A Simple Way To Think About Demand In Detail

Demand can feel both nebulous and very concrete at the same time. On one hand, we think about it as an abstract characteristic of the market, something difficult to define other than as a downward-sloping line from Econ 101. On the other hand, it feels very immediate - it's the customers who walk into your store, who buy your product, and who might buy your competitor's product instead.

How can we reconcile these views in a way that's useful for making decisions about price and demand management? That's the question we'll answer in this article.

Starting with the Econ 101 view, we'll work our way down to a view that feels more immediate and familiar. Then we'll explore why it's useful to think about demand in detail and how you can use that detail to optimize important decisions about pricing and demand management.

Breaking Down Demand By Example

Imagine you recently opened a pet supply store in a town whose residents feel very differently about dogs. Half the town loves them, but the other half is deathly allergic. Since the closest pet supply store is an hour away, you decide to start simple and sell only a single brand of dog food.

High Level Demand: Single Product, Single Market

Three months after opening, you sell $20,000 worth of dog food. Great!

You price the dog food at $50, and you do the math... 400 bags total, 133.3 bags a month - that translates to 4.44 bags a day. On average, the town has 2 dogs per household, so you have 2.22 customers buying 2 bags of dog food each day. That's about one paying customer every 3.6 hours. Not bad!

Your view of demand at this stage is at a pretty high level. Since you're the only pet supply store in town, the demand you see is essentially market demand.

Raise your prices next quarter, and you'll sell less than 400 bags because some customers will drive an hour to your competitor. Lower your prices, and you'll sell more than 400 bags. This is the classic downward-sloping demand curve you learn about in Econ 101.

High Level Demand: Multiple Products, Multiple Markets

After a successful first quarter, you decide to expand your product line. In addition to dog food, you decide to help out the allergic residents by selling a branded version of allergy medicine.

You sell $4,000 of allergy medicine in the second quarter at$ 20 per bottle. You do the math... 200 bottles total, 66.7 bottles a month - that translates to 2.22 bottles a day. Since a bottle lasts for three months, that's one paying customer every 3.6 hours. Same as the dog food shoppers!

Since you now have two types of paying customers each arriving every 3.6 hours, you will have some kind of paying customer every 1.8 hours. Since half the market has dogs and half the market has allergies, which kind of customer you will see comes down to a coin flip.

Again, your view of demand is still at a pretty high level. We basically have two markets and a market demand curve for each of them.

Detailed Demand: Multiple Products, Multiple Markets, Multiple Choices

In the third quarter, you decide to expand your product line even further by selling dog treats and a generic version of the allergy medicine. You also notice that during the 1.8 hours between paying customers, there are a few customers who browse but don't buy.

How do you think about pricing now that you're selling dog food and dog treats together as well as branded and generic medicine? Should you consider offering a discount when customers buy both dog food and dog treats? How does the price affect the number of customers who browse but don't buy?

The answer is to think in terms of probabilities. Since dog food, dog treats, and allergy medicine are all necessities, let's assume that if someone browses but doesn't buy then that means they buy from the competitor an hour away. You know that if someone walks into your store, then there's a 50% chance they're a dog lover and a 50% chance they're allergic.

If they're a dog lover, then they'll do one of four things:

Buy dog food
Buy dog treats
Buy dog food and dog treats
Buy from the competitor

If they're allergic, then they'll do one of three things:

Buy the branded allergy medicine
Buy the generic allergy medicine
Buy from the competitor

Clearly, price affects a lot of things here. Set the price too high and customers will go to your competitor. The price of dog food and dog treats will affect whether customers buy food and treats separately or together. How you price the allergy medicine will affect whether customers substitute generic for branded.

Setting optimal prices in this situation comes down to understanding how the probabilities of customer choices vary with price. You can estimate these effects by talking with customers, analyzing historical data, and creating a model. In this case, you can model the choices of dog lovers using a multivariate probit model and the choices of allergic customers with a multinomial logit model. A little complicated, sure, but it's doable.

What's interesting is that at this point we've gone from a very high-level view of demand to a very granular one.

Thinking in terms of customer segments that may arrive at your store at different times and rates, and then who make different choices based on a probability model introduces a lot of useful dynamics. In this case, we saw competitive dynamics, substitution effects, and effects of complementary products. Because we modeled with probability, we also introduced the dynamic where customer behavior is somewhat unpredictable.

Why It's Useful To Think About Demand In Detail

It's useful to think about demand in detail because we can optimize over it.

If different segments arrive at different times and rates, then we may adjust, price, quantity, or terms of trade to account for that. Once in store, we might organize the placement and groupings of our products to encourage customers to make more predictable decisions. It may even make sense to reward customers with discounts and promotions through a loyalty program in exchange for more getting more certainty about their purchase decisions.

Of course, this level of detail can be taken to an invasive extreme, but the framework is a useful one to better understand the composition of demand and be intentional about how to offer and price your products. Instead of thinking about demand as a single curve from Econ 101, you can break it down into more concrete chunks that you can work with to optimize demand management decisions.

Conclusion

Demand can feel both nebulous and very concrete at the same time. With the right framework, however, you can break down demand into more concrete chunks that you can use to optimize important demand management decisions. In the process, you can not only better understand the composition of demand, but also be intentional about how to offer and price your products.