Business Analytics in Steel Product Fabrication Cluster

Abstract

Forty micro and small steel products manufacturing enterprises in Salem District of Tamil Nadu, South India faced problems in value addition of the steel products like windows, grill gates, truss work and panel boards manufactured by them. They formed M/s Salem Steel Cluster Pvt Ltd; Salem, a special purpose vehicle, in 2012 by getting funds from the Government of Tamil Nadu and the Government of India through the Tamil Nadu Small Industries Development Corporation under Micro Small Enterprises Cluster Development Programme of Ministry of Micro, Small and Medium Enterprises, Government of India. The objective is to find the physical and financial performance of the Steel Product Fabrication Cluster (SPFC) before and after Cluster Development Approach (CDA) to find the productivity of the cluster by taking independent variables like number of units, employment and production and dependent variable like turnover, and to find the performance of SPFC before and after CDA. To find business analytics models like Diagnostic Analytics, Descriptive Analytics, Inferential Analytics, Predictive Analytics, Prescriptive Analytics and Decision Analytics. The methodology adopted is by collecting primary data like number of units, employment in numbers, production in crores and turnover in crores before and after CDA and analysing using Compound Annual Growth Rate, Descriptive Analysis, Correlation Analysis, Trend Analysis, Regression Analysis, Structural Equation Modelling and T-Test. There is a Difference in Difference between controlled units which have not adopted CDA and experimental units which have adopted CDA, where there is an increase in number of units, employment, profit and turnover. To conclude, there is an increase in number of units, employment, production and turnover after CDA when compared to before CDA, which leads to an increase in productivity thereby Sustainable Development Goals of 1, 4, 5, 8 and 9 are achieved.

Introduction

In the UNIDO context, clusters are defined as ‘geographical concentrations of inter-connected enterprises and associated institutions that face common challenges and opportunities’. UNIDO’s approach to cluster development contributes to achieving Sustainable Development Goals (SDGs) in various ways. UNIDO helps cluster entrepreneurs to overcome barriers to growth by increasing their collective efficiency and helping them to access new markets, thereby promoting inclusive and sustainable economic growth (SDG 8). Cluster development encourages cluster firms to develop new products and improve production processes. It thus contributes to innovation and ensures inclusive and sustainable industrial development (SDG 9). Cluster initiatives can thus also have a positive impact on improving education in project countries (SDG 4). In cluster initiatives, UNIDO often works with micro-enterprises in peripheral areas and marginalised populations, in particular, young people and women. Supporting them will hence have a positive impact on reducing poverty (SDG 1) and enhancing gender equality (SDG 5). By promoting the development of agro-food, tourism and creative industries, which are often concentrated in peripheral regions, the Integrated Cluster Approach contributes to reducing regional inequalities within national borders (SDG 10). Better communication between stakeholders of the three sectors together with a collaborative decision-making and governance structure fosters regional development not at the expense of but by preserving natural and cultural heritage, through raising awareness about the importance of biodiversity and sustainable utilisation of resources for the valorisation of territorial assets (SDG 15) (UNIDO, 2020).

Technical Survey

Significance of Clusters in Developed Economies

Economist Alfred Marshall (1880) has commented on the gains of clustering to enterprises in the textile and metal-working regions of England and Germany (Humphrey & Schmitz, 1996). The observations on ‘Clusters and Competition’ by Porter (Porter, 1990, 1985) have also extensively considered the contributions of clusters in giving developed nations a competitive advantage (NI-MSME, 2006).

Significance of Clusters in Developing Economies

Nadvi and Schmitz (1994) observe that sector-specific and geographically bounded clusters are frequent in SSE manufacturing in developing countries. One of the export successes is Sialkot stainless steel surgical instrument cluster. This cluster targets markets in the United States and Europe. The Pakistani clusters of electrical fans (Gujarat) and farm machinery (Daska) are also well known. Brazil’s leather shoe exports, Franca and the Sino’s Valley specialising in footwear for men and women, the roof tiles manufacturing clusters in Central Java, Indonesia and the carpenting and metal-working clusters of Ghana are examples of sectorally specialised clusters. Mexico’s footwear industry is also clustered around Leon and Guadalajara (NI-MSME, 2006).

