, and campaign data.
2) Are your cancellation reasons aligned toward the company’s structure?
Reasons/categories should be aligned along company owned churned (something your company did to cause churn) or customer owned churn (the customer churned because of something they own and is not the company’s fault). In addition, each category should represent a function in the company.
Price too high? Your CRM department should own that one.
Reliability Issues? Your Product/Engineering department should be held responsible for those.
3) How do you report your churn? Is it a single chart? A dashboard? What time frame do you use to report it (quarter or monthly)?
Your churn report should contain multiple views of the same data that let decision makers understand: a) who is churning, b) why they are churning, c) trends over time, and d) the general mood of the customer as they leave (do they hate you or are they going to miss you)?
4) Is the data accessible to everyone who plays a role in decision making? How do you know they review the data?
Setup a recurring monthly meeting a few days after the final report is published with the stakeholders and make sure everyone is on the same page as to what the report says and what action items there.
5) Does your churn report look better than it works? Does it have more animation than the last banner ad you saw?
A well designed churn report is not defined by animation, color palletes or the number of charts. A well designed churn report was built from the ground up with a firm understanding of how your business operates, a solid method of customer engagement, and well defined data viewpoints that help the decision maker understand what is dripping down the drain at a glance.
The more you can understand churn and help the decision makers understand churn, the better off your service company will be. In fact, if you can gather enough good data, you can start predicting churn with a degree accuracy, preventing a customer from churning before it even happens. How much better could that be?
So, what are you waiting for? Get down and dirty with churn, get your plumbing tools and start making sure that drain doesn’t drip any more than it needs to.
Read more of my content on: Customer Experience
I had the pleasure of sitting in on a demo of LogiXML today and must say that I am quite impressed. Super simple to use and integration is superb using .NET or Java. I downloaded the trial, so I will post some screenshots after I have to play! 😉
I am currently evaluating different solutions to solve my company’s business intelligence dilenma. The dilenma is that we don’t have one currently.
After tweeting about business intelligence and looking through some options, I was somewhat surprised to see a number of solutions that covered just reporting and dashboards, or in other words, were simply infrastructure that allows the sharing of data in various forms. There was no ETL solution, there was no database, there were just just numbers and charts offered up on a server.
This contradicts what business intelligence is in my mind. Business Intelligence is an end to end solution that includes three parts:
1) ETL solution that allows you to gather data, transform it and place it somewhere else. Talend and Kettle are examples.
2) A database that houses that transformed data and is reported in part 3. Vertica, Oracle, and MySQL are examples.
3) A data visualization tool that reads data from the database and reports it based on user parameters in various types of visualizations that allow end users to interact with the data. These may or may not include additional analytical JasperSoft, Pentaho, and BIRT are examples.
So when looking at these solutions, JasperSoft and Pentaho come to mind as true end to end solutions. Business Objects is also an example of an enterprise solution.
Another catch phrase, perhaps, I would like to mention is “BI Stack”. I firmly believe that this concept of BI Stack acknowledges that there aren’t many true end to end business intelligence solutions out there. In the case of the BI stack, we are talking about combining solutions that cover the three phases above. In fact, RightScale offers a BI Stack for the cloud that includes Talend (ETL), Vertica (database) and JasperSoft (reporting/visualization).
So what is Business Intelligence? Is it reporting? Is it ETL? Is it none of the above?