Designing a Lead Score Model for Digital Marketing Firms in Education Vertical in India

Background: Indian higher education sector is characterized by severe competition. The market size is 91.7 USD and its contribution to GDP is 3.8%. The admissions market is an evolving one with digital players’ signiﬁcant role. Students depend on these portals for making decisions pertaining to a selection of an institution. These companies make revenue by providing various digital marketing services to the educational institutions. Selling leads is one of the critical activities as part of their package oﬀer to various educational institutions. Objectives: The major problem faced by these companies is the low conversion rate of the leads at the client end. The main objective of this study is designing and developing a lead score model for these companies to qualify the leads. Methods: The research focuses on developing a lead score model for these companies to qualify the leads based on three parameters — explicit parameters, implicit parameters, and negative parameters. These parameters are chosen because they represent the lead behaviour and engagement level. The overall lead score is calculated by assigning scores for each parameter. The value for each parameter is determined based on their importance and also in consultation with the sales team. Then the leads are classiﬁed into hot, warm, and cold leads based on the lead score and a lead score matrix is created based on explicit and implicit scores. Findings: The major ﬁnding of the study shows that out of a sample of 1900 leads, 21% are hot leads which are sales qualiﬁed, 35% are warm leads also called as marketing qualiﬁed and 44% are cold leads. The lead score matrix is used to qualify the leads. The result of the matrix shows that 60% of the leads are qualiﬁed. Thus, it helps the digital companies to ﬁlter out unqualiﬁed leads and manage the leads in a better way which improves the quality of the leads delivered to the clients. This will raise the conversion rate at customer level. Such improved conversion rates reinforce the business model of digital marketing companies. Novelty: The lead score model is designed with customization and applied for digital marketing ﬁrms in education vertical in India.


Introduction
Leads are very important to the organisation. A lead is a person or a company who has some interest in the products or offering of the organisation. They are the source from which the potential customers are obtained. Thus, leads are very important to the organisation and hence it is essential for every organisation to manage their leads effectively.
The online Education sector in India is booming and offers excellent growth prospects. The higher education sector provides various business opportunities. The Indian education market size is about 91.7 billion USD and it contributes about 3.8% of the country's GDP. There are many digital marketing companies in India which provides information about colleges, courses and exams. These portals help students in searching colleges or courses according to their choice. They are used by students as well as parents. They also provide free resources such as discussion forums, counselling support, students review and also informative articles. These companies provide free listing as well as various digital marketing options to the educational institution such as running display advertisement campaign and Email campaigns. The business model of these companies is that they earn revenue by selling leads to educational institutions and by providing various marketing services to these institutions.
The major problem faced by these companies is the low conversion rate of the leads at the client end even though they generate huge volumes of leads through their digital platform. Industry trends indicate that the conversion rate of these leads generated by digital or online marketing companies is about approximately 5% of the overall leads generated. Most of the companies lack proper model to qualify the leads and hence nurturing the right type of leads is not possible. These problems can be addressed by using a lead score model. The lead score model helps the organisation to qualify the leads into Marketing qualified and sales qualified leads. By using this model, the sales team of the clients can focus their efforts on more qualified leads. This would lead to increasing chance of conversion and shortening the sales cycle.
This paper explains the process of developing a basic lead score model which is customized for those companies which provide information about colleges and courses through their digital platform. Lead scoring is a part of customer relationship management. Lead scoring is the process of assigning scores to leads based on various parameters like explicit parameters, implicit parameter and negative parameters. The scores are assigned based on the activities of the leads. These parameters reflect the demographic, behavioural and contextual characteristics of the leads. After calculating the lead score, the leads are categorized into Hot, Warm, and cold leads. The hot leads are sales qualified leads, the warm leads are marketing qualified leads and the cold leads usually have low interest towards the organisation. A lead score matrix is also used to qualify the leads. The matrix is built using implicit and explicit scores which represents target fit and engagement level of leads. Thus the qualified leads will have a high chance of conversion.

