Conjoint Analysis SPSS Syntax Tutorial

Every product is a bundle of trade-offs. A customer choosing a phone weighs battery life against price, camera quality against brand. The hard question for any business is: which of these features actually drives the buying decision? Conjoint analysis answers exactly that – and SPSS gives you a complete set of syntax commands to run it end to end.

This tutorial walks through conjoint analysis in SPSS using syntax. We cover the three core commands – ORTHOPLAN, PLANCARDS, and CONJOINT – with working examples you can adapt to your own study. By the end, you will understand how to design the experiment, generate cards, and read the part-worth utilities that reveal what customers truly value.

This is a practical, hands-on guide aimed at researchers who want to work directly with SPSS syntax rather than menus.

What Is Conjoint Analysis?

Conjoint analysis is a market research technique that measures how people value the individual features of a product or service. Instead of asking customers what they want directly, it shows them realistic product profiles and decodes the trade-offs they make.

It answers questions like:

• Which product attributes matter most to customers?
• Which levels of each attribute are most preferred?
• What is the predicted market share for different product designs?

The core output is a set of part-worth utilities – numerical values showing how much each attribute level contributes to overall preference. From these, every other result is derived, including attribute importance and market simulations.

This makes conjoint analysis one of the most powerful tools for product design, pricing, and feature prioritisation.

Why Use SPSS Syntax for Conjoint Analysis?

SPSS offers conjoint analysis through both menus and syntax. For this technique, syntax has real advantages:

• Repeatability – save and rerun your exact design every time
• Control – set factors, levels, and holdouts precisely
• Documentation – your syntax is a clear record of the study design
• Speed – adjust and rerun without clicking through dialogue boxes

Importantly, the CONJOINT command is only available through syntax – there is no menu equivalent. So if you want to complete a full conjoint study in SPSS, learning the syntax is essential, not optional.

The Three SPSS Conjoint Commands

SPSS Conjoint works through three procedures, each handling one stage of the workflow:

• ORTHOPLAN – generates an efficient experimental design (the profiles)
• PLANCARDS – turns those profiles into readable cards for respondents
• CONJOINT – analyses the responses and calculates utilities

Think of it as a pipeline: design, present, analyse. Let us walk through each command in order.

Step 1: Generate the Design with ORTHOPLAN

A full conjoint study with several attributes can create dozens or hundreds of possible product combinations. Asking respondents to rate all of them is impractical. ORTHOPLAN solves this by generating an orthogonal design – a fraction of all combinations that still captures the information you need.

Example Syntax

ORTHOPLAN

  /FACTORS=

    BATTERY ‘Battery Life’ (8 12 24)

    CAMERA ‘Camera’ (1 ‘Basic’ 2 ‘Standard’ 3 ‘Premium’)

    PRICE ‘Price’ (20000 40000 60000)

  /MINIMUM=9

  /HOLDOUT=4

  /OUTFILE=’phone_plan.sav’.

What Each Part Does

• /FACTORS – defines your attributes and their levels. Here, three attributes each have three levels.
• /MINIMUM – the minimum number of profiles to generate. SPSS creates the smallest efficient set.
• /HOLDOUT – extra profiles that respondents rate but that are not used to estimate utilities. They test how well the model predicts.
• /OUTFILE – saves the design to a plan file you will use in later steps.

The result is a manageable set of product profiles instead of every possible combination. For example, a 3x3x3 design with 27 combinations might reduce to just 9 profiles – far easier for respondents to handle while keeping the analysis valid.

Step 2: Create Cards with PLANCARDS

Once the design exists, respondents need to see the profiles in a clear, usable form. PLANCARDS converts the abstract plan into readable cards that respondents can sort and rank.

Example Syntax

PLANCARDS

  /TITLE=’Smartphone Profile #)CARD’

  /FOOTER=’RANK:’

  /OUTFILE=’phone_cards.txt’.

What Each Part Does

• /TITLE – the heading on each card. The )CARD element inserts the profile number automatically.
• /FOOTER – text at the bottom of each card, such as a space for respondents to write their rank.
• /OUTFILE – writes the cards to an external file for printing or distribution.

PLANCARDS produces two useful formats: a compact listing for the researcher (showing all profiles including holdouts) and clean individual cards for respondents (with holdout markers hidden). This is what respondents actually rank during data collection.

Step 3: Collect and Structure the Data

Before running the analysis, respondents rank or rate the profiles, and you enter their responses into SPSS. The data file structure matters here.

Key requirements:

• Each respondent’s rankings are entered as variables
• The sequence of responses must match the profile order in the plan file
• Rankings are usually entered with the most preferred profile first

This is the data collection stage, and its quality shapes the final result. Clean, correctly structured input is essential – a mismatch between profile order and response order produces meaningless utilities. Careful data quality practices at this point protect the entire analysis.

Step 4: Run the Analysis with CONJOINT

The CONJOINT command performs the analysis. It is a specially tailored version of regression that calculates part-worth utilities from the rankings. Remember, this command is syntax-only.

