Consumer Decision Tree (CDT): The Order Shoppers Actually Choose

When a consumer stands in front of a supermarket shelf, they do not evaluate every product simultaneously. Instead, they follow a hierarchical decision process — a series of sequential filters that narrow their consideration set from the entire category down to a single purchase.

The purchase decision hierarchy defines the order of these filters. For example, in biscuits, a shopper might decide:
1. First: What type? (Sweet biscuit vs. savoury cracker)
2. Second: What brand? (McVitie's vs. Oreo vs. private label)
3. Third: What format? (Single pack vs. multipack vs. sharing)
4. Fourth: What price? (Standard vs. on promotion)

This order is not universal. In some categories, price comes first. In others, format or occasion comes first. The hierarchy is category-specific, channel-specific, and sometimes segment-specific.

Why does the order matter? Because the attributes higher in the hierarchy are "gates" — they determine which products even get considered. If brand is the first gate and a consumer decides "McVitie's," then every non-McVitie's product is eliminated before price, format, or flavor are even evaluated. The higher your brand's position in the hierarchy, the more protected you are from competitive switching.

Decision Level Importance Score = Variance Explained by Attribute / Total Variance in Choice x 100

This is typically estimated from conjoint analysis, discrete choice models, or Shapley value decomposition of purchase drivers.

Example (Biscuits, Grocery channel):
- Segment/Type: 32% of decision variance
- Brand: 28%
- Pack format: 18%
- Price: 14%
- Flavor: 8%

Hierarchy: Segment > Brand > Pack > Price > Flavor

Brand Vulnerability Index = 100 - Brand Importance Score
If brand explains only 15% of the decision (e.g., in commoditized categories like bottled water), the brand is highly vulnerable to substitution based on price or format.

Switching Probability at Level n:
SP(n) = Consumers who change brands when attribute at level n changes / Total consumers
Higher-level attributes have lower switching probability (more "sticky").

Research in the frozen pizza category revealed two distinct decision hierarchies for two consumer segments:

Brand-loyal segment (45% of shoppers):
1. Brand (which manufacturer?)
2. Type (thin crust vs. deep dish vs. stuffed)
3. Size (single vs. twin vs. family)
4. Price (regular vs. on promotion)
5. Topping (pepperoni vs. cheese vs. margherita)

Price-driven segment (35% of shoppers):
1. Price (what's on promotion? what's under $5?)
2. Size (how much food for the money?)
3. Type (thin crust vs. deep dish)
4. Brand (reputable enough?)
5. Topping (preferred flavor)

For the brand-loyal segment, the manufacturer's priority should be brand awareness and visibility (be seen first). For the price-driven segment, the priority should be competitive pricing and prominent promotional display (be the best deal visible).

The remaining 20% were "occasion-driven" — their first decision was type/format based on what they were planning (quick weeknight dinner vs. party). This segment responded most to pack-format innovation.

Cross-lesson connection — PPA Lesson 2 (Pack Roles): the Size/Format level of the CDT maps directly to the Entry → Routine → Upsize → Upscale path from Lesson 2. A shopper who gates on "format first" is literally navigating the Pack Roles ladder through the CDT before brand or price enters. Cross-lesson to Pricing Lesson 1 (tier-specific elasticity): brand-loyal-segment shoppers (price gate 4) are relatively inelastic on their preferred brand — typically -1.0 to -1.5 own-price elasticity within the brand. Price-driven-segment shoppers (price gate 1) are much more elastic — typically -2.5 to -3.5 — and shop across the full price band. Occasion-driven shoppers sit between these extremes. The aggregate category elasticity is a weighted average of these path-specific elasticities; aggregating without CDT segmentation produces misleading forecasts.

Determining the actual decision hierarchy requires research, not assumption. Common methods:

1. Decision tree analysis from panel data: Analyze actual purchase sequences to infer the decision order. If consumers consistently buy the same segment but switch brands, segment is above brand in the hierarchy.

2. Conjoint analysis: Ask consumers to choose between product profiles varying on key attributes. The attribute with the highest part-worth utility is highest in the hierarchy.

3. Eye-tracking studies: In-store or virtual shelf eye-tracking reveals what shoppers look at first. The first fixation point often corresponds to the top of the decision hierarchy.

4. Shopper intercept surveys: Ask shoppers immediately after purchase: "When you approached the shelf, what was the first thing you decided?" Simple but effective.

5. Basket analysis: What substitutes for what when a product is out of stock? If consumers switch to the same brand in a different format, brand is higher than format. If they switch to the same format in a different brand, format is higher.

The most common mistake: assuming the hierarchy is the same across all consumer segments. Price-driven shoppers have price at the top. Brand-loyal shoppers have brand at the top. Your CDT should account for at least 2-3 segments.