# Competitive Analysis (Conjoint)

How to set up a Competitive Analysis and read its report, which measures how your preference share, revenue index and switching to competitors change as you change your price while competitor prices stay fixed.

Competitive Analysis (Conjoint) measures what happens when your product sits on the same shelf as its competitors and only your price changes. Conjoint is the general name for methods that measure preference by asking consumers to choose between options. At each price you test, the synthetic consumers (personas) choose your product, one of the competitors, or none of them (**Neither**). An example question: "If we raise our price from ₺800 to ₺1,000, how much share will we lose, and to which competitor?"

## When to use it

It suits questions like these:

- With competitors' prices known, how does your price affect your preference share?
- As your price rises, which competitor picks up the share you lose, and how much of it shifts to not buying at all?
- On a shelf with competitors, which of the prices you test brings in the highest revenue?
- Which consumer groups are more price-sensitive, and which switch to a competitor more easily?

To measure willingness to pay (WTP) alone, without competitors, [Price Analysis (WTP)](/en/docs/research/price-analysis) is a better fit; to measure how much a feature adds to the price, use [Feature Price Impact](/en/docs/research/feature-price-impact). If you can't decide, see [Choosing the right method](/en/docs/research/choosing-a-method).

### Two conjoint types

On the design page, you choose one of two conjoint types. Choice-Based Conjoint (CBC) is selected by default.

| Topic | Choice-Based Conjoint (CBC) | Dynamic Conjoint |
| --- | --- | --- |
| How it works | At each price point, every persona sees the shelf and makes a single choice. | Each persona is first asked willingness-to-pay (WTP) questions about your product. Demand and switching to competitors at each price are then calculated from these answers with a simulation. |
| Shares in the report | A count of the choices the personas made | Modelled preference probabilities |
| Extra report items | None | **Price Comparison Table**, **Segment-Based Optimal Price** table, shopping tendency breakdowns |
| When to choose it | When you want a directly observed shelf choice | When you want demand modelled from each persona's own willingness to pay |

In CBC, the order of the products is shuffled in every question so that being at the front of the shelf doesn't influence the choice; **Neither** is always last.

## Before you start

- Prepare the names of up to 5 competitors and their current shelf prices.
- Decide the lowest price, the highest price and the step you'll test for your product.
- Competitive Analysis is available in the **General Research** and **Fashion** verticals. Populations aren't shared between verticals, so set up the study in the vertical your population belongs to.
- Check your balance. In the **New Research** flow, you need at least 500 credits to create a General Research study, and at least 50 credits to create a new population in Fashion.
- If you plan to look at results by segment, choose a large population; small segments aren't shown in the charts.

## Setting up the study

### Choosing the method and population

1. In the sidebar, click **New Research**.
2. Choose the **General Research** or **Fashion** card.
3. Choose the **Competitive Analysis** card in General Research, or the **Rekabet Analizi (Conjoint)** (Competitive Analysis) card in Fashion.
4. Select an existing population or create a new one.

Once you've chosen the population, the design page opens: **Competition Analysis (Conjoint)** in General Research, **Rekabet Analizi (Conjoint)** in Fashion. The population step is covered in detail on [Starting a new study](/en/docs/research/new-study).

### Entering the product, competitors and prices

1. On the **Select Research Type** card, choose **Choice Based Conjoint (CBC)** or **Dynamic Conjoint**.
2. To start from a saved survey, choose one on the **Select from Saved Surveys** card. The list shows only surveys saved with the selected conjoint type. If you click **Remove selection**, what you had entered before comes back.
3. In the **Your Product** section, enter the **Product Name**.
4. In the **Product Description** field, enter the pack size, contents and key features. This text is shown to the personas.
5. In the **Competitors** section, enter the **Competitor Product** name and the **Competitor Price (₺)**. You can also fill in the **Competitor description (optional)** field.
6. To add another competitor, click **Add Competitor** and fill in the same fields. When there's more than one competitor, you can delete one with the bin icon on its row.
7. In the **Price Ladder** section, enter the **Minimum Price (₺)**, **Maximum Price (₺)** and **Price Step (₺)** values.
8. Check the prices generated under **Price Points (… items)**. To test a price that isn't on the ladder, click **Add Custom Price**.
9. Click **Next Step**. In General Research, click **Continue** in the confirmation dialog that opens.

At this point, the app runs a market price check based on your product name and price range. If the range looks out of line with the market, the **Price might not be realistic** dialog opens with **Typical range:** and **Estimated market price:** values. To fix the range, click **Cancel**. To continue with the same range, tick **I want to proceed anyway** and click **Continue**.

### Previewing and starting

1. On the **Research Preview** page, check the **Person Count**, **Price Point**, **Competitor Count** and **Credit Usage** values on the **Experiment Summary** card.
2. Click **Start**.
3. In the **Are you sure you want to continue?** dialog, click **Continue**.