Example Syntax

CONJOINT

  PLAN=’phone_plan.sav’

  /DATA=’phone_responses.sav’

  /SEQUENCE=PREF1 TO PREF9

  /SUBJECT=ID

  /FACTORS=

    BATTERY (LINEAR MORE)

    CAMERA (DISCRETE)

    PRICE (LINEAR LESS)

  /PRINT=ALL.

What Each Part Does

• PLAN – the design file generated by ORTHOPLAN.
• /DATA – the file containing respondent rankings.
• /SEQUENCE – lists the variables holding the preference data, in profile order.
• /SUBJECT – identifies each respondent.
• /FACTORS – specifies the expected relationship for each attribute:
  • DISCRETE – no assumed order (use for categories like brand or colour)
  • LINEAR MORE – higher levels are expected to be preferred (more battery is better)
  • LINEAR LESS – lower levels are expected to be preferred (lower price is better)
• /PRINT=ALL – produces the full set of output.

Specifying the expected relationship helps SPSS interpret the results correctly and flags any reversals where respondents behave unexpectedly.

How to Read the CONJOINT Output

The output tells you what customers value. The key elements to focus on:

Part-Worth Utilities

These are the core results. Each attribute level gets a utility value showing how much it contributes to preference. Higher utilities mean stronger preference. For example, a high utility for “24-hour battery” tells you that feature drives choice.

Importance Values

These show the relative importance of each attribute as a percentage. They reveal which factors weigh most heavily in the decision. If price has 50% importance and camera has 20%, price is the dominant driver.

Correlation Statistics

SPSS reports correlations (such as Pearson’s R and Kendall’s tau) that measure how well the model fits the observed rankings. Higher values indicate a stronger, more reliable model.

Holdout Validation

The correlations for holdout profiles show how well the model predicts profiles it did not use to estimate utilities. Strong holdout performance gives confidence that the model generalises beyond the original data.

Together, these turn raw rankings into clear, actionable insights about product design.

A Practical Example: Designing a Smartphone

Imagine a study on smartphone preferences using the three attributes above – battery, camera, and price. After running the full workflow, the output might show:

• Price has the highest importance at 45%
• Battery life follows at 35%
• Camera quality contributes 20%
• Within price, the lowest price has the highest utility
• Within battery, the 24-hour option is strongly preferred

The business takeaway is direct: customers are most sensitive to price, value long battery life highly, and care least about camera tier. A product team can use this to design a phone that wins on the attributes that matter most – and even simulate the market share of different designs.

Common Mistakes to Avoid

A few errors trip up many first-time conjoint users:

• Mismatched sequence – response order not matching the plan file produces invalid results
• Too many attributes – overloading respondents reduces data quality
• Skipping holdouts – without them, you cannot validate the model
• Wrong factor relationship – labelling price as DISCRETE instead of LINEAR LESS loses useful structure
• Ignoring data cleaning – messy rankings distort utilities

Avoiding these keeps your analysis valid and your conclusions trustworthy.

Industry Applications

Conjoint analysis is used wherever feature and pricing decisions matter:

• Consumer goods (FMCG): optimising product features and pack design
• Technology: prioritising features for new devices and software
• Financial services: designing product bundles and pricing tiers
• Automotive: balancing features against price points
• Healthcare: understanding patient and provider preferences

In each case, conjoint analysis turns customer trade-offs into clear guidance for product and pricing strategy – strong market intelligence for decision-makers.

How Linkinfotech Supports Conjoint Studies

Conjoint analysis is powerful but demanding. The design, card creation, data structuring, and syntax all need precision – and the results are only as reliable as the data behind them. Linkinfotech operates as a global research operations and technology partner, supporting market research firms and enterprise teams across the full conjoint workflow.

Our role spans every stage:

• Study design – building efficient orthogonal designs for your attributes
• Survey programming and data collection – structured, multi-mode collection across web, phone, and field
• Data processing and validation – analysis-ready datasets with correct profile sequencing
• Statistical analysis support – including conjoint, utility estimation, and simulations
• Real-time dashboards – clear visibility into preferences and importance scores
• Secure, scalable operations – ISO-certified processes and compliant data handling

Because we manage the foundation – clean, secure, well-structured data – the conjoint analysis built on top becomes far more dependable. That is what turns SPSS syntax output into product decisions you can confidently act on.

Final Thoughts

Conjoint analysis in SPSS gives you a complete, syntax-driven way to understand what customers truly value. The workflow is clear: design with ORTHOPLAN, present with PLANCARDS, and analyse with CONJOINT. Read the part-worth utilities and importance scores, and you have direct guidance for product design, pricing, and feature decisions.

As with any advanced technique, the quality of the result depends on the quality of the design and data. Clean, well-structured input is what makes conjoint analysis reliable.

If you want to run conjoint studies backed by precise design, clean data, and secure operations, Linkinfotech can help you build research processes that are accurate, scalable, and ready for confident decision-making.

Frequently Asked Questions