How to automate revenue monitoring

Effective revenue monitoring means monitoring all the metrics and KPIs that impact your revenue. Here’s how automation will help you stay on top, whatever revenue models you use

Share This Post

Share on facebook
Share on linkedin
Share on twitter
Share on email


Revenue is the ‘mother of all metrics’ for a very significant reason: Because all the data, metrics and KPIs you monitor ultimately stand to impact your revenue. If your site is experiencing downtime, this can hurt your revenue. If a specific item is under-performing, this can hurt your revenue. If your campaigns aren’t working, this can hurt your revenue.

However, monitoring metrics that could ultimately make significant changes to revenue is a huge, resource-heavy challenge for every business today. There’s just so many metrics that go into revenue, and keeping on top of all the fluctuations within those metrics is a huge task.

Fortunately, it’s a problem that can be easily solved with AI automation. But, before we dive into the exciting opportunities that automated KPI monitoring open up, let’s first take a look at the four of the most common revenue models and the unique challenges they present for revenue monitoring.

Single payment revenue models

Exactly what it says on the tin. A single payment is a one-time only payment for a product, or a number of products, that’s agreed upon in advance. For example, an electronics company might sell a TV for a single, one-time payment and collect $1200.


The monitoring challenges

Single payments are product-based and are very sensitive to fluctuations in metrics such as:

  • Product updates
  • Technical glitches
  • Product inventory changes
  • Purchase flow changes

To demonstrate the impact an under-performing metric can have on revenue, let’s look at the example of a cosmetics company. This company might make $200 million in a year and have an average order value (AOV) of $30. This means that they are processing hundreds, even thousands, of sales every single day. With such a huge volume of sales, it becomes incredible difficult to monitor potentially under-performing metrics. So, if a certain product isn’t selling as well as others, as a result of the sheer amount of data being processed this is likely to fly under the radar. Yet the company could be missing out on thousands in potential revenue by not investigating why this product is under-performing.

Subscription revenue models

Subscription plans are the most popular revenue model today, in part because they offer a more regular stream of revenue than a single-time payment. A subscription model is a regular, recurring payment that allows a user access for a period of time. The money usually comes out of an account on a monthly or yearly basis (often with a discount for those paying yearly). This is a common form of revenue model for SaaS tools such as Millimetric’s, gyms, streaming services like Netflix and publications, to name just a few examples.

Here’s an example of the Financial Times‘ subscription plan, allowing access to an unlimited number of articles for paying customers.


The monitoring challenges:

While subscription models don’t fluctuate as much as a single-time payment, they still come with their own complications. A subscription revenue model relies on monitoring metrics like:

  • The activity of subscribers (sometimes in their millions)
  • Number of renewals
  • Number of cancellations
  • Up-sells or cross-sells

For example, last year Spotify reached 124 million paying subscribers (adding 3 million every month in Q4 of 2019). In this case, tracking customer behaviour is incredibly difficult with so many users across so many different locations on so many potential subscription options. Overlooking a misbehaving metric can easily be done, but it can create havoc and damage revenue.

Micro transaction revenue models

Micro transactions, or MSX, are most common in the gaming world. Micro transactions are a digital-only revenue model where users can make a one-time, online purchase of virtual goods in-game.

Here’s an example of some micro transaction payment options in the freemium game Candy Crush.

The monitoring challenges

Much like single payments, micro-transactions are also product-based and therefore require monitoring KPIs such as:

  • Technical glitches
  • Product updates
  • Purchase flow changes

Affiliates, partners or ads revenue models

This fourth revenue model tends to be the most complex. Companies using it focus on connecting buyers and service providers rather than creating a product or service of their own. This revenue model finds an audience interested in a product or service and directs traffic towards it. Common examples are travel agencies like Tripadvisor, or e-commerce marketplaces like eBay. Money is made through this model by affiliations, partnering, ads or a mix of these three.


