Porter Value Chain Model: Components, Application, and Testing
What is the Porter Value Chain Model
The Porter Value Chain Model is an analytical framework that helps you examine how a business creates value for customers through its activities, from receiving inputs to delivering products or services and providing after-sales support. Michael Porter introduced this framework in research on competitive advantage. In a thesis, the model helps turn a broad question such as “How does a business create competitive advantage?” into groups of factors that can be observed and measured with a questionnaire.
The value chain has two groups of activities. The first group is primary activities, which are directly related to creating, selling, and delivering products. The second group is support activities, which enable the primary activities to operate effectively. A business can create competitive advantage by reducing costs, increasing perceived value, or combining both approaches.
You should distinguish this model from What are Porter's Five Competitive Forces. The Five Forces framework examines the competitive structure of an industry, while the value chain focuses on activities inside the business. If your topic requires both the external environment and internal capabilities, you can place the two frameworks in separate sections instead of mixing their variables into one scale.
Components of the model
Each activity in the value chain can be treated as a group of independent variables, an aspect of organizational capability, or a set of indicators used to describe the current situation. You do not have to include all nine activities in a quantitative model. Your choice of variables should depend on the industry, the research question, and your ability to collect data.
| Group | Component | What to examine in the business |
|---|---|---|
| Primary activities | Inbound logistics | Receiving, storing, and distributing raw materials or inputs |
| Primary activities | Operations | Processes that transform inputs into products or services |
| Primary activities | Outbound logistics | Storing products, processing orders, and delivering products to customers |
| Primary activities | Marketing and sales | Pricing, communication, selling, and distribution channel management |
| Primary activities | Service | Warranties, support, complaint handling, and after-sales care |
| Support activities | Procurement | Purchasing inputs, equipment, technology, and necessary services |
| Support activities | Technology development | Technology, data, product innovation, and process improvement |
| Support activities | Human resource management | Recruitment, training, evaluation, and employee retention |
| Support activities | Firm infrastructure | Finance, legal affairs, management, planning, and control systems |
For example, in a study of business performance in retail companies, “Marketing and sales” could be represented by customer reach, the clarity of product information, and the effectiveness of sales channel management. “Service” could concern response speed, complaint handling, and post-purchase support. These two groups could affect a dependent variable such as competitive advantage or business performance.
If you are looking for a tool to analyze a product portfolio, What is the BCG Model is more suitable for questions about market share and growth rate. If you need to evaluate objectives and performance indicators, What is the Balanced Scorecard offers a different approach. These models can complement one another, but you should not describe all of them as components of the Porter value chain.
The original model and common extensions
The original model divides activities according to the logic of value creation, from inputs to outputs and after-sales service. When you apply it, retain the connection between each activity, the costs it creates, and the value received by customers. A business may have effective marketing but still lose its advantage if its logistics are poor or its operating process creates too many errors.
In applied research, the model is commonly extended in three directions. The first is to combine activities into higher-level capabilities, such as operating capability, marketing capability, and service capability. The second is to add outcome variables such as competitive advantage, business performance, or customer-perceived value. The third is to add contextual variables such as business size, competitive intensity, or digital transformation.
You need to state clearly whether you are using the value chain as an analytical framework or as the foundation for a quantitative model. If you are using it only to analyze the current situation, you can present each activity with descriptive data. If you are testing hypotheses, each activity group needs specific indicators, a scale, and a clearly stated hypothesized relationship.
A common extension treats technology as a factor affecting several activities rather than placing it only under “Technology development”. For example, a warehouse management system may improve inbound logistics, operations, and outbound logistics at the same time. You then need to decide whether technology is a separate independent variable, a mediator, or only an indicator of operating capability. That decision must appear in the conceptual model and in the rationale for the hypotheses.
