Discounting Budget

Mathematically there is no difference between the discounting budget and the price reduction. However, discounting instruments allow one to price the product differently for different user segments.

The most important KPI for discounting-related spending is Incremental ROI (or Incremental revenue per incremental unit cost)

Types of Discounts that are commonly used

1. New User Discount – Typically given to a user who is new to the platform.
2. Specific Item-related Discount – Typically an additional price reduction, is given on select merchandise that is either not selling or offered for the first time
3. Bundled Discount – Typically given to a user when their cart size crosses a meaningful milestone – say crossed 1000 bucks or something.
4. Limited Offer kind of Discount – A variant of 2, typically done for boosting the demand in the lull periods, for example – weekday discount in the restaurants business is quite common, another variant of this is Mega Sales, where the methodology is same, but the concept is to maximize the revenue during SALE times.
5. Loyalty Discount – Typically given to a user who is frequently buying your products or services. The idea is that since this is a high LTV user or the user has paid an additional amount to subscribe to the loyalty program, you can pass on some price leverage to them. P.S. – This is usually executed via platform currency points e.g. frequent flyer miles.
6. Banks / Financial Instrument-based Discount – Given to users who transact using a particular bank credit card, or a certain payment service. This is usually partially or fully funded by the financial partner as it helps them in getting a certain transaction, while the end user benefits with the reduced price. At the same time, your business gets a good deal, without any cost.

How to measure the Performance of the Discounting Budget

In order to set up the performance measurement, you need to identify the atomic unit of the discounting budget.

The atomic unit of the Discounting Budget is a unique combination of the following

1. Objective – which is typically revenue, could also be conversion
2. Target User Segment, who will be exposed to this
3. Applicable Supply, where this discount will work
4. Discount Amount or %age
5. Base Price

Atomic Unit = Combination of (Objective, Segment, Supply, Discount, Base Price)

Now, Total Discount Budget = Sum of all the discounts in operation

How to measure the performance of a defined atomic unit, say ‘a‘

• firstly in absolute terms
• and secondly in comparison to other forms of discount

To measure the performance in absolute terms, typically you would want to determine, that for a fixed segment and supply, what is the relation between discount %age and objective achievement? In order to test this out, generally, we follow the below technique

1. Setup 2 Sample User Sets – Test and Control
2. Expose the Discount to Test and keep Control as is.
3. Measure the following across the 2 sets
   1. Revenue (Assuming this is also the objective)
   2. Conversion
   3. Cost
   4. Significance

Incremental ROI = (Revenue _Test – Revenue _Control ) / (Cost _Test – Cost _Control )

Ideally, every discount should have enough run time and iterations to find their own maxima. Next, a comparison is made across the optima of every discounting budget to identify a relative winner across multiple possible discount offers.

One callout – New user discounts need to be seen in the context of future cash flows that the Customer is likely to bring in. Hence, even if new user discounts provide lower incremental ROI compared to repeat user discounts, one needs to consider the user LTV in the overall calculations.

Reference for Confidence Calculation and Theory if you want to know about calculating significance calculation in detail.

How often should you measure the performance of Discounting Budgets?

Short Answer: Depends on your appetite and the scale of business.

The very first discount you put on your platform mostly works. But as the discounts become more and more available, users dont see any exclusivity and tend to behave as if there is no discount in operation.

Typically ROI is established on certain discount types and teams usually operate with the same heuristics for a long time. Doing test and control measurements continuously may not be very effort effective. However, teams should still revisit the benchmarks on a 6-month to 1-year basis.

Another technique utilized is to keep a certain segment of users in control all the time, and put all the ensuing experiments on all but this segment (In one of my experiences the teams used to refer to one such segment as Pure Organic – for this segment behaves closest to the organic nature of the user). Using this technique you will be able to see the performance of the discounting techniques over time, and also be able to measure the performance of one discounting technique compared to the other.

Questions for the Reader

1. What should be the ratio of the discounting budget vs the acquisition budget for a company
2. What is the true CAC, if the total expenditure on the first transacting user also includes the cost of getting the user to transact on the platform (i.e. new user discount)?

---

If you haven’t read my article about the Acquisition Budget – You can find it here.

The 4 Types of Marketing Budgets and how to measure if they are working for you (1/4) Acquisition Budget

Next, let’s look at the third category of the Marketing Budget – Reactivation+Engagement Budget. Coming Soon…

---