The monitoring challenges

Affiliate, partner or ads revenue models are vulnerable to similar challenges as marketing campaigns, like:

  • Traffic
  • Conversions
  • Referrals
  • Purchase volume
  • Technical glitches and downtime
  • Impressions

For example, if eBay experienced a 50% drop in traffic to a particular product, this would be hard to detect among the millions of products available on the website, but will probably halve the number of conversions for that particular supplier. Because it’s difficult to monitor so much data for so many partners, organisations using this revenue model tend to focus on monitoring metrics of their top partners – and experience a high level of churn from smaller partners as a result, losing thousands in potential revenue each year.


RELATED: How to choose marketing KPIs


Blended revenue models

Many companies today also use a blended approach where they utilise a mix of these four revenue models in order to meet the needs of the largest amount of customers. Amazon, for example, mostly utilises the affiliate marketing model to market other companies’ products. But they also sell many of their own products for single payments, as well as operating a subscription model for their delivery and Prime video services.

Blending revenue models might be an effective way of supercharging revenue, but it creates even more opportunities for unruly metrics to damage your bottom line.

Monitoring revenue: The requirements

Revenue monitoring is a complex ecosystem, where anomalies and fluctuating metrics can at any point devastate a company’s profit margins. But, with so many metrics that can potentially impact revenue to monitor (and companies using a mix of of revenue models to boot), it can be incredibly difficult to stay on top of all the KPIs affecting revenue.

There are three crucial requirements to ensuring you are able to fully monitor the metrics impacting revenue:

  • Time: To fully understand your revenue and all the metrics that impact your bottom line, significant time needs to be put in, week in week out, to comb through the relevant KPIs
  • Resources: To stay on top of all your revenue metrics, resources need to be put towards acquiring and cleaning data, budgeting for the tools used to understand it and paying the salaries of analysts and data scientists
  • Talent: Depending on the size of your company, you may need a whole team of analysts to help you understand behaviours and performance opportunities that impact revenue. This isn’t just a costly investment, it’s also difficult to source data talent due to high demand

Time, resources and particularly data and analytics talent are in short supply for many organisations. And, with millions of pieces of data to comb through, human error is always a risk to your revenue when you approach revenue monitoring manually.



Luckily, today we have automated KPI monitoring technologies to streamline the process of monitoring revenue and ensure your profit is protected from anomalies and fluctuating metrics at all times.


RELATED: What to do when there’s ‘too much’ data


Automated revenue monitoring

The millions upon millions of pieces of data that make up the metrics that impact revenue might be difficult for humans to keep track of, but analysing large datasets is exactly where AI flourishes.

Automated KPI analysis tools like Millimetric use AI to analyse metrics across multiple dashboards 24/7, alerting users to unusual activity before it damages revenue. Using our AI model, we’ve helped clients across a range of industries to:

  • Identify technical issues causing users to abandon webpages in real time
  • Unearth anomalies in their campaigns that are decreasing conversions
  • Optimise their SEO presence to attract more customers
  • Discover product glitches as they happen
  • See the products that are struggling, and drill down into why

In addition to finding many more digital performance opportunities that helped our users supercharge revenue.

Try our SaaS tool for free today to enjoy benefits including:

  • 24/7 analysis revealing any anomalies, trends and digital performance opportunities as they arise
  • More free time to focus on driving performance and the creative endeavours humans excel at
  • More reliable results and data-driven decisions. AI doesn’t make the mistakes that humans do
  • ‘Analytics-on-demand’. Our AI alerts users to fluctuating metrics and anomalies so they don’t have to search for them.
  • Anomaly scoring. A clear number system that lets users know how much each anomaly stands to impact revenue
  • Increased efficiency of data analysis across ALL dashboards

Or contact us at to learn more about our bespoke Enterprise model that offers full customisation to suit any workflow, agency-style consulting from our team of specialists and fully-supported set up to make the integration process easy – saving you time and money to drive better performance.


More To Explore

Ready to get started?

  • Try it Free
  • No credit card required
  • No code required