Background
Every organisation is spending huge amount of time and resources in following up the leads. Without a proper method to qualify the leads, many marketing qualified leads (MQL) may be left unnoticed. To overcome this issue, a lead score model is used. Lead scoring refers to the practice of calculating and assigning scores to the leads of the company. This score helps the company to select target, establish contact priorities and personalise marketing actions (1) . It also helps the sales and marketing team to save time and resources by targeting their efforts on more qualified leads.
A lead score model also helps to track the leads behaviour and their web activities. Digital marketing has become an important part of companies in the process of attracting new customers. Digitalization has led to substantial changes in the ways consumers and businesses search for information and do their research before making a purchase, a major shift has been observed regarding how digital communication influences the purchasing decision within the B2B sector. Today's marketing technology allows organizations an unprecedented degree of insight into their leads' decision-making, and lead scoring is the foundation of actionable insights that will improve the return on investment (2) .
A huge amount of data is generated by the company, but it is very important to select relevant data that can be used to qualify the leads. Lead scoring is used to guide companies in prioritizing which leads to target. As a starting point, companies may choose to start qualifying leads according to the data they have on them. The idea here is that sales people should only spend their time on contacts that have a high lead score, which assuming a reliable scoring procedure implies that they will also have a high sales conversion probability (3) .
Sales representatives often do not have the time or resources to rationally select the best leads to call. As a result, they rely on gut feeling and arbitrary rules to qualify leads. Model-based decision support systems make this process less subjective (4) . Major questions facing the company include how to differentiate immediate prospects from suspects and the idly curious, and how to advance the prospect who has an immediate need toward accepting a sales offer (5) . Knowledge in qualifying leads can help systemize the process and give better standardized information about leads to both management and the sales representatives managing them (6) . Sales person through a call or sales meeting can gather valuable information from the clients (7) . Activities https://www.indjst.org/ in sales funnels can be categorized into two broad types: sales lead generation and sales lead conversion into sales through presentation and persuasion (8) . The lead qualification model help marketing representatives assign leads more effectively to sales representatives, thereby fostering follow-up of marketing generated leads (9) .
A basic lead score model can be developed from behavioural, contextual and negative parameters based on which the lead score can be assigned. The lead scoring method determines the level of interest of a lead in the product (10) . A business should classify their customer's value chain into value-creating, value-charging (monetizing) and value-eroding activities. Hence, the value-creating and value-eroding activities representing the implicit data and negative data respectively (11) . A matrix based on profile fit and level of engagement is created which classifies the leads into various categories. This matrix is used to identify more qualified leads (12) . The parameters used for building the lead score model is very important as they reflect the leads behaviour and their level of engagement. The lead score model built using the literature will not be suitable for digital marketing companies in Education sector, since some of the parameters used in the model does not accurately reflect the leads behaviour in online educational platform. This paper aims at building the lead score model by redefining and restructuring the parameters, which will suit the online digital marketing companies in educational sector and helps them to qualify their leads.

Methodology
The objective of the study is to design and develop a lead score model for online digital marketing companies in Education vertical. The lead score model helps these companies in qualifying the leads which in turn will increase their chances of conversion. The approach towards building the lead score involves the following stages: • Identification of possible and appropriate parameters in online Education sector to assign the lead scores. Typically, the lead score model is built using demographic, behavioural and contextual characteristics of the leads. In this paper, the parameters are chosen in a way which accurately reflects the leads behaviour in online educational platforms. These parameters are classified in three types -explicit parameters which describes the demographic characteristics, implicit parameters which describes the behavioural characteristics and Negative parameters. The detailed list of parameters used in the study is explained in the next section. • The response of the leads towards these parameters are recorded and scores are assigned to arrive at the overall lead score. • The leads are then classified into three categories that give their utility to the sales department. • Various statistical tools (Multiple Regression and Chi-square) are deployed to understand the specific relationship between the specific parameter and the lead category.

Business Problem
The online digital marketing companies in educational sector generates large number of leads through their digital platform (portal). These companies sell leads to the educational institutions and the major issue faced by them is poor lead conversion rate. Most of these companies does not have a proper model to qualify the leads and thus a lot of marketing qualified leads goes unnoticed. This has an impact on the overall sales performance of the company as the clients demand better return on investment for their digital advertising expenditure. The companies also spend large amount of time and resources in following up the leads. This problem can be addressed by using a lead score model for qualifying the leads. The main objective of the paper is to design and build a lead score model for better qualification of leads which will also increase the lead conversion rate.

Lead Score Model
The lead score model is a model used for qualifying leads by assigning scores to each lead based on various parameters. The scores for each parameter are then totaled to arrive at the overall lead score.

Determining parameters for leads score model
The parameters are identified through literature and they are redefined to suit the online marketing companies in educational space. The parameters reflect the leads behaviour and engagement level. The scores are assigned to each parameter based on their importance in predicting the leads behaviour. The score ranges between -10 to +10. These scores are added up to arrive at the final lead score. These parameters are classified as explicit, implicit and negative parameters. The explicit parameters describe the demographic characteristics of the leads. These parameters are used to identify whether the leads fit into the ideal target profile desired by the company. The explicit parameters include company size, location, company region, Industry type and the designation of the leads. The Table 1 (List of explicit parameters) show the explicit parameters used in the case illustration. The implicit parameters describe the behavioural and contextual characteristics of the leads. These parameters can be used to identify the level of engagement that the lead shows towards the institution. These parameters should be given more importance because these activities highlight the interest level of the leads towards the institution. The implicit parameters include opening emails, click through rate, downloading contents, signing up for newsletter, participation in forums and blogs and response to telemarketing campaigns. Some of these parameters might be more important and hence more weightage can be given to those parameters. The Table 2 (List of implicit parameters) shows the implicit parameters used in the case illustration.  The negative parameters represent the negative responses of the leads; these activities indicate that the leads might have a less interest towards the institution. The negative parameters include unopened emails, unsubscribe request from newsletter, joining do not call list. The negative parameters should be reduced from the overall lead score. The Table 3 (List of negative parameters) shows the negative parameters used in the case illustration.