Significance of Clusters in India

In India, regionally concentrated industry and specialised clusters of small firms produce hand block printed textiles in Jaipur, locks at Aligarh, machine tools at Batala, leather footwear at Agra and Kanpur, power looms at Bhiwandi, roofing tiles at Morvi, diesel engines at Phagwara, Rajkot and Coimbatore, electronics at Noida, hosiery at Tirupur, coir at Alleppey, brass parts at Jamnagar and diamond polishing in Surat. These are all examples of the 350-odd industrial clusters and over 1,700 artisan clusters in the country (NI-MSME, 2006).

Significance of Cluster in Tamil Nadu

The Ministry of Micro, Small and Medium Enterprises (MSME), Government of India (GoI) has adopted the Cluster Development Approach (CDA) as a key strategy for enhancing the productivity and competitiveness as per Micro Small and Medium Enterprises Development Act, 2006 (Ministry of Micro, S. a. (2023) as well as capacity building of Micro and Small Enterprises (MSEs) and their collectives in the country. Government Policies on CDA will improve Entrepreneurship in India.

Many studies have been made to identify performance for clusters under CDA like match, printing, auto components, leather, plastic, hosiery, textile, lorry body building, pharmaceutical, ceramic, wet grinder, jewellery and rice mill. Tamil Nadu is the first to implement 24 clusters in India as per Figure 1, which includes Steel Product Fabrication Cluster (SPFC), Salem (Micro, 2022).

Figure 1.

Source: Tamil Nadu Small Industries Development Corporation (TANSIDCO) (Micro, 2022).

Forty micro and small steel products manufacturing enterprises in Salem faced problems in the value addition of the steel products (window, grill gate, truss work and panel board) manufactured by them. They manufacture stainless steel pipes are and also involved in various sheet metal processes like laser cutting, bending, forming, fabrication and material testing. They currently serve defence and aerospace, agricultural implements, rigs, railways, automotive, etc. They are also looking forward to doing business with the marine and shipbuilding sector (Limited, 2022).

The special purpose vehicle (SPV) approached the Government of Tamil Nadu (GoTN) and GoI through Tamil Nadu Small Industries Development Corporation (TANSIDCO) and formed M/s Salem Steel Cluster Pvt Ltd; Salem in 2012 at Salem, Tamil Nadu. The performance of the SPFC, Salem is to be studied for the benefit of MSMEs and policymakers like the GoI and the GoTN. However, the SPFC is not studied yet and this leads to a study on the performance of SPFC before and after CDA.

Objectives of the Study

The objectives of the study are to:

1. 1.

Find the physical and financial performance of SPFC before and after CDA.

2.

Find the productivity of the SPFC by taking independent variables like number of units, employment, and production and dependent variable like turnover.

3.

Find the integrated performance of SPFC before and after CDA.

4.

Find business analytics models like Diagnostic Analytics, Descriptive Analysis, Inferential Analytics, Predictive Analytics, Prescriptive Analytics and Decision Analytics.

5.

Find the Difference in Difference (DID) on control variable where the units have not adopted CDA and experimental variables where the units have adopted CDA.

Methodology of the Study

The methodology adopted by collecting the primary data like number of units (Un), employment in numbers (En), production in crores (Pcr) and turnover in crores (Tcr) from the SPFC before and after CDA and analysing using Compound Annual Growth Rate (CAGR), Descriptive Analysis, Correlation Analysis, Trend Analysis, Regression Analysis and Structural Equation Modelling (SEM), T-Test, Value Chain Analysis and also using DID. Diagnostic Analytics, Descriptive Analytics, Inferential Analytics, Predictive Analytics, Prescriptive Analytics and Decision Analytics are studied in detail. The conceptual framework is given in Figure 2 and the performance study of Cluster Development Programme is given in Table 1.

Figure 2.

Conceptual Framework.

Table 1.

Integrated Performance Study of Cluster Development Programme.

