The Lean Start-up Ninja (Part I) – Your idea sucks, but if you are Agile you can charge on!

 In Agile

Why this series?

We’ve often been asked what it’s like working in a start-up in the new Silicon Roundabout area in London UK. Our answer is Well pretty much like start-ups everywhere else we’ve seen! To be fair we are lucky in that we’re involved with Shoreditch Works, who are nothing short of awesome). Despite cultural or environmental differences, the truth is that all start-ups share things in common:

  1. Challenges (“can haz revenue?”)
  2. Fears (“Can haz revenue?”)
  3. And hopefully Successes! (“haz revenue!”).

However, start-ups are unique and people on the outside want hear about them, because it all seems so liberating and romantic…. right? Ben and I recently spoke about our experiences covering a host of topics from your idea, being lean & agile, finance, technical, product development, start-up culture and more!  You can find the full “The Lean Start-up Ninja” talk on SlideShare. As per usual with our talks there was a heavy dose of British humour, we certainly think the whole lean start-up Ninja hype (as portrayed in some Hacker News threads) are complete nonsense, the reality is this:

“Ninjas and rock Stars leave behind dead bodies and trashed hotel rooms”

Still, there was something appealing about this, even when we were 6 years old…..

3014649-cast-of-teenage-mutant-ninja-turtles-5

This is the first post in the series, covering how your idea doesn’t survive its first impact with the market, the reality of pivoting and why being agile matters. Of course you should take any of the advice here with a grain of salt, jClarity has been around for 18 months which means we’ve bettered the 7/8 start-ups that fail within their first year, but we’re not claiming to be world experts in this space. In particular we’re going to cover this from our personal experiences which we hope will provide a little more context to the generic advice out there.

What people think Lean Start-up Ninja life is like

Colombian_Party_Cartel._Photo_by_Ellen_Ordóñez_for_Colombian_Party_Cartel

Image courtesy of Ellen Ordóñez

Sorry, but that scene is unlikely to happen. Unless you’re the 1 in 1,000 that get serious funding or the 1 in 10,000 that has a huge exit. However, you can grow a successful, sustainable business and that should be celebrated even more than the occasional moon-shot as it actually helps grow a sustainable start-up ecosystem. For us this meant avoiding the hype around the new Silicon Roundabout, with it’s endless rounds of nightly Meetup.com events, visiting Members of Parliament and tech giants coming in to nurture (read ‘consume’) new start-ups in the area. Instead we knuckled down, joined like-minded folks at Shoreditch Works and set about trying to create some amazing products in the Java/JVM performance analysis space.

Your idea and how it’ll change the world

Basically your idea means almost nothing, so don’t be overly precious about holding it close to your chest. In a world of 7 Billion+ people, chances are, someone else is working on the very same thing. The takeaway here is that building a team, a product and a company around that idea is what really counts.

“Your idea is not a unique and beautiful Snowflake”

12309476-brains-in-the-light-bulb-isolated-on-white
bill-murray-you-suck
Your Idea! The truth!

Images courtesy of www.comicvine.com and www.123rf.com

For our first product (Censum, the Java /JVM Garbage Collection log analyser) we had the leading product in its space due to us adding in real analytics for the first time. Instead of providing graphs and raw data, we also gave plain English answers to memory problems based on some analysis we were able to perform. About 2 months later, we saw an immediate response from the competitors in our space. They very rapidly set about trying to close the gap between themselves and some of the new innovations we’d added into the space. In a few months we’d gone from being clearly the leading product in the market to only retaining a narrow lead in terms of a feature set, a first amongst equals as it were. Although it was tempting to fall into despair, we simply took the attitude of “Let’s go one further” and are now building the next generation tool that simply no-one on the market has provided yet.

For our second product (jClarity, the on-demand Java/JVM application performance analyser) we’ve seen a host of new start-ups begin to follow in our direction. Again here the concepts of using Machine Learning to analyse complex performance issues and provide plain answers was new to the market, but certainly won’t remain that way. It’s not an easy domain to work in and already we’ve seen a couple of small interesting efforts by other individuals and teams fall by the wayside as the effort to take the idea to market is realised.

