• Follow
• Contact

Sunday 20 October 2019

# Don't trust average numbers in your startup - they're a bad metric

## Startup diary

Richard Rodger

Can you predict the success of a company based on a single mathematical equation?

Considering how complex the factors are that determine success, you'd have to think it would be a pretty complex equation.

Isaac Asimov, the great science fiction writer, took this one step further. In his 'Foundation Series' of novels, he suggested that you could predict the thousands of years of history of a galactic civilisation with a mathematical equation, if it was sufficiently complex.

Your intuition will tell you this probably isn't possible and you'd be right. Scientifically, it's the same reason we can't predict the weather - it's what's known as a chaotic system. If you just slightly change things at the start (the proverbial butterfly flapping it's wings somewhere in South America), then you end up with wildly different outcomes a few days later.

You can even have the perfect equation, correct in all details and incredibly long, and it still won't help - your measurements will never be perfect and even a slightly misreading makes the prediction useless within a short time. And yet, if you accept that perfect prediction is impossible, you can run your company using a simple mathematical equation.

At least, you can run your Software-as-a-Service (SaaS) company on this basis. The trick is to find the relationships between a very small set of factors that can give you useful information that you can act on.

Think of a doctor taking your blood pressure and heart rate. These measurements won't predict your health exactly, but they're still very useful to the process of diagnosis.

Here, then, is the equation of SaaS: New Revenue equals 'Acquisition' by 'Conversion' by 'Average Revenue' (per user).

Just before we dig into this equation, let's think about the wider context of the way a SaaS business works.

Once it is up and running, a SaaS business starts every year with an existing set of subscribers, paying monthly subscriptions (to a first approximation - we'll ignore any special pricing offers).

That means you know on January 1 how much money you will collect if you do nothing all year.

That's why SaaS is such a great business model - once established it's a cash machine!

There is a fly in the ointment: churn. You're going to lose some customers due to natural attrition. Companies close down, or get bought, or just don't need you anymore.

You can expect to lose some customers every year no matter how fantastic your product.

So, in reality, if you do nothing all year, your revenue declines 5pc-10pc, on average. You need to counter this with new business, and you need to grow - that's where the equation comes in.

Let's look at it again. New revenue in the coming year will be 'Acquisition' by 'Conversion' by 'Average Revenue' (per user).

In a SaaS business you live or die by this equation. It's not a perfect model of your business, just a map for where you should be going.

Here's the breakdown: 'Acquisition' means the number of people you can get in front of.

Visitors to your website is a big component of this traditionally, but that's just one way to build an audience.

In our case we've focused on creating newsletters and other high-engagement content. As I've discussed at length in this column, this is a great way to build a loyal audience.

An audience is no good if you can't turn them into customers. That's what 'Conversion' means.

Here's a tip: there's not much point building your audience if you can't convert them.

That's why we haven't gone for a strategy of writing lots of generic content on our blog - pre-launch our conversion rate is zero.

The conversion rate of newsletter subscribers to users will be a very big number for us, and you'll get to go along that journey as a I write about it.

There's no chance to sugar coat the results as you see the consequences of our decisions in real time.

Finally, your have 'Average Revenue' (per user). This turns the result of the equation into money.

Of every 1,000 audience members, you might convert 2pc (that's the relatively normal conversion rate you see for web content), giving you 20 new customers.

If your average revenue for each customer annually is €100, then you'll get €2,000 in new revenue this year.

Averages are terrible things and very misleading. Consider the common belief that most people in the Stone Age were dead be the age of 20 and most people in the Middle Ages were dead by the age of 30.

You'll often see these numbers expressed as average life expectancy. But almost everyone who made it out of early childhood in those times actually got much more of an innings - at least 60 years. Early childhood mortality was so high that it dragged the average down.

Averages are terrible things, mathematically. Don't trust them.

But they are useful for aggregating data.

Even though your SaaS pricing might have multiple tiers and discounts, so that your top enterprise customers are paying €500 a year, and your individual accounts are paying just €50, you can still use the average to predict overall revenue.

Just remember that the average does not represent any actual customers.

To make you SaaS business a success, you need to improve each term of this equation: get more eyeballs ('Acquistion'), more users ('Conversion') and more revenue ('Average Revenue').

If you can focus on improving each term, you might just get to an IPO.

Over the next three weeks, I'll cover each of these terms in more detail and talk about how Voxgig will drive each one for our business, scientifically.

Marketing update: The speaker's newsletter is still at 5,335 subscribers and has an open rate of 13pc.

Our hypothesis is that there are two effects here: lower numbers of new subscribers over the holiday season and people clearing out their newsletter subscriptions for the new year. I'll let you know once we figure it out.

Our relaunched 'Event Professionals' newsletter aimed at conference organisers is now at 59 subscribers and an open rate of 54pc (that will fall quickly as subscriber numbers climb).

The podcast is now at 58 downloads per week - we've started adding a section for this to the 'Speakers' newsletter, so we'll soon see how that helps.

Richard Rodger is the founder of Voxgig. He is a former co-founder of Nearform, a technology consultancy firm based in Waterford.