Phase	UNIDO Methodology	Business Analytics	Statistical Analysis
1	Cluster selection	Diagnostic Analytics	SWOT Analysis
2	Selection of a Cluster Development Agent (CDA)	Descriptive Analytics	Compound Annual Growth Rate (CAGR), Descriptive Analysis
3	Cluster diagnostic	Inferential Analytics	Correlation Analysis
4	Vision building and action planning	Predictive Analytics	Trend Analysis, Regression Analysis
5	Implementation	Prescriptive Analytics	Structural Equation Modelling, Value Chain Analysis
6	Monitoring and evaluation	Decision Analytics	T-Test, Difference in Difference

Source: UNIDO methodology from NI-MSME (2006) and business analytics developed by the researcher.

Technical Analysis and Results

Diagnostic Analytics

The diagnostic study was conducted and the primary data collected are shown in Table 2.

Table 2.

Status of the Steel Product Fabrication Cluster (SPFC).

Name of the Cluster	M/s Salem Steel Cluster Pvt Ltd
Created in	2012
Category of products	Steel products (window, grill gate, truss work and panel board)
Extent of land	1.12 acres (73,534 sqft)
No. of units intended	100
No. of units created/established	40
No. of occupancies	40
Present board of management	4 directors
Regularity of conducting meetings	As per ROC regulations
No. of members	38 SPV members
Access to members/others (as per original plan)	Direct access
Existing facilities/maintenance	As per the sanction of GoI
Financial performance	Starving for working capital
Product refining, etc. as per original proposal	Converting stainless steel sheets into pipes and tubes, press components and laser cutting

Source: Primary data.

Note: ROC, Registrars of Companies; SPV, special purpose vehicle; GoI, Government of India.

The diagnostic study was conducted to know the various interventions needed for empowering the entrepreneurs engaged in SPFC in the areas of social, technological, infrastructure-related, financial and marketing for the successful promotion of the cluster.

Before adopting the CDA, the MSEs manufactured some parts, and the majority of them were outsourced.

SWOT Analysis

SWOT Analysis is given in Table 3.

Table 3.

SWOT Analysis.

	Strength	Weakness	Opportunity	Threats
Markets	Only local	Unorganised sectorLimited market	Export	Over-dependence on traders
Technology	Indigenous technology	The cluster members do not use modern and sophisticated technology. Lack of in-house testing facilities	Need for modern machinery for product making which is capable of making large number of homogeneous products in a short period	Lack of credit/limited resources for investment in the latest technology machineries
Inputs availability	Highly available	Obsolete machinery with cluster members—need for technology infusion	Good	Need for a reduction in the cost of production and wastage to compete with big players
Innovation capabilities	Highly qualified innovators	Many units still follow traditional methods	More innovative products	Global innovators
Skills	Highly skilled labour is available	Poor R&D activity—no new design development by cluster units Lack of training to work with modern machinery	Opportunity for skilled labourers	The establishment of a CFC for the cluster members will bring radical change for the cluster members in terms of cost reduction, quality improvement, increased productivity, product diversification through design creation, etc., and enable them to compete with organised players and create their own brand and market
Business environment	Good	Poor backward (raw material and advanced machinery) and forward integration (job order, design, wholesale market and export)	Opportunity to explore the global business environment	Investment in the latest technology machinery for various activities beyond the reach of these MSEs

Note: R&D, research and development; CFC, Common Facility Centre; MSEs, Micro and Small Enterprises.

Works at Cluster Units Before CDA

The process was done before CDA using machines like computer numerical control (CNC), lathe, welding, etc. Figure 3 also depicts a map indicating various linkages and actors that exist in a cluster.

Figure 3.

Map Indicating Various Linkages and Actors That Exist in a Cluster.

Under the Micro Small Enterprises Cluster Development Programme (MSE-CDP) of the Ministry of MSME, GoI, the SPFC has got funding from the GoI and the GoTN (MSME, 2022). The SPV also contributed and obtained a bank loan to finish the project of about ₹3.56 crore (Micro, 2022). The details of the project cost are given in Figure 4.

Figure 4.

Source: Computed data.

Need for a Common Facility Centre (CFC)

Facilities Created in CFC

To improve productivity, the following facilities were created in the cluster during 2017, also shown in Figure 5.

Figure 5.

Source: M/s Salem Steel Cluster Pvt Ltd.

