How DataMilk improves gross profit for its customers
Traffic division between Original and Smart traffic
DataMilk splits the traffic to each customer's website randomly into two parts: Original traffic and Smart traffic (or Optimized traffic). The role of the Original set is to always maintain a part of the traffic where DataMilk does not modify the user interface. This acts as the baseline for measuring DataMilk's performance on a customer website. DataMilk uses this baseline to report how much Smart visits improve revenue.
The sources of gross profit uplift
The additional gross profit that DataMilk generates arises from a multiplication of the following factors:
- total traffic to the site
- the ratio of Smart traffic to the total traffic
- revenue uplift per 1000 visits
- gross profit margin
Let us consider an example store with 1,000,000 visits and $2,000,000 revenue per month. If the Smart traffic ratio is 50%, then DataMilk optimizes 500,000 visits per month, which would generate $1,000,000 revenue without DataMilk's Smart Components.
If DataMilk achieves 6% revenue uplift, then this means that the Smart traffic has generated $1,060,000 revenue instead of $1m. If we assume that the gross profit margin is a fixed 30%, then this $60,000 extra revenue is $60,000 × 0.3 = $18,000 extra gross profit per month.
Let us look at these factors in detail.
Total traffic
DataMilk does not increase traffic volume to the store. By ensuring high uptime and very low latency, DataMilk maintains the same level of website loading performance that shoppers already experience.
Traffic ratio setting
DataMilk launches stores with a 50% Smart : 50% Original traffic ratio. The system gathers data and demonstrates the performance that is gained by the Smart visits. As time progresses, DataMilk recommends turning up the ratio of Smart traffic in order to increase the revenue uplift for the whole store instead of just the original 50% of total traffic. Both DataMilk and the store benefit from increasing the traffic ratio as DataMilk earns more revenue, and the store earns more revenue and profit for the additional Smart visits. Statistical computations determine how high the Smart traffic ratio can be such that the Original traffic volume still provides enough statistical certainty (e.g. 90% Smart traffic and 10% Original traffic). DataMilk and the store choose the traffic ratio together to get the best results.
Revenue uplift per 1000 visits
Given a certain traffic volume in Smart traffic, how much revenue is booked for it?
Note that conversion and order are used interchangeably. The conversion rate is the number of conversions divided by the number of visits. As metrics of averages, the revenue per conversion is the same as the average order value (AOV). Therefore
Revenue per 1000 visits =
= Number of conversions per 1000 visits × Revenue per conversion
= Conversion rate (per visit) × 1000 visits × AOV.
In our example, revenue per 1000 visits is $2000. The conversion rate can be 2%, giving an AOV of
Revenue per 1000 visits / (1000 visits) / Conversion rate =
= ($2000/1000 visits) / 0.02
= $2 / 0.02
= $100.
DataMilk impacts revenue per 1000 visits in two ways: the Smart Components increase conversion rate and they change AOV. AOV change can be positive or negative.
Let us assume that in our example the conversion rate uplift is 4%, the AOV uplift is 3%. How does the revenue change?
Optimized revenue per 1000 visits =
= (Original conversion rate × 1.04) × 1000 visits × (Original AOV × 1.03)
= (Original conversion rate × 1000 visits × Original AOV) × (1.04 × 1.03)
= Original revenue per 1000 visits × (1.04 × 1.03)
= Original revenue per 1000 visits × 1.0712.
We see that a conversion rate uplift of 4% and an AOV uplift of 3% combine to give a revenue uplift of 7.12%.
Gross profit margin
The most data-driven stores transmit to DataMilk not only the value of every individual transaction in the store, but also the gross profit booked for each transaction. This allows DataMilk to work toward generating directly the most gross profit for these stores.
The alternative solution is to specify a single number for the entire store, an approximate, fixed gross profit margin. This margin varies greatly between different commercial sectors. In this setup, revenue and the gross profit DataMilk can report are proportional, e.g. gross profit = 30% of revenue. By generating x% revenue uplift, DataMilk reports that the system generated x% gross profit uplift. DataMilk’s goal is to maximize revenue uplift.