Building products people want


A talk I gave at London College of Communication on my experiences with startups & building products that people actually want. My second outing in public speaking and it really shows in my nerves. I think I still need much more exposure before I’ll feel comfortable on stage.

Raw Transcript

Good afternoon everyone, My name’s Deon and Today I want to talk to you about startups, failing fast, building products that people actually want, and if you’re really lucky making some money in the process. These are some things that unfortunately I’ve had to learn about the hard way — quite a few times now, as a result of starting my own company. But before we get into any of that let me give you some background for context.

About a year and a half ago I co-founded Ocean Labs with Fionn (& Charlie) off the back of some cool prototypes and an initial seed investment. We’re a product focused company. We currently have three products in market:

  1. Huggleup - Huggleup is a straightforward business to consumer e-commerce website. It allows you as a user to turn your photos into all manner of personalised products, things like magnets, postcards, posters, stickers, frames and so on
  2. Kite - Kite is the platform the underlies Huggleup. It’s a B2B company. It allows you as a business or independent developer to offer our products to your end users in a way that’s totally branded to you. You can charge them what you want, we’ll take a small cut and deal with some of the tricky product creation, packaging and shipping stuff.
  3. Loci - Loci is a totally different proposition from the previous two. It’s a social play and in its current form it allows you as a user to get a realtime view of social content coming out of a given geographic area.

In the early days of Ocean Labs we invested a significant amount of time in into building beautiful products and features that nobody wanted. And beautiful is maybe debatable by modern trends so forgive some of these screenshots. Our first product was Ps Postcards, it was actually and somewhat surprisingly the one that allowed us to raise our initial round of investment - It was a simple app the allowed you to turn your photos into physical postcards and have them posted anywhere in the world all from your iPhone.

This app, at least in our totally unbiased opinions blew all our competitors out the water in terms of user experience and functionality. It had all the bells and whistles and we invested a great deal of time in polishing the experience with cool subtle animations, and features to wow our customers. All in all it took us around a year from when we first started dedicating our time to the idea to to an app store release. Only the last three months of that timeframe was full time work but it’s not unrealistic that a full man year or so of time across UX, design, project management, development and testing was invested in the project. That’s an awfully long time and a massive opportunity cost if the project fails to gain any traction. So how much money did we make for our time investment?

Just over £3000. Which for a business with three full-time employees at that time may as well round down to £0. Let that sink in, approximately 1 man year of work resulted in a measly £3000 in revenue. If you subtract COGs gross profit is near £1500. For every hour of my time I invested in that project, I earn’t well below £1 for my time. That’s quite a lot below minimum wage.

To give you some perspective. Before starting Ocean Labs we used to build mobile applications for clients at several digital agencies. If a client were to approach us with a project involving similar effort and team it would likely be charged out at several hundred thousand pounds. So we were failing big time in comparison.

What’s worse though is that after the reality sunk in (the fact that we weren’t going to be overnight postcard millionaires) we still kept blindly grinding away building products and features that no-one wanted for several months following that app store launch. We’d have relatively long design-build-release cycles typically taking north of 1 month to get anything new out. Before yet again discovering no-one really cared and that feature we just built had no significant impact on our bottom line. And we’d repeat this process. We were very good at failing slowly.

A startup is a temporary organization designed to search for a repeatable and scalable business model – Steve Blank

I think we were (and arguably still are) a startup in the truest sense of the word. And that is a temporary company with effectively no business model. But one that’s searching for a scalable, repeatable one. We were trying to sell what we could make, not making what we can sell. I think it’s a trap many people fall into when taking seed investment to start a company, it’s why so many companies fail.

So if investment buys you a finite runway (or if you’re bootstrapping your company and thus it’s simply time investment and opportunity cost) how can you find that scalable, repeatable business model and get to traction on your journey to product market fit as quickly as possible?

That’s a question I was definitely pondering at the beginning of this year. I think all of us in the team were looking for and answer at that time and it’s probably not all that surprising that we all seemed to discover upon the lean startup movement somewhat simultaneously.

The Lean Startup provides a scientific approach to creating and managing startups and to get a desired product to customers’ hands faster – Eric Ries

So what does (the above quote) mean? Well the methodology was started based on the theory that too many startups (ourselves included) begin with an idea for a product that they think people want. The startup then spends months, if not years, building the product, perfecting it, before actually taking it to market and without ever showing it to the potential customer. When they fail to gain that traction in market, it is often because they never spoke to potential customers to determine whether or not the product was actually of interest. They never got out the building and spoke potential customers. This is a recipe for failure, and is exactly what we were guilty of in the early days of Ocean Labs.