Common scales for each construct
The Porter Value Chain Model does not provide one fixed set of items for every industry. You need to find scales that fit your context and adjust the wording. For a quantitative thesis, start with constructs that have a clear scope and avoid overly broad variables such as “the company’s entire value chain”.
| Suggested construct | Measurement content | Reference source or foundation |
|---|---|---|
| Inbound logistics capability | Ability to receive, control, and supply inputs on time | Porter value chain framework, adjusted to the industry |
| Operating capability | Stability, efficiency, and ability to improve processes | Porter value chain framework, adjusted to the industry |
| Marketing and sales capability | Understanding customers, managing channels, and communicating value | Can be related to the service quality foundation of (Parasuraman et al., 1988) |
| Service capability | Response speed, support, and problem resolution after a transaction | Can be related to the service quality foundation of (Parasuraman et al., 1988) |
| Technology capability | Ability to apply technology to improve products and processes | Indicators developed according to the research context |
| Competitive advantage | Ability to create differentiated value or operate at an appropriate cost | Outcome construct, defined according to the topic |
| Business performance | Financial or non-financial results of the business | Indicators selected according to what respondents can answer |
Likert scales are commonly used for questionnaire items about attitudes and perceptions, based on the foundation provided by (Likert, 1932). You should use the same direction of scoring for all items. If you include reverse-coded items, recode them before running Cronbach's Alpha or clearly document the treatment in your dataset.
A good item addresses one specific issue. The statement “The business manages its value chain well and creates high competitive advantage” combines several ideas, which makes it difficult for respondents to choose one level of agreement. You could separate it into statements such as “The business controls delivery time effectively” and “The business responds quickly when customers encounter a problem”. These are examples for developing a questionnaire, not a scale that has already been validated for your topic.
Applying the model to your topic
Before drawing the model, write a research question that can be tested. For example: “How do value chain activities affect the competitive advantage of retail businesses in Ho Chi Minh City?” From there, select four groups of factors that can be measured through a questionnaire instead of placing all nine activities into a model that is too broad.
An illustrative model could include operating capability, marketing and sales capability, service capability, and technology capability as independent variables. Competitive advantage would be the dependent variable. The conceptual diagram should show four arrows running from the independent variables to competitive advantage. This is illustrative data and an illustrative model, not the result of a real study.
You can write the hypotheses as follows:
- H1: Operating capability has a positive effect on the competitive advantage of retail businesses.
- H2: Marketing and sales capability has a positive effect on the competitive advantage of retail businesses.
- H3: Service capability has a positive effect on the competitive advantage of retail businesses.
- H4: Technology capability has a positive effect on the competitive advantage of retail businesses.
These hypotheses state the expected direction of the relationships but do not prove that the relationships exist. You need to identify the respondents, unit of analysis, and data source before designing the questionnaire. If respondents are employees, the items should concern activities they are able to evaluate. If respondents are customers, variables such as internal logistics or human resource management may concern information they do not have.
You could adapt the following wording for your thesis: “Based on the Porter Value Chain Model, this study proposes that [factor 1], [factor 2], [factor 3], and [factor 4] affect [dependent variable] in the context of [research population and location]. The hypotheses are tested using survey data from [number] respondents.” Replace every bracketed section with your actual information.
How to test the model with SPSS or SmartPLS
SPSS is suitable when you use composite variables to run linear regression, test reliability, and conduct EFA. The usual process includes checking the data, running Cronbach's Alpha, running EFA, calculating representative scores for each factor, and running regression. In SPSS, go to Analyze > Scale > Reliability Analysis to check the scale, then Analyze > Dimension Reduction > Factor to run EFA, and Analyze > Regression > Linear to test the main relationships.
If your model contains several latent variables, each measured by multiple indicators, SmartPLS may be more convenient. Import the .csv file, create the constructs, draw arrows between the constructs, and run PLS Algorithm. Then use Bootstrapping to examine the path coefficient, t-value, and p-value. When reporting PLS-SEM, assess the measurement model before the structural model, following the two-step approach of (Anderson and Gerbing, 1988).