You can leave the progress screen; the study keeps running in the background. If you stay on the screen, you're taken to the report page when the study finishes. Error states are covered on [Starting a new study](/en/docs/research/new-study).

[Argus](/en/docs/argus/working-with-argus) can also fill in this form for you. You can rerun a saved competitive survey with [Quick Research](/en/docs/research/quick-research).

## Settings and limits

| Setting | Limit |
| --- | --- |
| Conjoint type | **Choice Based Conjoint (CBC)** (default) or **Dynamic Conjoint** |
| **Product Name** | Required, up to 255 characters |
| **Product Description** | Optional; up to 1,000 characters in General Research, up to 2,000 in Fashion |
| Number of competitors | At least 1, at most 5 |
| Competitor name and price | Both required; name up to 255 characters. At least one competitor's price must be greater than 0 |
| Competitor description | Optional; same character limits as the product description |
| Number of price points | The ladder built from the lowest price, highest price and step must contain 3–16 prices |
| **Price Step (₺)** | At least ₺0.01 |
| Number of personas | The number of personas in the population (population size 10–2,001) |

The ladder starts at the lowest price and rises in increments of the step. The highest price is included only if it falls exactly on one of these steps. If fewer than 3 or more than 16 prices result, a warning beginning with **Number of price points must be between 3 and 16.** appears; change the range or the step.

## Credit calculation

The estimated credits appear in the **Credit Usage** field on the **Research Preview** page, in "balance/estimate" format. The number of competitors doesn't change the credits.

| Type and vertical | Estimated credits |
| --- | --- |
| Choice-Based Conjoint (CBC), both verticals | price points × personas |
| Dynamic Conjoint, Fashion | (price points + 2) × personas |
| Dynamic Conjoint, General Research | price points × personas |

Dynamic Conjoint's willingness-to-pay questions also cover one step below and one step above your ladder; that's where the +2 in the Fashion formula comes from. Each price you add with **Add Custom Price** also counts as a price point.

Example: in a 100-persona population, say the lowest price is ₺500, the highest price is ₺1,000 and the step is ₺100. The ladder has 6 prices (₺500, ₺600, ₺700, ₺800, ₺900 and ₺1,000). For CBC in General Research, the estimate is 6 × 100 = 600 credits.

> [!NOTE]
> In Fashion, the study also tests prices above your highest price. The ladder extends with the same step and goes past your highest price by two-thirds of the range width (highest − lowest); if a competitor is more expensive, it goes up to at least that competitor's price. In the example, the ladder extends to ₺1,400 and has 10 prices: 10 × 100 = 1,000 credits for CBC, (10 + 2) × 100 = 1,200 credits for Dynamic Conjoint. That's why, in Fashion, **Credit Usage** looks larger than **Price Point** × **Person Count**.

Credits are deducted after responses are generated successfully; a study whose response generation fails spends no credits. For CBC in General Research, the credits deducted are based on the responses the study actually produced. For Dynamic Conjoint in General Research, the preview shows price points × personas and your balance must cover that amount for the study to start; when the study completes, however, 1 credit per persona is deducted. In General Research, if the study fails at the analysis step after the responses have been saved, the deducted credits aren't refunded automatically; before restarting, send a request in the **Billing** category from the **Support** page.

In Fashion, you can't click **Start** if your balance doesn't cover the estimate. In General Research, **Start** stays clickable. If your balance is zero, the warning "Not enough credits to start the survey." appears and the study doesn't start; if your balance doesn't cover the estimate, the study fails without generating responses and no credits are deducted. For the general rules, see [Credits](/en/docs/account/credits).

## Reading the results

You can find the report later on the **Reports** page with the **Competition** filter. At the top of the report are the name of the conjoint type (**Choice Based Conjoint (CBC)** or **Dynamic Conjoint**) and the **Refresh**, **Download PDF**, **Download Excel** and **Delete** buttons. The structure shared by all report pages is covered in [Reading a report](/en/docs/results/reading-a-report).

### Summary boxes

The revenue index used in the report is the price multiplied by the number of personas who chose your product at that price.

| Box | Meaning |
| --- | --- |
| **Participants** | Number of personas included in the analysis |
| **Optimal Price** | The tested price with the highest revenue index |
| **Preference at Optimal Price** | Share of personas who chose your product at the optimal price |
| **Price Points** | Number of prices tested |
| **Competitors** | Number of competitors on the shelf |

### Demand & Revenue

- **Preference Distribution**: Shows, as lines, the share of your product, each competitor (with its name and fixed price) and **Neither** at each price. The shares at the same price add up to 100%; the optimal price is marked.
- **Revenue Curve**: Shows the revenue index at each price. This value isn't a sales forecast in lira, and it grows with the population size; use it to compare prices with each other.
- **Conversion Rate by Price**: Shows only your product's preference percentage at each price. The more steeply the curve falls, the more price-sensitive the population is.
- **Segment-Based Analysis**: Choose a dimension in **Breakdown:** and **Purchase Rate (%)**, **Demand** or **Revenue** in **Metric:**. The chart draws a separate line across prices for each segment. The **Purchase-Rate Difference** metric, available only in Dynamic Conjoint, subtracts the whole population's share at the same price from the segment's share.
- **Price Comparison Table** (Dynamic Conjoint only): Gives the **Preference %**, **Demand** and **Expected Revenue** values for each price; the optimal price's row is highlighted.

