This is a piece I have been meaning to write for a few months. It sets out the concept of a Revenue Equation, how that can be used to determine your KPIs / metrics, and how to derive the Controllable Input Metric (CIM) which is the most important metric to focus on. I also share how to set incentives to drive desirable behaviour re the metric. Lots of examples and illustrations make these concepts easy to understand. Let us go!

Every business can be defined as a revenue equation.

Any startup’s business offering can be distilled into a ‘revenue equation’, for example, an ecommerce business’ revenue equation at the surface level can be defined as [Items Sold * Average Selling Price or ASP * take rate].

We can drill down into each variable till you get a base variable or atomic variable that cannot be broken down further.

For the above ecommerce equation, we can say

  • [Items sold = Site visitors * Conversion]
  • in turn [Site visitors = New visitors + Returning visitors]
  • in turn [New visitors = People seeing campaign * Conversion rate of ad]
  • and [Old visitors = Past site traffic * Return rate]

And hence we have a drilled down revenue equation of

  • [(((People seeing campaign * conversion rate of ad) * conversion rate of new visitors) * ASP of new visitors * take rate) + (((Past Site Traffic * return rate) * conversion rate) * ASP of old visitors * take rate)]

We can break the above equation down even further. ASP is a function of merchandising and site display and offers; Conversion rate is a function of campaign spend for offers, and performance marketing / remarketing in general etc. At some point you will arrive at an equation that can’t be broken down further. This is the Atomic or Base revenue equation. Every variable (or operand) in the action is an atomic variable (it cannot be broken down further).

You can have revenue equations like the above for a food delivery marketplace or a home services fulfilment app (see the revenue equation for Chefkart in appendix below), and even a Direct-to-Consumer brand. Every business or startup offering can be distilled into a revenue equation. Determining the atomic revenue equation is hugely helpful, not just because it helps you better understand your business and how each variable interacts with the other, but also because the revenue equation is the first step in determining your KPIs (Key Performance Indicators).

A hierarchy of variables; and how the Revenue Equation determines your KPIs.

Each variable in the equation can be ranked based on extent of influence or degree of control, and degree of leverage. Degree of control is how much you can influence the variable and drive its change. Degree of leverage is how significant the variable is in terms of its impact on the business equation.

The most important variables are the ones with the highest degree of leverage where you also have the highest degree of control. Then the ones with the highest degree of leverage, where you albeit have limited degree of control. Finally, the ones with limited leverage but where you have a high degree of control. Rank these variables in descending order of importance. Now, for each variable, identify a metric or a key performance indicator (KPI) that directly or indirectly signifies some level of achievement of that variable.

When selecting a metric / KPI, it should ideally be what is called a Controllable Input Metric (CIM) – one which is an indicator of your activity, e.g., mails sent requesting demo meetings is a CIM or controllable input metric, while demo meetings booked (which results from mails sent) is an outcome metric. Ideally the CIM should not have an increasing cost of achievement. Performance marketing is a CIM with a rapidly growing cost of achievement – the more you spend on performance marketing the more costly it gets.

Do note that the concepts of input and output (or outcome) metrics are allied, though not synonymous, to the concept of leading and lagging metrics. Input metrics and leading metrics are more important than outcome metrics and lagging metrics. Input metrics have the quality of being actionable or changeable / influenceable, as opposed to outcome metrics which are a given. Likewise, leading metrics help us give clues to the future performance or what is portending, as opposed to lagging metrics, where the metrics are reporting what we already know.

Let us understand how Postman, the API platform distinguishes between these. From a recent private fireside chat of Abhinav Asthana (cofounder, Postman) with Anupam Rastogi of Emergent Ventures, I captured the following quote: “For most of (Product-Led Growth or PLG) companies like Postman, ARR is a lagging indicator.…actually, just being driven by ARR will lead you to take a lot of suboptimal decisions. So, the number one thing that we track is usage. And then of course, as users are coming in how many new users are coming, how kind of existing users are becoming more engaged users, and how we then convert users to paying users and those kinds of stuffs, but usage is what drives us.“ As we can see, Abhinav’s focus is not the ARR which is a lagging / output metric, but usage which is a leading metric, and one which is also influenceable / controllable.

The North Star metric is not your most important metric / KPI; the CIM is.

I contend that the most important metric for a founder is not the North Star metric but the controllable input metric(s) that influences the North Star Metric the most. North Star metric is typically an output or outcome metric that indicates progress on revenue or customer traffic or engagement. What the founder / CEO needs to focus on is the controllable input metric (CIMs) influencing the North Star metric the most, e.g., for a top-down motion-led enterprise-focused SaaS co, your North Star Metric could be ARR, check metric is churn, but the key metric may be the number of mails sent to ICPs (Ideal Customer Persona) which lead to demos that convert to sales.

Two examples to clarify the above.

  • Instead of only tracking site visitors (an outcome or a lagging metric; not a Controllable Input Metric or CIM), look at performance marketing spend which determines site visitors, and is a leading metric as well as a CIM).
  • In addition to tracking conversion on site, look closely at what is influencing it. This could be merchandising, say the presence of top 10 brands, at low prices or with special offers, which excites visitors to purchase. Arriving at CIMs requires analysis and a deeper understanding of the ultimate drivers and extraction of rules, but once done, we can arrive at these CIMs and track how they are progressing.