We’re retaining a philosophy of smart people and sharp tools in order to build ideas into reality, the next difficult piece is to get the marketing and sales right, on the good old limited start-up budget of “very little”!

Pivot like there’s no tomorrow

Start-ups transition…. a lot. At jClarity we’ve transitioned products twice and markets three times as we explored the space around a gap for Java/JVM performance analysis tools and the emerging trend of main-stream enterprises moving into the cloud. If you have a business plan, please make sure it’s of the Don’t make my eyes bleed variety, anything further than that and you’re probably just wasting ink.

“The street finds its’ own uses for things” – William Gibson

This will happen to you, a lot. The good news is that it’s a sign of a successful product! We laugh at our original business plan :-).

In particular we thought the jClarity product was ripe for the emerging Java/JVM Platform as a Service (PaaS) market, in 2012 about 10-15 new players had arrived! However, the reality was that consumers weren’t ready for Java/JVM PaaS in 2012/13 and in fact are still coming to terms with it in late 2013! This meant in order to stay alive, we had to go back and focus on Enterprise customers that were looking for a new light-weight, accurate performance analysis tool and also sell more of Censum to help customers solve at least one major performance problem.

Lean and Agile like a Mongoose

So there are a bunch of lean/agile practices that have helped jClarity pivot when required, and remain afloat as economic factors and markets changed.

  1. Daily Stand-ups – Gives you a 24-hour warning on challenges that arise and are great for team building (share those successes!). We actually have 5-minute ‘sit-downs’ as that’s what the team is most comfortable with.
  2. Weekly Retrospectives – The more open and honest the better. Start-ups tend to be tight-knit families and its good to let off steam and forgive each other before going off for the weekend (for those lucky enough to get weekends). We have a 10-question retrospective covering points such as: “Did we keep the customer happy?”, “Are there any Trust and/or Support issues in the team?”, “Did we keep it simple?” and so on.
  3. Small, 1-week, achievable sprints – Crunches == bad code which leads to a crap product, don’t fall into that trap.  We have 1-week sprints usually focused on something we can deliver to a real customer such as a new analytic, a UI refresh or similar. There are exceptions to this, especially when we are working on some real R&D work such as retraining the jClarity analytic engine.
  4. Honesty with clients – We tell people what really happened when we screw up. The response is overwhelmingly supportive and we keep customer’s we’d otherwise had lost! One one occasion we lost customer production data, something that we’re not proud of. However, we immediately owned up and took the customer step by step through how we’d lost the data, what steps we’d taken to ensure it wasn’t repeated and then prioritised one of their feature requests as way of an apology. Because of this approach, there was no public backlash, no Lawyers thrown at each other or other horror stories.

The opposite of Lean and Agile

We got plenty of things wrong and it’s the #fail stories which are the best right! All of the items below basically boil down to:

“Get revenue first, revenue allows you to solve all of your other so called problems”

  1. We didn’t validate our product against real end users from Day 1. A classic mistake! We’ve since opened up the ways in which you can trial our product, removing Credit Card barriers etc.
  2. We didn’t aim to generate revenue from Day 1. 2nd classic mistake! – always have revenue coming in, preferably from your product(s) from day 1. We took ~6 months to start
    gaining revenue from our first product (Censum) as we’d wasted time on a technology preview on something that was never going to have mass market appeal.
  3. We didn’t bring in a Sales/Marketing lead from Day 1. We’re now taking some great mentorship from our friends at ZeroTurnaround, but nothing beats having marketing ingrown into your culture from day one.

In the next part we’ll cover How to be a Ninja, discussing the automation tools we use, how we treat each part of the business and the all important criteria in picking the right technology.

Thanks for reading, comments welcome!

Martijn (CEO – & and the jClarity Team)

No more memory leaks and application pauses!

Find out why my app is slow and tell me how to fix it!

Recent Posts