Lead categorization
Thus, after determining the parameters, the values are assigned to each parameter based on the engagement level of leads towards that parameter. (For example, the parameter -willingness to join in the current academic year is given a full score of 5, when the lead is willing to join in the current year. If the leads are willing to join next year, then they will be given a score of 3 and no score will be provided, if the leads are unwilling to join. The overall lead score is calculated by adding up scores of explicit and implicit parameters and subtracting the scores of negative parameters. Based on the lead score, the leads are classified into three categories. They are hot, warm and cold leads. The scores for each categorization is determined based on the combination of various important parameters and in consultation with the marketing and sales team of the company. Hot leads have high chance of conversion and require less marketing effort, these are sales qualified leads and can be directly passed to the sales team. These are high value leads for the business. Warm leads are leads that require marketing effort to nurture them and increase their chances of conversion. These leads are classified as marketing qualified leads, and they also have value to the organization. Cold leads have little interest in the company, and they have a low chance of conversion.

Lead Score Matrix
A lead score matrix is created using explicit and implicit data with X-axis representing leads engagement and the Y-axis representing demographic fit.
The profile fit score can be obtained from calculating the explicit scores. The leads with explicit scores greater than 75% are assigned rating ' A' . The leads with an explicit score between 50-75% are assigned rating 'B' . The leads with an explicit score between 25-50% are assigned rating 'C' and the leads with an explicit score of less than 25% are assigned rating 'D' .
The engagement score can be obtained from calculating the implicit scores. The leads with implicit scores greater than 75% are assigned rating '1' . The leads with the implicit score between 50-75% are assigned rating '2' . The leads with the implicit score between 25-50% are assigned rating '3' and the leads with an implicit score of less than 25% are assigned rating '4' . Thus, the matrix results in 16 possibilities.
https://www.indjst.org/ The clusters A1, A2, A3, B1, B2, and C1 have good fit and high engagement which appear at the upper right of the matrix and have the highest priority. They are on the verge of buying and can often go directly to the sales team for closure. The clusters B3, C2, D1 forms the next category. Although they may not be customers yet, they have a higher than average likelihood of becoming buyers in the future with the right marketing messages over time. The cluster A4, B4, C3, C4, D2, D3, D4 are the leads with low engagement and an unlikely fit. Their lead scores are the lowest and indicate that they are the lowest priority for your marketing department.

Case Illustration
The proposed lead score model is tested using the sample data derived from the various sources related to the online educational platforms.

Data
For this study, a sample of 1900 leads are selected. The implicit and explicit parameters are selected and the scores are assigned for these parameters based on the actions performed by the leads which can also be identified through the company's website. Table 1 (List of explicit parameters), Table 2 (List of implicit parameters) and Table 3 (List of negative parameters) indicate the parameters used in the case illustration.

Validation
After implementing the lead score model, the lead scores are calculated. The result showed that out of 1900 leads, 21% are hot leads which are sales qualified leads, 35% are warm leads also called as marketing qualified leads and 44% are cold leads. Thus the company can focus more on hot and warm leads which helps them to save crucial resources.
The cold leads may contain some leads which can be converted, and hence a lead score matrix is used. It is used to identify the marketing qualified leads, which the company can continue to follow.
https://www.indjst.org/ Before implementing the lead score model, all the lead generated are passed to the sales team. For example, if 100 leads are generated all the leads are passed to the sales team. After implementing the lead score model, the leads that belong to the cluster A1-A3, B1-B3, C1, C2 are qualified. Thus out of 1900 leads, 1140 (60%) leads are qualified based on the lead score and are passed on to the sales team. Thus, by using the lead score model the company can filter out uninterested leads (cold leads) at the initial stage of the sales funnel. Thus, it helps the organisation in saving time and resources. Since the leads are also more qualified, the company can personalize their marketing efforts and it increases their chances of conversion.
We have also used various statistical tools to establish further relationships with the data. The statistical tools multiple regression and chi square test have been used to test the relationship between implicit parameters, explicit parameters and the lead score.
In determining the relationship between the implicit parameters and lead score, the results of multiple regression analysis show that the implicit parameters can predict 85% of the variation in the overall lead score. The regression analysis shows that the explicit parameter can predict 25% of variation in the overall lead score. The chi square test is used to identify the strength of association between each implicit parameters and the lead categorization. The strength of the association can be identified through Pearson's chi square value. The results of chi square test indicate that all the parameters have an association with the lead categorization. Thus it is evident that the lead score is influenced by all the parameters used.

Conclusion
The study has focused on the design and development of a lead score model for online marketing companies in Education market. The lead score model is built using parameters which accurately reflects the behaviour and engagement levels of the leads. The proposed lead score model will help the company to qualify the leads better and increases their chances of conversion. The lead score model is tested using a case illustration. Thus out of 1900 leads, 21% were classified as hot leads, 35% as warm leads and 44% as cold leads. We have also used a lead score matrix to identify the qualified leads. The lead score matrix is built using explicit and implicit scores. The matrix defines the profile fit and engagement level of the leads. The result of the matrix showed that 1140 (60%) out of 1900 leads were qualified based on their lead score. The online marketing companies generate huge amount of leads, thus by filtering out the unqualified leads, the efficiency of the marketing team and the sales performance of the company is improved by focusing their efforts on more qualified leads as the conversion rate is also increased.