» Ideas Are Easy, Implementation Is Hard « – Guy Kawasaki
I have met numerous entrepreneurs, startup teams and companies with brilliant ideas and strong visions. Creativity, knowledge and multidisciplinary teams foster disruptive ideas.
Implementing disruptive ideas is a complex problem, though. When it comes to complex problems, the relationship between cause and effect, by definition, is not clear. In order to figure out, if an idea is truely a great idea, we have to try it out!
Validated Learning
Both, the lean startup methology and the validated learning process, provide a proven approach for quick and early idea validation. Not until we performed some early testing, could we rank and compare multiple ideas and paths.
While, to a certain extent, validation can be done by employing surveys and paper prototypes, in most cases I would recommend to build an interactive minimum viable product (MVP). This heavily facilitates our decision-making process for selecting the most promising idea.
» The minimum viable product is that version of a new product which allows a team to collect the maximum amount of validated learning about customers with the least effort. « – Eric Ries
With the help of an MVP, we are able to measure and assess user behavior for a potential innovative product or service. By employing tracking technologies – such as Google Analytics or Mouseflow – we gather deep insights and can perform quantiative testing. It should be our goal to enrich these findings with qualitative information. We can do so by acquiring potential consumers for interviews, listen to their needs and dreams, as well as establishing a relationship for testing future versions.
At this point it is usally not required to deliver the actual product or service!
Fake it ’til you make it
Especially in a (pre-)seed stage of a startup it is usually sufficient to “fake” a product or service, in order to gather as much learnings as possible, with the least effort. In a second step, it is recommended to manually perform value creation and product delivery (concierge / wizard of oz test). Not until we validated our product or service idea, should we start to automate our value creation. This might not sound efficient at all. But it is. It tremendously reduces the cost of failure.
» Want to increase innovation? Lower the cost of failure. « – Joichi Ito
Instead of investing thousands of Euros in software design and development, to create the most sophisticated high-tech solution, we walk step-by-step. Based on the first version of our product (MVP) and the learnings from the early market testing performed, we are planning the next iteration (version) of our product or service.
How to NOT build an MVP
A couple of years ago I also used old-school ways to build overengineered or static MVPs that do not support validation of the actual product idea.
Examples
- Do NOT just build a static landing page to validate an idea based on artificial intelligence (AI). In order to prove if consumers will honor automation, you would need at least some interactivity and intelligence in your MVP.
- Do NOT just build a bare MVP without any trust signals. Use a “satisfaction guaranteed” badge, testimonials, user ratings and SSL encryption. It is little effort that pays off: it can dramatically reduce user acquisition cost and increase comparability to your final product.
- Do NOT build a mobile app. Do not get me wrong, apps are great – for certain scenarios or in a later stage. Even with cross-platform development tools, such as Titanium, it is still pricier compared to typical web development. But the worst drawback is: it slows down your early innovation process. Even with the most efficient compiling and deployment process and a hybrid-app-approach, it costs valuable time. Use a mobile optimized website (adaptive or responsive) instead.
- Do NOT use a “hot new” programming language, framework or tool, just because it is “cool” or could increase productivity by 0.1 %. When it comes to IT, there’s more than one way to skin a cat. Focus on a large developer community, well-maintained libraries and availablity of freelance developers and competitive rates. I prefer working with PHP, Java and Javascript, instead of Ruby.
- Do NOT host your first version on Amazon Web Services (AWS). AWS is great for scaling. On the other hand it is not cheap and it takes precious time, as well as some expert knowledge, to configure. Start with an off-the-shelf web space or virtual server. If your business grows, move parts of your infrastructure to the cloud.
- Do NOT hire software developers for “everything that is related to coding”. Get a product manager or marketing ninja who knows at least some basics of full-stack development. It does not take a dedicated software developer to write some Javascript for sending tracking events to Google Tag Manager / Analytics. If there is more work to do: Get a full-stack developer. Place him in your office, with product management on his right side, and marketing on his left side. If there is more work to do, think about hiring a team or outsource parts of your development. Think about having a fast-paced internal IT that works closely with product management and an expert team outside your organisation, for well-defined tasks with an already validated benefit.
Further increasing efficiency
Building MVPs and validating ideas requires deep knowledge in various areas.
Success factors
- A deep understanding of Lean Startup, Validated Learning, Design Thinking and Scrum.
- Great product management and digital marketing skills, as well as entrepreneurial spirit.
- Domain-related / industry knowledge – or at least the experience how to absorb it quickly.
- Data and experience in how to make data-driven decisions.
- Up-to-date knowledge about consumer trends, technology trends, design patterns and digital business models.
- A continuous drive to increase efficiency and automation as well as optimizing business and product KPIs.
Try to find someone, or a team, that combines as many success factors as possible in order to increase efficiency.
Additionally, having a standardized, but still flexible, workflow and a proven tool-set, can tremendously increase efficiency. Especially when the workflow and tool-set itself are continously optimized.