So it is definitely something I can relate to. And it’s something that the lean startup methodology attempts to address by providing process or framework that can be used to guide your company throughout it’s journey.

So lets look at this process or framework in a bit more detail. A core component of Lean Startup methodology is the build-measure-learn feedback loop.

build measure learn feedback loop

The 1st step in this loop is figuring out the problem that needs to be solved followed by developing a minimum viable product (MVP) so that we can begin the process of learning as quickly as possible. Once we’ve developed the MVP, we can work on tuning the engine. This will involve measurement and learning and must include actionable metrics that can demonstrate cause and effect question. Although it’s referred to as the BML feedback loop, you actually want to progress through it in reverse. You want to take the validated learning approach.

The first step in this approach is deciding what it is you want to learn, you want to form your hypotheses that you are going to either validate or invalidate by carrying out an experiment. For example: lets say we’ve identified that our postcard proposition is failing, as a team we get together and brainstorm potential directions we can pivot the company. We decide that we want to explore selling personalised shoes. We hypothesise that personalised shoes are what the market wants.

Next up we look at how are we going to measure whether our experiment is a success or failure. What need to identify the actionable metrics we’re going to use to validate or invalidate our hypotheses that the market want’s customisable shoes? Perhaps we could look at the sign up rate or revenue within a given timeframe. How does this compare with the equivalent Postcard metrics, is it equally as lacklustre.

Regardless of your metric it’s advisable to draw some lines in the sand ahead of time that are going to define the success or failure criteria of the experiment.

Drawing these lines in the sand ahead of time is quite important. Without them you can’t really tell if you’re making adequate progress fast enough to meet your goals. They can also give you an idea as to how much effort to invest in your current course and direction, whether knuckle down or try something else.

Next we want to build a minimum viable product to support our validated learning. What’s interesting is this might not actually involve any development work, in fact I’d personally build or develop as a last resort now. You need to ask yourself what’s the minimum amount of effort required to validate or invalidate the hypotheses laid out earlier.

Given past experience this is usually the most time consuming part of process as it’s a comfortable place for most teams to be. Sell it before you build it if at all possible.

The quicker you can get through this loop the more opportunities you have of getting to product market fit within your remaining runway. You absolutely want to minimise the amount of effort you expend here so that you can fail faster and give your company the best chance of success.

Going back to our personalised shoe proposition, we might approach the build phase by throwing up a quick landing page on the web offering personalised shoes for sale. We’d probably and add a “Get Started” or “Buy it Now” button. Clicking this button could just inform the user that the product is coming soon and show a form to sign up to be notified when it’s available.

To minimise effort we could avoid building out any of the actual personalised shoe printing functionality. We might even be able to avoid wasting any time talking to shoe suppliers and rather focus on driving users to the landing page to help in validating our proposition. If we get the kind of interest that we want based on the metrics we care about then we can explore building out the proposition further. If we fail to gain the traction we’d like then we’ve invested little effort and the opportunity cost is relatively low.

So how have things changed at Ocean Labs now that we’re a little bit leaner.

Well lets take a look at how we actually applied this lean startup process to our latest product proposition: ZenCam

ZenCam is trying to bring back the disposable camera in digital form. The idea is that the app will have a digital camera reel that functions not unlike a disposable camera. You’ll only be able to take a limited number of photos, say 20 before exhausting the reel. When you’ve taken your 20 pictures and exhausted the camera reel we’ll print and deliver those photos to your door. You then have to buy another digital camera reel if you want to take more photos. The idea behind the proposition is that if we limit the number of photos you can take, you’ll slow down and take more meaningful pictures.

Now the old Ocean Labs would have jumped straight in to development and built the proposition out in it’s entirety. Being a little bit wiser now though we decided that our MVP would be a KickStarter campaign. So we’ve thrown up a beautifully designed page with an awesome video to get the proposition across. We’ll let customers vote with their money to validate whether this is actually something we want to build out further. Essentially we’re trying to sell it before we build it.

We were failing way before we adopted several lean startup principles. And this stuff should be taken with a pinch of salt, there are no silver bullets. But now we’re failing slightly faster and with a bit of luck hopefully we’ll find that success before we run out of runway :)