The CFC has a laser cutting, shearing machine, power press, bending hydraulic, milling shaping and slitting machine for manufacturing steel pipes for sectors like agriculture, defence, engineering, etc. After CDA, there is an increase in the productivity of individual MSMEs.

Physical Performance

The physical performance is shown in Figure 6.

Figure 6.

Source: Computed data.

Financial Performance

The financial performance is shown in Figure 7.

Figure 7.

Source: Computed data.

Descriptive Analytics

The descriptive analysis is given in Table 4.

Table 4.

Descriptive Analysis.

		Un	En	Pcr	Tcr
N	Valid	4	4	4	4
	Missing	0	0	0	0
Mean		36.50	40.00	1.0600	1.2400
Standard error of mean		0.645	0.816	0.15943	0.19996
Median		36.50	40.00	1.0350	1.2250
Mode		35a	40	0.70a	0.77a
Standard deviation		1.291	1.633	0.31885	0.39992
Variance		1.667	2.667	0.102	0.160
Skewness		0.000	0.000	0.448	0.217
Standard error of skewness		1.014	1.014	1.014	1.014
Kurtosis		–1.200	1.500	1.043	1.010
Standard error of kurtosis		2.619	2.619	2.619	2.619
Range		3	4	0.77	0.97
Minimum		35	38	0.70	0.77
Maximum		38	42	1.47	1.74
Sum		146	160	4.24	4.96

Source: Computed data.

Note: ^aMultiple modes exist. The smallest value is shown.

In descriptive analysis, what happened is there was an increase in the mean value of number of units, employment, production and turnover after the CDA.

Correlation Analysis/Inferential Analytics

The correlation analysis is given in Table 5.

Table 5.

Correlations Analysis.

		Un	En	Pcr	Tcr
Un	Pearson correlation	1	0.949	0.980*	0.981*
	Sig. (two-tailed)		0.051	0.020	0.019
	N	4	4	4	4
En	Pearson correlation	0.949	1	0.986*	0.990**
	Sig. (two-tailed)	0.051		0.014	0.010
	N	4	4	4	4
Pcr	Pearson correlation	0.980*	0.986*	1	0.999**
	Sig. (two-tailed)	0.020	0.014		0.001
	N	4	4	4	4
Tcr	Pearson correlation	0.981*	0.990**	0.999**	1
	Sig. (two-tailed)	0.019	0.010	0.001
	N	4	4	4	4

Source: Computed data.

Notes: ^*Correlation is significant at the 0.05 level (two-tailed).

^**Correlation is significant at the 0.01 level (two-tailed).

H_o1: There is no relationship between number of units and turnover (Rejected: r = 0.949, p = .05).

H_a1: There is a relationship between number of units and turnover (Accepted).

H_o2: There is no relationship between employment and turnover (Rejected: r = 0.98, p = .02).

H_a2: There is a relationship between employment and turnover (Accepted).

H_o3: There is no relationship between production and turnover (Rejected: r = 0.98, p = .19).

H_a3: There is a relationship between production and turnover (Accepted).

Trend Analysis

$$\text{Un}=34+T\left[p=.01,R^2=1\right]$$

(1)

The annual average increase in number of units is 1.

$$\text{ En }=37+1.2T\left[p=.51,R^2=0.90\right]$$

(2)

The annual average increase in employment is 1.

$$\text{Pcr}=0.46+0.24T\left[p=.02,R^2=0.96\right]$$

(3)

The annual average increase in production is ₹0.24 crores.

$$\text{Tcr}=0.48+0.304T\left[p=.02,R^2=0.96\right]$$

(4)

The annual average increase in turnover is ₹0.95 crores.

In diagnostic analysis, why it happened was due to the usage of the CFC, there is an annual average increase in number of units, employment, production and turnover in the after-CDA.

As per Table 6, there is increase in no. of units, employment, production and turnover after Cluster Development Approach when compared to before Cluster Development Approach.

Table 6.

Paired Samples Statistics.

