My team at TownSqua.re just went through an intensive planning week down in Austin(remind me to post the video). We spent time going over customer feedback, reviewing and refining strategy and analysing our metrics. Then we got to work planning what the product road map would look like this quarter. We just hired two new bad ass developers and a killer marketing beast(Gorgous George).

When it come time to plan out the features it is tempting to build out features that will 'help us grow' but is that what our startup needs? We launched last quarter and found that co-working spaces, student groups and other communities found value in our tool but... we did little in the way of validating our primary business model. Even worse we had not designed and tested a scalable customer aqusition process.

Growth Hacking a cross section of test users:

This is where the chicken and the egg scenerio starts. We wanted to know, not just feel, like we had a solid value to customers. In order to do that we need to throw some customers in there. We got a hustle machine sales guy to pound the pavment and start getting the leaders of these communities on to our system. Now this is not a scalable customer aqusition process, but we needed the data. Building out the scalable customer aqusiton processs with out knowing we provide a value is futile.

Validating the product:

Once we had the users on we got feedback, oh boy did we get feedback. We saw what they liked, what they hated. We saw some behaviors we did not expect, but better yet we saw organically with out building out a specific feature set some behaviors that strongly indicate that our end monitization strategy is valid.

With those indicators in mind this quarter I made it a huge priority to test our monitization strategy. We proved we have value to the customer and have at least one revinue stream but it pales in compairison to the billion(yes with a 'B') dollar play.

This is where I ran into a little bit of an up hill battle. Some of our team wanted to jump right to focusing on scaling. "Lets prefect our customer aqusition process", "Lets add more features so we can capture a larger cross section of the market" both to which my response was "Hell No", not yet.

Some of my team mates looked at me like I was crazy, I even had to put a phone call in to one of our team members not in Austin to make sure I wasnt insane. After listening to Do More Faster by the founders of Tech Stars I was far more confident in my resolve to test what really matters: Can we build a valuable product that we can then monitize and scale?(In that order)

So I devised a set of expirements that would prove our business can be monitized and that we have a competitive edge. I ranked the expirences as follows using a baseball analogy:

• The Bunt - The lowest risk and lowest reward if we can prove it
• The Single - The medium risk play, others have done it but it has a higher reward if we can add our spin to it
• The Home Run - A high risk cause no one has done it quite this way that gives us a decent competitive edge if we pull it off
• The Grand Slam - A crazy high risk because it looks like no one has thought of it yet but if we can pull it off it has exponential growth.

I also came up with a specific set up metrix that would show just exactly how exponential our lifetime customer value is.

I have purposly been being vague for two reasons:

1. This model should be appied to any startup
2. Whats the fun if I just laid out my grand scheme for you right now, figure it out yourself. Its more fun that way

I will give you this. If I am right our grand monitization play is a two sided marketplace. This means you need to have buyers and sellers. I also aim to find the exact right amount of buyers to sellers. That ratio is a key metric. Then ideally I will use that to design our customer aqusition strategy that grows both halves of the market place in close to optimal ratio.

Validating the customer aqusition process:

So next we need to know that we can take the market one customer, then the next. Like I said earlier we have a bull dog hunting down potential customers and manually walking them through our system. This is great because he is learning a lot about our customers but this will not scale.

We could do what large enterprise software companies do and hire an army of sale people but that type of overhead would crush our margins. Besides our customers, the leaders of the coworking spaces and entripinurial hubs are smart enough to figure out how to sign up for our service.

So the next step is to figure out what marketing activities will push users to the site. Then what landing page will lead them to sign up. Then what steps and tools will help them get their communitiy fully on boarded.

Currently we have tasked our biz dev machine(who turns out to be about as process oriented as myself) with designing this process. I will work with him and our lead product designer to bring his proprosed scalable customer aqusition process together with the product then test the results.

If we find that x dollars spent on marketing funnels y customers through the signup page and they are likely to have a total lifetime value of z Then we just have to look at the ratio of x to z. If it is as we hope, and z is exponentially larger than x then and only then is it time to scale this and really go to market.

Finally Go to market strategy:

Once we know for a fact...

• We solve real pains and provide a value
• We can monetize in a scalable way
• We have a system to get customers that is scalable

Now it is time to go to market and focus on growth, not before! If we focus on it before hand its just blind luck.

Other things to note:

Controls:

I try to be as scientific as possible. I am no acedemic but I do understand Sceintific Method, specifically Sceintific Controls(or at least I have read the wikipedia pages...).

In order to properly conduct an experiment you need to isolate variables and change only one variable at a time to get accurate results. If you change several variables and get the desired result then how do you know which variable is the magic one.

Changing only small incremental changes instead of scrapping your entire plan or large chunks of your product every time to fail to get the results you want requires discipline. I strongly suggest being methodical and not reactionary. If you are going to pivot pivot one particular part of your product first measure the result and use that to decide what to pivot and how much to change.

Establish a baseline:

Once you have isolated your variables you want to test you need to start out by establishing a baseline. As we speak I am slack chatting with our sales guy. I intend to quietly push out a new feature that I can use to gather essential data about our proposed business model. He requested that we I let him know ahead of time so we can do a big announcement and push people into it. I told him that is a bad idea. First I want to simply put a small call to action link on the side of our TownSquare news feed. Ideally I will track the click throughs and see how many got through step 1,2,3 that would lead us to making some cash. Once I have those initial click through rates I can calculate that:

1. x% of our total active users click through.
2. y% of the user on that page engage with the activity that make money
3. Each engagement is worth $z to us.

Once I have that base line then we can tweak other variables like trying to increase x, or y but without that baseline we have no frame of reference and therefore can not be smart about what we need to change next in order to increase the efficiency of our model(and make more money).

One is Luck, Two is chance, three is ....

We have had some interesting successes. Early on we have a big success with our bigest community on TownSquare, 100State but does that mean that we proved anything? No. Like I said earlier there are other variables that need to be examined, for example we started 100state so obviously we had some really strong ties. We did NOT start every other coworking and eutripural hotbed on the planet so it is NOT safe to assume we will have the same results.

There is a saying I use when coaching gymnastics and an athlete completes a new skill for the first time: "One is luck, two is chance, there is skill". This means if they complete the skill once they got lucky so if they want to cement the skill in their arsenal of skills they need to get up and try it again to be sure. If they can complete the skill a second time chances are it's not luck but its still not enough to say "I got this". It's only after completing the skill 3 times (ideally in a row) that they can say confidently "I got this".

SIDE NOTE: It's always a huge adrenaline rush when an athlete rocks a new skill for the first time

The same goes for startups. We have one big success, but we kinda got lucky and there were definitely outside variables that will not be the same for other customers. So we need to have another big success. Even then that could be just chance. When we have 3 large successes then we can say with some confidence we might have the magic formula.

So get to it, fail faster, but you need at least 3 wins under similar circumstances to prove you've got it. Some scientists would argue you need a lot more to be certain and I agree but if you're in startups you don't really have the luxury of a huge test bed. If you're still alive and kicking in 3 years then we will really know if your hypothesis was right, that is the real test.

Conclusion:

This is probably why people look at me like I am crazy. Perhaps it is because when I shoot down features that would obviously help us scale they don't understand why. Well now you know. Scaling without knowing exactly or even approximately how is a fool's errand. Figure out your numbers.