Before a Form: Predictive Analytics for Sales Scott D. Meyer, CEO/Brofounder 9 Clouds Sioux Falls, SD scott@9clouds.com 855-925-6837 1
The views and opinions presented in this educational program and any accompanying handout material are those of the speakers, and do not necessarily represent the views or opinions of NADA. The speakers are not NADA representatives, and their presence on the program is not a NADA endorsement or sponsorship of the speaker or the speaker s company, product, or services. Nothing that is presented during this educational program is intended as legal advice, and this program may not address all federal, state, or local regulatory or other legal issues raised by the subject matter it addresses. The purpose of the program is to help dealers improve the effectiveness of their business practices. The information presented is also not intended to urge or suggest that dealers adopt any specific practices or policies for their dealerships, nor is it intended to encourage concerted action among competitors or any other action on the part of dealers that would in any manner fix or stabilize the price or any element of the price of any good or service. 2
THE MARKETING CLIFF We are running towards the marketing cliff. We send more and more email and push harder and harder for form submissions. As we do that, fewer and fewer people open the emails or fill out forms. The current approach to marketing is focused on blasting as many people as possible with the same message trying to get them to fill out a form or visit the dealer so we can then blast them all again with the same message to get them to buy. This is desperate (and doesn t work). In fact, most buyers are making their decisions without filling out forms. They research online and come into the dealership ready to buy. PREDICTIVE ANALYTICS SOLUTION The solution is identifying who is interested based on their behavior. Then the sales team can act proactively to contact interested buyers and the marketing team can send targeted messages that educate and help the buyer. No more one-size fits all solution. Instead, a data-driven approach that uses big data from your specific store and predictive analytics to know who is ready to buy, before a form submission. THREE RESULTS FROM PREDICTIVE ANALYTICS double email open and click rates increase sale quota fulfillment by 9.3% know customers are ready before they tell you 3
HOW TO GET THE DATA To begin using predictive analytics for sales, you need to export your customer data. Every CRM system will have an option to export customer data to a CSV or Excel file. Here is where you will find export options on three different CRM systems: ReyRey 4
ADP DealerSocket 5
HOW TO GET BENCHMARK (HISTORIC) DATA Once you have exported the data, you want to find your benchmarks. These are the moments when your customers typically take an action such as buying or servicing a new vehicle. There are different types of benchmarks your store can find. Each benchmark can be created at a store level, or better yet, within new or used, segments (e.g. - cars, trucks, etc.) specific makes and even for specific sales consultants. Time Based Benchmarks DAYS BEFORE FIRST SERVICE Know how long before your customers visit the service bay. 1. Filter to view customers. 2. View purchase date and service date. Hide other columns. 3. Create a blank column next to purchase date and service date. 4. Create a formula to calculate the days between purchase and service. If your blank column is column D, you can use a typical subtraction formula such as: D1=B1-C1. 5. Find the mean number of days between purchase and service. Marketing action: Invite customers in for service as they approach your average days before service. DAYS BEFORE PURCHASE Find the number of days from a customer s first purchase to their second purchase. This is similar to the days to first service formula. 1. View customers who have made two purchases. View purchase date one and purchase date two. Hide the other columns. 2. Create a blank column next to purchase date two. 3. Create a formula to calculate the days between purchase one and purchase two. If your blank column is column D, you can use a typical subtraction formula such as: D1=B1-C1. 4. Find the mean number of days between purchases. Marketing action: Contact customers as they approach your average days before purchase. 6
