Subscription analytics basics: Data-driven workflow

In the previous articles we have discussed how to measure three core concepts in any subscription business: volume, revenue and costs. In this article, we will go through three additional areas that are important for working in a data driven way.

Segment the analysis

In the previous articles we discussed how to measure volume, revenue and costs – at a high level. By high level we mean, for example, the entire subscription base. However, the analysis and insights will not be interesting until we start segmenting. Sure, it may be important to understand for example, the churn on a total level, but it is even more relevant to look at churn per product or churn per sales channel. It is only then that conclusions can be drawn that lead to concrete actions.

We recommend starting to segment your analysis according to the following dimensions:

  • Product: The product referred to in the subscription
  • Offer: What offer the subscriber has, or got in on
  • Sales channel: Which sales channel or source it was recruited through
  • Customer data: All other customer data we have, such as demographics

There are obviously more dimensions and more data that can be applied to enrich the analysis. After you’ve mastered the above dimensions, we would recommend taking a closer look at engagement and usage data – i.e. metrics that relate to activity, behavior, engagement and so on.

Build transparency

Data can become the common language of the organization and ensure that everyone strives for the same goal. To achieve this, one must be transparent with the data and make it available internally. In this way, a mindset is created where the data points are continuously discussed and solutions to problems are developed.

In order to achieve this in practice, one must ensure that the organization consumes the data on an ongoing basis and that they understand it. The first point can be solved by, for example, putting up dashboard displays in the office or through well-designed dashboards online. The second point, to create understanding, may require more work. Here you can, for example, hold courses for different departments, where you go through the concepts and talk about what is important to influence the numbers in the right direction.

Also, keep in mind that transparency can be sensitive. Suddenly, it becomes difficult for employees and managers to hide their own or the department’s failures. This may mean short-term loss of prestige, but it is also the whole point of transparency. Especially when things are looking down, the entire organization must gather and understand how things can be improved.

Prioritize actions based on data

When the organization understands how data drives the business, one can also begin to prioritize actions and activities based on historical data and/or hypotheses about the future. Here you should take a holistic view of the activities the organization is doing and ask yourself: Are we doing the right things? Are we doing enough to reach our goals?

With relatively simple means, most activities (both existing and new) can be quantified in a “common currency”. When quantifying the measures in a common currency, one can prioritize them and choose the ones that provide the most impact. Here it can also be important to think about what effort is required. Such a common currency is usually expressed in one or more of the core concepts we have previously reviewed – such as volume, revenue and costs.

Here is a practical example. Let’s say we want to prioritize and choose between three activities:

  1. Try a new sales channel
  2. Up sell to a more expensive product
  3. Reduce costs in the organization

Starting from the back with the costs, we make some assumptions and calculations, and establish that this activity can reduce costs by 2 million per year. Costs directly affect the profits, so the change in contribution will be +2 million per year.

Next comes the up selling to the more expensive product. Let’s assume that we manage to up sell 1000 subscribers at the same time, and that the new product costs 100 more per month (compared to the one they had before). The increased revenue will thus be +1.2 million (1000 * 100 * 12) in 12 months, which is also a positive contribution impact with the same amount. Costs are not expected to be affected by this activity.

Lastly, we want to quantify the new sales channel. Let’s say we assume that we can sell 300 new subscriptions per month, at a sales cost (CAC) of 100 each and an average revenue of 160 per month. This case is somewhat more complicated where one may need to make a quick forecast to understand the development of a period of time, and also make assumptions about churn. We set the churn to 5% per month and roughly estimate that in the first 12 months we will get 1.2 million in revenue with sales costs of about 0.2 million. I.e. the change in contribution will be around +1.0 million.

So, which activity would you choose? Do you want 1 million, 1.2 or 2? Here, consideration should also be given to what efforts are needed to make the actions happen, as well as how confident one is of the estimates. It can be a good idea to work with team when you quantify, to understand the highest and lowest numbers you think are plausible, as well as getting their input early on.

To find these new activities, one has to think creatively and dare to try. Brainstorm in cross-functional groups and be sure to look at what other players in the market are doing, which might already have been tested and proven. When you’re done with the quantification and prioritization, it is important to set the right process for following up the activities and learning from the results. This is not about holding individuals accountable for achieving the result, but rather how to learn and become better together as an organization.

Our final advice is: keep it simple. Do not work with “big data”, work with “relevant data”. Especially when it comes to making sure that the entire organisation understands and works with the data, then simplicity is critical.

Recuro

Recuro

Do you need help related to this topic? Feel free to reach out on hello@recuro.io for a discussion.