Read switching to competitors from the **Preference Distribution** chart. As your price rises, your product's line falls; the competitor whose line picks up the share you lose is the one the personas are switching to. If the **Neither** line rises, the personas you lose aren't switching to a competitor; they're not buying anything.

### Segment Analysis, Price Scenario and Crosstab

- **Segment Analysis**: Choose a dimension with **Demographic Breakdown:**. **Segment Preference Distribution** shows the share of your product, the competitors and **Neither** in each segment, averaged over all prices. In Dynamic Conjoint, the **Segment-Based Optimal Price** table also gives each segment's optimal price.
- **Price Scenario**: Choose a dimension in **Breakdown:** and a price in **Price:**. The chart shows each segment's accept and refuse rates at that price. Accept is the share who chose your product; refuse is the share who chose a competitor or **Neither**.
- **Crosstab**: Appears when there are at least two breakdown dimensions. It crosses two dimensions (e.g. age and gender) and gives each combination's optimal price in the **Segment Combinations — Optimal Price & Revenue** table.

> [!NOTE]
> If an older price study was run with the competition option, its report opens on a separate report page called **Competitive Pricing Analysis**.

## Breaking results down by segment

The breakdown lists include **Age**, **Gender**, **Education**, **Income Group**, **SES** and, if the population has segmentation, **Market Segment**. In CBC, if you added a [custom feature](/en/docs/populations/custom-features) to the population, **Custom Feature** is listed too.

Dynamic Conjoint also offers six shopping tendencies: **Price-Value Consciousness (PVC)**, **Quality-Brand Orientation (QBO)**, **Deal Proneness (DEA)**, **Novelty Seeking (NOV)**, **Shopping Hedonism (HED)** and **Spending Tendency (POP)**. For what these scores measure, see [Persona profiles](/en/docs/populations/persona-profiles). Each tendency is split into **Low (<45)**, **Typical (45–55)** and **High (>55)** groups.

- Dynamic Conjoint automatically selects the dimension that shows the largest difference between segments.
- Segments with fewer than 5 personas are hidden; if no segments are left, the message **Insufficient segment data (at least 5 participants per segment required)** appears.
- When you change the dimension, the chart updates immediately; no rerun or credits are needed.

For the difference between income group and socio-economic status (SES), and for the other dimensions, see [Segmentation](/en/docs/results/segmentation).

## Exporting

- **Download Excel**: While the file is being prepared, the button reads **Downloading Excel...**. The Excel file has six sheets: Meta, Executive Summary, Market Share Curves, Segment Impact, Population and Raw. The Market Share Curves sheet gives each option's share at each price and the number of personas counted at that price (**Valid N**). In CBC, personas whose response couldn't be obtained at a price are left out of that price's calculation; they aren't counted as **Neither**. In Dynamic Conjoint, the summary sheet also includes willingness-to-pay values such as **Median WTP** and **Mean WTP**.
- **Download PDF**: In the **PDF segment selection** dialog, tick the breakdown dimensions to include in the report and click **Generate PDF**. The first three eligible dimensions are preselected; you can change the selection with **Select all** and **Clear**.

For file names and download details, see [Exporting and downloading](/en/docs/results/exporting).

## Tips and common mistakes

- Enter competitor prices as their current shelf prices. These prices stay fixed throughout the study, and all results are read against them.
- Fill in the product and competitor descriptions; personas compare the products by their names, descriptions and prices. In CBC, personas are reminded that a high price doesn't by itself mean better quality.
- A competitor row with a name but no price is dropped without warning if there's at least one complete competitor. Before starting, check the **Competitor Count** in the preview.
- A high **Neither** share can mean the whole shelf is too expensive or irrelevant for this population.
- **Segment Preference Distribution** is an average over all prices, whereas **Preference at Optimal Price** is the share at the optimal price only. It's normal for the two not to match.
- Results are language model output; they show direction and magnitude. Validate high-stakes pricing decisions with real consumers. For details, see [Accuracy and limitations](/en/docs/getting-started/accuracy-and-limitations).

> [!TIP]
> **Optimal Price** can only be one of the prices you tested. Add in-between prices with **Add Custom Price** in the range where you expect the optimal price to be. Each extra price costs one more credit per persona.

## Related pages

- [Price Analysis (WTP)](/en/docs/research/price-analysis)
- [Choosing the right method](/en/docs/research/choosing-a-method)
- [Segmentation](/en/docs/results/segmentation)
- [Exporting and downloading](/en/docs/results/exporting)
- [Credits](/en/docs/account/credits)