PULL AHEAD LEASES Move customers into a new lease 1. Filter new customers by recent sales dates. 2. Subtract T1 lease end dates from sales dates. (If you do not have T1 data, use the previous month s customer lease date, copy it to the new spreadsheet and subtract the sales end date from the sales date.) 3. This is the number of days/months left in the average customer s lease when they lease a new vehicle. Marketing action: Contact customers approaching the average days left in a lease and encourage them to re-lease. Event Based Benchmarks AVERAGE MILEAGE AT PURCHASE 1. Open your customer records in Excel 2. Filter to only customers with sales dates (These are people who have purchased.) 3. View T1 mileage. (This is the mileage at trade-in. If this is not available, you will have to use a previous month s contact record.) 4. Calculate the T1 mean. This is your average mileage at purchase. Marketing action: Contact customers within 5,000 miles of your benchmark AVERAGE MILEAGE BEFORE SERVICE 1. Open your customer records in Excel 2. Filter to only customers with Number of Service RO s=1 (This is the first service of customers. Watch out if a customer purchases accessories and has the first RO before driving the vehicle.) 3. View RO mileage in. (This is the mileage at service.) 4. Calculate the RO mileage in. This is your average mileage at service. Marketing action: Estimate the average miles a customer puts on their vehicle in one month. Contact customers one month before you estimate they approach the average mileage at service. 7
VEHICLE UP-SELLS Up-Sell your customer to the next vehicle make. 1. Create your vehicle up-sell graph. For example, your store would like a current Honda Civic owner to upgrade to a Honda Accord. 2. Sort customers by vehicle make. 3. Use your average mileage at purchase date to filter these customers to potential buyers. Marketing action: Contact the potential buyers and focus the conversation on the vehicle chosen as the next up-sell. Forecasting Benchmarks VEHICLE EQUITY Calculate when customers will buy based on 1. Filter customers who have a purchase price and a buy back price. 2. Average the purchase price and the buy back price. 3. Divide buy back price by average purchase price. This is the average equity for your customers. Marketing action: Contact customers whose vehicle buy back divided by potential purchase equals your average (or ideal) equity number. Sales action: Monitor your store s average equity. Increase equity month-over-month. Forecast action: Identify the number of customers whose buy back divided by purchase price is within 10% of your average equity number. Use this data to plan on used or CPO inventory. LEASE FORECASTING See the future of your store s trade-in business. 1. Filter customers by the number of months left in their lease. For example, filter 72 months and write down the number of customers. Then 71, 70, 69, etc. 2. Create a bar graph of how many customers are in each month. 3. Predict the future pull ahead leases based on your average pull ahead lease data. Forecast action: Predict whether you should expect a high or low number of pull ahead leases. Create incentives for sales consultants and customers based on what will happen. Competitive insight: If you are considering purchasing or consolidating with another dealership, use this data to understand the future health of the store. 8
HOW TO GET PREDICTIVE DATA (ONE DIMENSIONAL) The benchmarks tell you when purchases will typically happen. Predictive data tells you who is ready to buy based on your dealership s history. The easy way to get predictive data is to look at which customers will hit your key benchmarks in the coming days, weeks or months. This list will show customers who are potential customers based on one data point. If you want to increase their qualification, see if they are approaching multiple benchmarks. This method, however, does not tell you which benchmarks are most important. To do this, we need multi-dimensional predictive data. HOW TO GET PREDICTIVE DATA (MULTI DIMENSIONAL) To accurately predict who is a qualified lead and to know which benchmarks are most important, multi-dimensional predictive analytics are required. In math jargon, you can use logistic regression based on weight of evidence to identify the variables that predict when someone is in buying mode. In plain English this means, you can look at your data to know based on a combination of factors who is going to buy (even if they haven t filled out a form). This type of calculation is beyond Excel. An algorithm is required that does the following: Creates a set number of bins in which to test dependent and predictor variables Calculates logistic regression using Weight of Evidence as predictor variables Uses a sale of the vehicle as the dependent variable Calculates key benchmarks as the predictor variables Identifies the predictor variables that have the biggest influence on purchase Provides the names of potential customers qualified based on all factors 9
CUSTOMERS DON T TELL YOU WHAT THEY WANT, THEY SHOW YOU WHAT THEY WANT Predictive analytics will help your store: double email open and click rates increase sale quota fulfillment by 9.3% know customers are ready before they tell you You don t have to use all of it, just the data that will help you. Start with one customer segment, create a benchmark and use it to improve marketing and sales. HOW TO USE THE DATA - SALES Both historic and predictive data helps you proactively approach potential sales and improve marketing to send specific messages to interested buyers. Here are eight data recipes that will help your sales and marketing teams create smart marketing based on data and personalized information. 10
11
12
13
14
THREE RESULTS FROM PREDICTIVE ANALYTICS double email open and click rates increase sale quota fulfillment by 9.3% know customers are ready before they tell you 15