		Mean	N	Standard Deviation	Standard Error Mean
Pair 1	Una	37.0000a	3	1.00000	0.57735
	Unb	36.0000a	3	1.00000	0.57735
Pair 2	Ena	40.6667	3	1.15470	0.66667
	Enb	39.3333	3	1.15470	0.66667
Pair 3	Pna	1.1800	3	0.25710	0.14844
	Pnb	0.9233	3	0.20108	0.11609
Pair 4	Tna	1.3967	3	0.30436	0.17572
	Tnb	1.0733	3	0.27062	0.15624

Source: Computed data.

Note: ^aThe correlation and t cannot be computed because the standard error of the difference is 0.

As per Table 7, it is revealed that there is moderate relationship between before and after Cluster Development Approach on number of employments. However, there is high degree of relationship between before and after Cluster Development Approach on production and there is also high degree of relationship between before and after cluster development approach on turnover.

Table 7.

Paired Samples Correlations.

		N	Correlation	Sig.
Pair 2	Ena and Enb	3	0.500	0.667
Pair 3	Pna and Pnb	3	0.850	0.353
Pair 4	Tna and Tnb	3	0.831	0.375

Source: Computed data.

As per Table 8, the T-Test reveals that there is no significant difference between before and after cluster development approach on number of employments. However, there is significant difference between before and after cluster development approach on production and there is significant difference between before and after cluster development approach on turnover.

Table 8.

Paired Samples Test.

		Paired Differences					t	df	Sig. (Two-tailed)
		Mean	Standard Deviation	Standard Error Mean	95% Confidence Interval of the Difference
					Lower	Upper
Pair 2	Ena–Enb	1.33	1.15	0.66	–1.5	4.20	2.000	2	0.184
Pair 3	Pna–Pnb	0.25	0.13	0.07	–0.08	0.59	3.257	2	0.083
Pair 4	Tna–Tnb	0.32	0.17	0.09	–0.09	0.74	3.292	2	0.081

Source: Computed data.

Table 9 exposes the Value Chain Analysis of the cluster wherein the profit/ turnover is calculated based on infrastructure cost, technology cost, procurement cost, production cost and marketing cost.

Table 9.

Value Chain Analysis of a Steel Product Fabrication Cluster (SPFC).

		Infrastructure (Ii)	Cluster has taken the land and constructed building for CFC. Arrange funds, carry out accounting, compliance, legal (statutory) and payroll functions and perform other general management tasks. Information resources for discrete activities
Support activities	Infrastructure interrelationships	Human resource management (Hi)	Supervising, clearing agents, training stores and personnel	Supervising/training workers job workers	Supervising delivery personnel	Supervising/training advertising and promotion personnel	Supervising sales agents- employees	Profit/turnover (PRi)
	Technology interrelationships	Technology development (Ti)	Improve inventory management and storing methods	Develop new products/designs and cost-effective production methods	Track delivery to retailers	New promotional material, media- channels, sales promotion tools	Improve method to track product movement off the shelf
	Procurement interrelationships	Procurement (Ri)	Transportation/hire containers, etc.	Buy/import new steel, accessories, etc.	Transportation rent trucks, etc.	Buy space at retail counters (POP), shop in shop counters, media time, etc.	Tools for tracking sales off the retailers’ shelves
			Primary activities	Transport raw material/supplies to factory/warehousing	Manufacturing/processing	Transport products to retailers	Develop brand image		Ensure timely delivery of re-order
				Inbound logistics	Operations	Outbound logistics	Marketing/sales		Service
				Production interrelationships (Pi)			Marketing interrelationships (Mi)

SEM

The SEM is given in Figure 8..

Figure 8.

Source: Computed data.

$$\text{Tcr}=-489+0.064\text{Un}+0.079\text{En}+0.600\text{Pcr}\left[p=.00<.05,R^2=1.00\right]$$

(5)

For one unit increase in number of units, turnover increases by 0.064, for one unit increase in employment, turnover increases by 0.079 and for one unit increase in production, turnover increases by 0.60. The regression graphs are given in Figure 9.

Figure 9.

Source: Computed data.

Regression Analysis/Predictive Analytics

$$\text{Pcr}=-7.355+0.11\text{Un}+0.11\text{En}\left[p=.91>.05,R^2=0.992\right]$$

(6)

For one unit increase in units, production increases by 0.11 units and one unit increase in employment, production increases by 0.11 units.

$$\text{Tcr}=-0.088+1.25\text{Pcr}\left[p=.01<.05,R^2=0.998\right]$$

(7)

For one unit increase in production, turnover increases by 1.25 units.

