n-of-1 is better than n-of-none

Figuring out what your users want when you don’t have any users.

We thought we were off to the races. Our minimum viable product was ready to go. Our funky little landing page was up and running. We were tracking our funnels and had a sweet analytics framework.

Things seemed great.

We were working out of a living room in an old victorian house. 5 people all pushing to build something great. Something that would bring real value to anyone in a competitive industry. We were pumped full of unbridled optimism.

Then we launched!

But guess what? Like so many thousands of other startups, our initial launch was underwhelming. When you’ve got so much energy and so much optimism, just about anything other than hitting the moon will feel a bit disappointing.

I could count our users on my fingers and toes. Almost none were actively engaged with the product.

We took the bullseye approach to marketing. Ran small tests on tons of different channels. But we couldn’t get traction. It felt like everything we tried fell flat.

But, every once in a while, we’d get a little “win”. A user here or there. Maybe an occasional snippet of positive feedback.

Making the data work

With those wins we started collecting a small data set… emphasis on small. There were times that our sample size was 1 person. In the beginning it was never big enough to have a high confidence level. Our data scientist reminded me of this very very frequently.

But what I learned was that we had to work with the data that was available. We could fill in the gaps with intuition.

Before I started my startup career I got really lucky and had an opportunity to have breakfast with Marc Randolph (co-founder of Netflix). One of the things that stood out in our conversation was his advice about making big decisions with barely any information.

“What characteristics have I seen in great entrepreneurs? One of the biggest is having a solid strategy for making good decisions with limited data.” He told me about his experiences having to steer his company without enough information to be confident in a positive outcome. Navigating those types of situations is a hallmark of great startup leadership.

Work with the data you have

So I put his advice into practice. I took what we had and made the most out of it. I figured out the profiles of every user that converted. I figured out where we were achieving a small (but respectable) viral coefficient. And we even started doing things that wouldn’t scale, as long as it meant adding one more user.

When it came down to it, we were looking at small numbers, making big assumptions and fighting hard to bring on one user at a time.

Our user journey. It taught me 3 big lessons. 1) Talk to your customers 2) Use your own product and map out the user journey 3) Find ways to get more data.

In the process I learned some important lessons.

  1. Start a dialogue with your existing and target customers.
    — you may not have many users, but you need to understand the ones you do have. Like really really well.
    — because you have so few users, they may not represent your “ideal user”. So, go find your ideal users. Make friends with them. Read their blogs. Talk to them on twitter. Ask them for feedback. Bring one on as an adviser. Do whatever you can to understand their needs.
  2. Understand your user’s experience (BECOME YOUR OWN USER).
    — when you don’t have users, you need to become one yourself. You should be using your own product/service…everydamnday. It’s the first thing you should do when you wake up in the morning. Every day ask yourself “is this really useful?”, “does it actually add value” and “would I honestly use this if I hadn’t built it?” (if not, why?).
    — you should know every bug before a user reports it. A good way to do this is to treat yourself like the QA guy. Think through every interaction possibility and test it out. Do this for every step of the users journey. Start with onboarding process for a new user, then move to using every feature in your product/service.
  3. Get more data, even if you have to pay.
    — there are lots of hyper-targeted survey services out there. You can pay a very reasonable price to get questions answered by your target customers. This can be a great way to prioritize features, pricing, marketing channels, etc. Do a bit of research on building effective surveys before you start. If everyone who responds says they’ll only pay $20 for your $1000 product, you may need to rethink your pricing (true story… this happened on one of my side projects).

It’s all about building context. A better understanding of your target users plus a better understanding of your own product equals a better understanding of your data. When you see that sudden drop-off in your users at a certain step in your funnel, you won’t just see a percentage, you’ll see an entire experience.

Using it in the real world

My favorite way to put this newfound knowledge into action? Using Google Analytics Behavior Flow chart.

If you have Google Analytics this chart is under the behavior section in the lefthand navigation bar.

The behavior flow chart becomes exponentially more useful the better you understand your users and product. Even if you only have a few users, you can quickly and efficiently explore ever single stage of their journey.

The process is simple. You’ll see a red symbol next to a stage where a dropoff occurred. Click it to highlight the users journey of anyone at that stage. At the top of the page you can switch between page flow and event flow (or both).

On the left you can choose how to segment the flow. I prefer to use source/medium so I know if I need to address issues with a specific referrer. Want to hone in on one specific users journey? Use a filter and restrict the traffic to their network or their specific city.

Spend 30 minutes exploring this chart and I guarantee you’ll find at least one opportunity to improve your conversions, user retention, or bounce rates.

So, is this something you can work into your routine? I know it’s a lot easier to talk about than it is to actually do it. But I honestly believe it will pay off in dividends for those just getting started and working with barely any data. You’ll be exponentially better at finding signal in the noise. It will make every insight more actionable. Try it out!

If you need any help getting started feel free to get in contact. I’m happy to help. cottrellconsultancy (at) g mail (dot) com

Data Studio

50% Data 50% Design

Josh Cottrell-Schloemer

Written by

Ex-Entrepreneur. Startups acquired=1. Comp Intel and Monitoring Expert. Google Data Studio Consultant. Based in Tokyo/Chiang Mai. Founder-G2 Performance Agency

Data Studio

50% Data 50% Design