If we find that tracking CIM activity doesn’t lead to progress on the desired metric – for example, increasing the number of top brands and giving low prices doesn’t lead to higher conversion, then the link between these is tenuous and you need to replace them with another metric which has better explanatory power.

Let us take one more example, say the return rate of past site traffic, now this is an important but lagging metric. If we know this is influenced by nudges / notifications / retargeting etc., then these become CIMs, and we can measure the marketing team’s performance by how much they are able to drive this without increasing ad spend proportionately.

Similarly, if we determine that take rate in certain categories is determined by the number of private label brands you have then, this can be influenced directly by increasing private label presence. This becomes the Controllable Input Metric.

Basis the above, we have the following CIMs

  • Ad spend (for new visitors)
  • Retargeting (ad spend) / Notifications (for old visitors)
  • Percentage of private label (to drive up take rate)

There will be more CIMs; rank them basis degree of leverage, and degree of influence. Ideally you will have anywhere from 8 to 15 CIMs, and another 25-30 outcome or lagging metrics / KPIs (Key Performance Indicators).

Keep it CIMble!

Now, ideally you want each of these CIMs (and their associated metrics) to be owned by an individual (DRI or Directly Responsible Individual). As the founder, you should have a cadence for reviewing these CIMs. The most important CIMs merit a weekly review (especially in the 0 to 1 phase), sometimes daily. The lesser CIMs could do with a fortnightly or monthly review. In the review look at whether the CIMs (and outcome metrics / KPIs) are moving in the right direction; if they are not, ask why and unblock any hurdles in the way of the DRI or team.

An important caveat. In the 0 to 1 phase, never link incentives to outcome metrics alone. At this stage, you are still unsure what input metrics and activities will lead to the desired outcome. Linking incentives and rewards to outcome metrics may lead to your sales or revenue team not experimenting with different metrics. At the early stage when you are seeking PMF (product-market fit) you want to continuously iterate on arriving at the most scalable cost-efficient GTM, and hence linking incentives to outcome metrics leads to lower willingness on the part of the sales teams to experiment (especially given any new GTM has some teething issues).

Ashwin Damera of Eruditus holds that input metrics are relevant even at the growth stage. He says “What I like to focus on in managing people is input metrics. Don’t focus just on output metrics. For example, if the hire is a Head of Sales, I’m going to look at how many meetings that person is getting. Do they have a strategy? Are they going after one or two industry verticals, are they spreading themselves thin? Are they hiring good people under them? Are they having meetings? Are they getting proposals in? They may not get a dollar, but I can give them more resources to grow. Now, that person is very motivated because he or she is being measured on input metrics and is being asked to grow and create something that’s even larger.”

Ideally, your incentives should be set as a combination of x% + y%; x% to encourage behaviour that moves the CIMs in the right direction, and y% to the north star metric / outcome metrics. The ideal combination of x to y depends on the stage of the co, and the strength of the relationship between the CIM and the North Star Metric or outcome metrics. Incentives are fundamentally how you effectuate (such an ugly word! but it means ‘put into force or action’) strategy.

Another important caveat is that for the VC, the metrics that matter most are outcome metrics; if you are at the early stage, then the VC looks at product metrics, and specifically cohort metrics indicating retention / churn / stickiness. As an early founder your goal is to get to PMF and then concentrate on building a fast-growing sustainable business; hence at this stage, focus on input metrics that help you grow predictably with sustainable unit economics. You will invariably find that focusing on the right input metrics also yields the right outcome metrics. Do not confuse what the VC is looking for with what you need to focus on.

Summing up

Founders should build out the revenue equation to the most granular, atomic level possible. This helps you understand the key levers that matter for your business.  Now, after distilling the atomic revenue equation, for each variable in the equation, identify the controllable input metric(s) (CIMs) that influences the variable the most. Rank the CIMs in order of control and leverage. Align each CIM with an executive. Have a cadence for reviewing these CIMs. Set incentives linked to CIMs primarily, especially if you are an early-stage startup. Periodically remove CIMs that no longer correspond to the desirable outcome metrics, and replace them with relevant ones. Repeat the above. Repeat, rinse and repeat.

Appendix

Here is another distillation of a revenue equation. This time for a home services fulfilment app called Chefkart (a Blume portfolio co), which provides long-term cooks to households

At the surface level, Revenue = Take Rate * Number of Months Serviced.

  • In turn, Number of Months Serviced = Number of Customers Active in each serviceable area * Cooks Available for that serviceable area
  • Number of Customers Active for a serviceable area = (Number of Existing Customers in serviceable area (Minus) Monthly Churn of those customers) + Number of New Customers
  • Cooks Available for that serviceable area = (Number of Cooks Contracted for that area (minus) churn of those cooks) + Number of New Cooks
  • Number of New Customers = Advertising spends to onboard new customers * effectiveness of the promotion
  • Number of New Cooks = Promotion / activation spends to onboard new cooks * effectiveness of the promotion / activation

And so on till you arrive at an atomic equation. Key Metrics become

  • Cook churn
  • Customer churn
  • Ad spends to onboard new customers
  • Ad spends to onboard new cooks
  • Take rate

Churn and take rate are lagging metrics; there will be CIMs that influence them, but I am not going into specifics here. I am also not going to work this out to the atomic level. This example was shared to provide an additional illustration of how the revenue equation is constructed.

Queries / Feedback

Founders / operators: Would love your thoughts / queries on the above. You can reach me directly (email in my LinkedIn profile) or you can share your thoughts / questions in the comments below, and I will reply as soon as I can.