a canadian startup

Always Be Nimble

After reading this post by StartupNorth I decided to finally write out this article on a related subject.

We tested out a new revenue generator yesterday, and the preliminary tests are wonderful.

Joy.

This is the greatest rush in innovation: you try something new — a neat little trick that is just beyond obvious — you test it out, and it works. Holy macaroni, it works! Why doesn’t everyone do it? Shh.. don’t tell them.

One of my most wisdom enhancing experiences this past year was at the Founders & Funders dinner organized by Jevon and David. I spent the entire night sitting with Mark Skapinker, one of the bestest and inspiring guys in the Canadian startup scene.

His biggest lesson to me came in the form of an observation he made about a strategy that startups must take in order to be successful. Startups must be agile.

We live in an age, Mark told me, unlike any before with regards to the ease at which we can invest small amounts of time and money into tests for the marketplace. We can spend a day building a prototype of a new site and seeing if people will go for it, we can change the colors on our site, or build a new login mechanism — all with minimal investment. Do it quickly, cheaply, and see if it works.

The best startups are doing this constantly — testing the waters until they find something that works. Set success criteria and measure your results.

From watching startups present their business plans, one of the things that scares me most is when I see startups making guesses, requesting huge sums of money to see if they work, and doing it without shame. You need how many million dollars before we can even be sure this is a real business? Huh?

It’s not that I mind that you’re probably going to waste a lot of other people’s money. It’s your family, significant others, and your children that I really care about. Please don’t do this to them.

You need to build a plan that is nimble — one where you can do multiple experiments for low cost until you find the one that works.

An example — when we were first pitching the idea of Well.ca, people asked us how we were going to market the site. We could have come up with ideas: we’re going to spend 500K on Facebook ads, and another 500K on making fake blogs that point to us. But as long as these ideas are unproven, you might as well spend 500K on magic beans.

What we did, instead, was to allocate tiny amounts of money to specific testing grounds, and we set up metrics right from the start to see if they’re working. We can test out the university market, for example, with $100 student newspaper ad with a special URL in a single chosen university. If it works, we can do this for every student newspaper in Canada. If not, we’ve only lost $100.

If you look at our site, you can see this in different places: we spend minimal amounts of development effort testing out some new ideas, such as customer reviews, and sat back to see if they work. If they work, we’ll fix our prototypes and make them into full-featured products/features. If not, we just wasted a couple of hours of AJAX scripting.

The strategy of quick prototyping with assigned metrics is also the greatest way to win arguments in a startup. Let’s say someone has an idea that you think won’t work. The ultimate response is: “find a way to build a prototype in half a day, if it increases sales, I’ll believe you”

If you’re right, you’ve only wasted half a day of dev time. If you’re wrong, well you can use the new revenue to buy jewelery engraved with the words “I’m sorry for ever doubting you.”

It’s also a great way to get funding. Instead of pitching an idea, you can show that what you’re saying already works. (And all you need funding for is the Aeron chairs).

My experience with this kind of business model is that the end result is that you will discover one thousand new business practices that are lousy wastes of money. For every thousand, however, you will find one that works magically — usually the one that you thought was too simple to work. I try to spread the same philosophy to new people at Well.ca — sometimes I let people work on projects I am pretty sure will never work (when it’s their idea) but I make sure to say “Yeah, we can try that, put X amount of time into for a week, then we’ll try it out for two weeks and we can say whether or not it is successful only if Y or Z happens”. This teaches people the philosphy of being agile. It also teaches people to focus on achieving results, rather than implementing ideas (which is really the point of this entire post).

Multi-faceted, agile, iterative strategy:

1. Do many simultaneous experiments
2. Invest minimal time in each one
3. Set success criteria
4. Let them run
5. Revisit them (and walk away if it didn’t work)

My other experience with the agile strategy is that many ideas will only work for a certain amount of time. Someone will copy your idea if it’s great, and they will have the advantage of not having wasted time on all the failed experiments.

An example: at one point in time, Well.ca dominated the (Canadian) health market for Google Adwords. It was a new service that I learned right as it was released, and I used every tool available to make sure we used it to its full extent. Our campaigns were designed with two stages: we would have big groups of test adwords and then we would promote them to a different group if they did well. Budgets were small, but if things worked, we would increase them. We did well.

But after a year or so, everyone who was anyone was paying for adwords and our positions went down. Adwords are paid for based on an auction model, and it got to a point where competitors with higher price margins (i.e. competitors that were charging customers too much) could push us out.

The market got saturated. We’ve moved on.

This happens with all new markets.

Friends, your glory days wills soon be over.

So my advice is to continue being agile, always looking for the next thing. Design a business model that sustains continual agility and experimentation.

As Jevon puts it:

Build the product, test the market, sell. GOTO 10