In predictive analysis, what will happen is there is/there will be an increase in number of units, employment, production and turnover after the CDA.

T-Test Paired Two Samples for Means

H_o4: Una = Unb, p = .01 < .05 (Rejected)

H_a4: Una # Unb (Accepted) Una > Unb

H_o5: Ena = Enb, p = .184 > .05 (Accepted)

H_a5: Ena # Enb (Rejected); however, Ena > Enb and Mean Ena > Enb

H_o6: Pcra = Pcrb, p = .083 < .10 (Rejected)

H_a6: Pcra # Pcrb (Accepted) Pcra > Pcrb

H_o7: Tcra = Tcrb, p = .081 < .10 (Rejected)

H_a7: Tcrb # Tcrb (Accepted) Tcra > Tcrb

Prescriptive Analytics

In prescriptive analysis, how can we make it happen by networking all micro manufacturing steel fabrication enterprises so as to increase individual units on employment, production and turnover after the CDA which is proved in DID. Figure 10 gives a cluster map indicating various linkages and actors that exist in a cluster and Table 9 exposes Value Chain Analysis of an SPFC

Figure 10.

Cluster Map Indicating Various Linkages and Actors that Exist in a Cluster.

$$PRi=To-Ii+Hi+Ti+Ri+Pi+Mi$$

(8)

As per equation (8), there is an increase in profit of individual unit after CDA.

DID/Decision Analytics

The DID is used to find the difference between the control variable (firms that have not adopted CDA) and the experimental variable (firms that have adopted CDA) which is given in Table 10.

Table 10.

Difference in Difference (DID).

Model	Equation	p	R 2	Hypothesis
1	Un = 35 + 2 T + t + 4.4E-16 Tt	.00	1.00	Significant
2	En = 38 + 2 T+ 2t + 0 Tt	.00	1.00	Significant
3	Pcr = 0.7 + 0.39 T + 0.28t + 0.1 Tt	.00	1.00	Significant
4	Tcr = 0.77 + 0.52 T + 0.39t + 0.06 Tt	.00	1.00	Significant

$$\text{Un}=35+2T+t+4.4E-16Tt,p=.00,R^2=1.00$$

(9)

As per equation (9), there is not much difference in number of units after CDA.

$$\text{En}=38+2T+2t+0Tt,p=.00,R^2=1.00$$

(10)

As per equation (10), there is not much difference in employment after CDA.

$$\text{Pcr}=0.7+0.39T+0.28t+0.1Tt,p=.00,R^2=1.00a$$

(11)

As per equation (11), there is difference of decrease of ₹0.1 lakhs on profit after CDA.

$$\text{ Tcr }=0.77+0.52T+0.39t+0.06Ttp=.00,R^2=1.00$$

(12)

As per equation (12), there is difference of increase of ₹0.06 lakhs on turnover after CDA.

As per table 10, Model 1 to 4 are significant, where there is significant increase in no. of units, employment, production and turnover after Cluster Development Approach.

Findings, Suggestions and Conclusion

A study was conducted to find the productivity of SPFC before and after CDA. There is an increase in CAGR after CDA. There is an increase in mean value after CDA. There exists a strong relationship between dependent variables like turnover and independent variables like number of units, employment and production. There is an annual average increase in number of units, employment, production and turnover. There is an increase in number of units, employment, production and turnover after CDA.

There is DID between controlled units which have not adopted CDA and experimental units which have adopted CDA, where there is an increase in number of units, employment, profit and turnover. There is cost reduction in individual unit after CDA. Due to CDA, steel fabrication products are diversified, new design has been developed, direct export by cluster units has taken place and brand creation has been developed in steel fabrication products. To conclude, by using business analytics techniques like Diagnostic Analytics, Descriptive Analytics, Predictive Analytics, Prescriptive Analytics and Decision Analysis, it is found that there is an increase in turnover and productivity, thereby SDGs of 1, 4, 5, 8 and 9 are achieved.