Startup diary: The art of conversion: how to turn your online visitors into customers
Last week I covered the 'acquisition' term of the 'fundamental equation of SaaS'. This week I'll focus on the 'conversion' term and we'll conclude with 'average revenue per user' next week.
Briefly, the idea is to measure some key metrics to understand the health of your software-as-a-service (SaaS) startup.
There are mountains of these metrics, but the best tools are ones that keep things simple. Your team can't focus on 50 different numbers.
This fundamental SaaS equation is a mental model for decision-making based on the numbers you're seeing on your website.
The equation is: new revenue equals acquisition-times-conversion-times-average revenue (per user).
The output is the amount of revenue you can expect this year in your startup.
Acquisition is how many visitors you attract, or newsletter. Conversion is the percentage of those visitors that buy - our concern this week.
Average revenue (per user) is exactly what is says and lets you tot up the amount of new cash you'll bring in this year.
If you've ever run an online advertising campaign through Google or Facebook, you'll be familiar with the term 'conversion'. Since you pay for each click or view of your ad, you care very deeply how many of people convert into paying customers.
Trust me what I tell you that you can lose a lot of money very quickly paying for online advertising.
In our case, conversion means something a little less stressful.
We just want people to sign up our service, and at first, we just want them to use the free version.
Now, that doesn't mean it costs nothing to get them there. The three acquisition tactics I discussed last week, variety, events, and virality, all cost money to execute. They're a lot cheaper than online advertising, but you still have the same basic dynamic - you need to get value for money by seeing conversions.
In a previous life, I worked in consumer ecommerce. You're lucky to see 1pc conversions there. The site I worked on was heavily dependent on search engine optimisation (SEO), meaning we had to place high up in Google results for searches for our product.
And we considered 0.5pc conversions to be a good day.
For Voxgig, we're in the business-tool space, and we're using inbound marketing and guerrilla tactics to generate visitors. The industry average in our case is around 2pc conversions.
We're going to be measuring this number and watching it very closely - it will drive changes to our marketing activities, and to our product.
A conversion for us does not mean a sale. That comes later, when users upgrade to get more features - after we've proven ourselves useful. First, you have to persuade people to invest some time in trying out your product. Time is not free and people are busy in work, so having a free option by itself is not enough.
The key tactic is make signing up as easy a possible. When you register with Voxgig for the first time, we won't ask you for a password. All we'll do is use your email address to send you a special one-time login link. This is still quite secure - it's how password resets work after all.
We will ask you to improve the security of your account if you're an active user (that is, if you keep coming back).
Just like LinkedIn and other sites, we'll have a profile completion score to encourage you to fill out your user profile fully. You'll get points for setting a password. This also provides us with a good proof point that you're a serious user.
Making the registration form as simple as possible is a tactical decision.
We hypothesise that it will lead to more sign-ups. We won't be able to demonstrate this for quite some time however.
The way that you confirm if ideas like this work is called 'A/B testing'.
The idea is that you show 50pc of your users one version of your website, call it 'A', and the other 50pc a different version, call it 'B'. Then you measure user actions for a month and use basic statistics to see which of A and B is better.
For user registration, you could have a form with and without a password.
We may well do that down the road.
The trouble with soft launching is that you do not have the user volumes to generate meaningful data.
So, for now we guess and measure what we can, and make a judgement call.
If you have been involved in a startup, a new product launch, or just been involved with building a new website for your company, you'll be familiar with the opinionated arguments the fly around the table about what will and won't work.
Opinions are a bad way to run a business.
That's where A/B testing comes in. It's a great tool for making scientific decisions. We look forward to using it.
What can you do in the meantime? Before you have data, you have graft. There's no substitute for getting out there and hand-holding your first customers as they register and use the system.
Not only will you learn a great deal about what your customers really need, but you'll also convert them.
That's what I'll be doing over the next few months - putting a lot time into registering users one by one.
This technique was famously used by the Collison brothers, Patrick and John, when they were launching stripe.com, their online payments solution.
The road to a billion-dollar company in Silicon Valley from Dromineer, Co Tipperary is paved one demo at a time.
As they demoed their system, they would do the work on registering their prospect - literally typing in their email address right there and then.
When they left the room, the prospect was almost up and running.
The lesson here is that in the early days you need to be out there meeting potential customers and going with them on their journey as they use your system.
It's a great way to learn what you're missing and what blocks people from converting.
Speaking of launches, this week we soft-launched our speaker tools.
We're registering very small sets of users to begin with, to iron out any big issues.
Each week we bring a new group on-board (tweet me @rjrodger if you want an invite).
We're keeping the numbers very low for the moment, because we want to get the most value out of the early feedback.
The great Joel Spolsky (one of the first bloggers, a project manager of the Excel team in Microsoft, a founder of Trello.com and StackOverflow.com) wrote a blog post about launching many years ago, and it still resonates today.
If if you're starting a software company, you should pretty just read all of his stuff at joelonsoftware.com.
The key insight is that if you have 1,000 users on day one and you have a big obvious bug, you'll get 999 bug reports telling you the same thing.
That means you wasted 998 bug reports that could have really improved your product.
It's tempting to just throw open the shop doors, but when you don't know what you're selling (this is fundamental truth of business software on day one), you'll just annoy a lot of people and they'll walk out with their money.
Marketing update: The speaker's newsletter has 5,452 subscribers, and an open rate of 13pc. We've confirmed the January dip with other newsletter publishers - it's real.
The EventProfs (Event Professionals) newsletter is now at 163 subscribers and an open rate of 48pc (statistical noise) and the podcast is at 63 downloads.
Richard Rodger is the founder of Voxgig. He is a former co-founder of Nearform, a technology consultancy firm based in Waterford