The following is illustrative output.
| Check | Statistic to examine | Reference threshold and source |
|---|---|---|
| Scale reliability | Cronbach's Alpha | 0.7 or above, (Nunnally, 1978); exploratory research may consider 0.6 according to (Hair et al., 2010) |
| Item-total correlation | Corrected Item-Total Correlation | 0.3 or above, (Nunnally and Bernstein, 1994) |
| EFA conditions | KMO and Bartlett's Test | KMO from 0.5, Bartlett significant below 0.05, (Kaiser, 1974) |
| EFA | Factor loading | 0.5 or above, (Hair et al., 2010) |
| EFA | Total Variance Explained | 50% or above, (Hair et al., 2010) |
| PLS-SEM | Outer loading | 0.7 or above, (Chin, 1998) or (Hair et al., 2022) |
| PLS-SEM | CR and AVE | CR from 0.7, AVE from 0.5, (Fornell and Larcker, 1981) |
| Discriminant validity | HTMT | Below 0.85 or 0.90 for closely related constructs, (Henseler et al., 2015) |
| Multicollinearity | VIF | Below 5, (Hair et al., 2019) |
| Explanatory power | R² | 0.75 substantial, 0.50 moderate, 0.25 weak, (Hair et al., 2011) |
| Prediction | Q² | Greater than 0, (Stone, 1974) or (Hair et al., 2022) |
The values in the table are reference points, not automatic reasons to delete an item. If an outer loading is low, review the item content, cross-loadings, CR, and AVE before deciding. Deleting items only to make the statistics look better can remove the theoretical meaning of the model.
SmartPLS does not use CFI, TLI, and RMSEA for evaluation in the same way as CB-SEM. These indices belong to covariance-based model evaluation and can be consulted through (Hu and Bentler, 1999) in a CB-SEM context. If you run SmartPLS, focus on outer loading, CR, AVE, HTMT, VIF, R², Q², and path coefficients.
Common mistakes when using this model
The first mistake is treating the nine activities as nine mandatory independent variables. With a small sample, a model with too many variables and indicators becomes difficult to test, and respondents may not have enough information to evaluate every activity. Select the groups that fit your industry and research objective.
The second mistake is using the activity name as the measured variable. “Operations” is only a conceptual label. You still need specific statements about the capability, quality, or effectiveness of that activity. Each statement should have a clear subject, content, and context.
The third mistake is mixing the model with the conclusion. An arrow from service capability to competitive advantage is a hypothesis to test, not evidence that service certainly creates an advantage. In the results chapter, report the coefficient, significance level, and confidence interval from the actual data.
The fourth mistake is copying items unchanged from another industry. A manufacturing logistics scale may not fit a service business. Review the terminology, respondent group, and observability of each indicator before distributing the questionnaire.
The final mistake is using the Porter value chain to answer every strategic question. If your topic concerns competitors, suppliers, and substitute products, also consider the SWOT Model or the Five Forces framework. If your topic focuses on relationships with stakeholder groups, What are stakeholders can help you choose the right analytical direction.
Frequently asked questions
Is the Porter Value Chain Model a complete quantitative model?
No. It is a theoretical framework for classifying value-creating activities and suggesting how to build a model. For quantitative research, you still need to select variables, develop indicators, collect data, and test the relationships between variables.
Should I use the Porter Value Chain Model or Porter's Five Competitive Forces?
Choose according to your research question. The value chain fits activities inside the business, while the Five Forces framework fits pressure from competitors, customers, suppliers, substitute products, and potential entrants. Both frameworks can appear in the theoretical foundation, but you need to keep their conceptual boundaries clear.
Do I need to include all nine activities in the questionnaire?
No. You can select the activities that are directly related to the industry and dependent variable. Explain your selection criteria, check whether respondents can answer the items, and avoid adding variables simply because they appear in the original model.
Should I run SPSS or SmartPLS for the Porter Value Chain Model?
SPSS is suitable when you want to test scales with Cronbach's Alpha and EFA before running regression with factor scores. SmartPLS is suitable when you model latent variables and indicators directly and need to assess the outer model and structural model. Your methods section should state clearly why you selected the software.
Should I delete an item immediately if Cronbach's Alpha or EFA does not meet the threshold?
First check the coding, missing data, reverse-coded items, and item content. Then review the Corrected Item-Total Correlation, factor loading, and theoretical meaning. If you delete an item, record the reason and the number of runs instead of deleting several items simply to reach a preferred value.
Open your data file, identify the industry context, select only a few activity groups that can be measured, and prepare the indicator table before running the tests. If you need to run the model on your own .sav or .csv file, you can use M4 DoThesis analysis.