Wses Sales Case Study

Executive Summary:

The major dilemma faced by WSES was to develop a better model of allotment of probability of successful conversion of leads into sales. Based on this, WSES planned to invest marketing budget more judiciously into leads received. As well, WSES wished to better the overall rate of lead conversion. Marketing personnel used to allot winning percentage based on sales value, and partially based on their biases. The intent of this analysis is to perform hypothesis testing on major assumptions that marketing professionals made which influenced their decision to allot budget to individual leads. With the help of outcome of these hypothesis testing, marketing team can direct their efforts in a better way.

To accomplish the same, WSES`s last 4 year sales data was acquired and from that database, a random sample of 1000 data was used for analysis. Different models were used to test different hypothesis such as z hypothesis testing on sample mean and proportions, t-test two sample assuming equal variances, two factor ANOVA test etc. Since the sample size is large, it was a fair assumption to treat the sampling distribution as Normally distributed in most of the cases. The variance of the population was assumed to be equal to population variation for most of the analysis except for two mean t-test for comparing two product categories. The alpha value is assumed to be 5%.

Following were the main biases of marketing team which required analysis: chances of winning the deal against deal value, product category, geography, relative strength of WSES`s product in the industry in specific product category etc.

Our analysis was able to break into many of these biases to either confirm or break them, such as that the average deal size was not more than USD 8M etc. Detailed analysis can be found in appendixes attached towards end of this document.

To start with, the marketing team used to invest 6% of the lead sale value on the marketing for the lead conversion. WSES marketing team was working under the assumption that average sales value of deals was over USD 8M, and based on this assumption they arrived at their yearly market expenses. Our analysis of sampling distribution of mean of leads sales value revealed that the average lead sales value was less than USD 8M. This meant that marketing team was making a Type-I error which is statistically costlier. The cost implications was that if marketing team assumed an average lead value of USD 8.3M, it would have resulted into an excess expenditure on marketing of at least USD 18000 while the revenue may be actually low; thus resulting into losses! Further analysis proves that there would have been no substantial costs had the marketing team been making a type-II error i.e the actual marketing cost was more than USD 8M—as the beta value (type II error probability) decreases sharply.

Further, the Chief Marketing Office (CMO) believed that their sales lead conversion ratio is more than 50%. Upon analysing a sampling distribution of sample proportion, it was found that leads sales conversion ratio was less than 50%. the sales lead that they receive exceeds 50%. It destroys the marketing office belief that WSES enjoys a 50% lead conversion ratio which was deemed as best in industry.

As well, not all biases of WSES employees were false. Such as, the claim of LearnSys product line team that LearnSys enjoys a higher sales lead conversion ratio than FinSys, the companies` one of the 4 product lines wherein company was market leader, turned out to be true. A comprehensive analysis of the claim using Two Sample Mean test (z-two mean test, t-two mean test for equal Variances and t-two mean test for unequal variances) reveals that indeed LearnSys conversion ratio was more than FinSys lead conversion ratio. WSES should prod the FinSys team to improve their conversion rate as the company may lose market leadership in this product line if they don’t convert more of such leads. The FinSys product line can very well take ideas from LifeSys team for the same.